Frontier Computing on Industrial Applications Volume 1

In order to realize automatic detection of surface defects of frozen dumplings, a U-shaped semantic segmentation ResNet_Unet based on U-Net is proposed in this paper. ResNet50 is used as the encoder of the semantic segmentation network to enhance the feature extraction capability. In order to enhance the recognition accuracy of small targets, expanding the receptive field of the multi-scale output feature map of the encoder is proposed. And embedding BN normalizes the data before decoding module convolution, which can speed up training and improve generalization of the model. In addition, the label is smoothed when the loss is calculated to prevent the model from placing too much faith in the predictions. The experimental result shows that the mPA and mIoU of the ResNet_Unet network are 83.20% and 80.44%, respectively. This study provides a reference for intelligent segmentation of surface defects in frozen dumplings.

With the rapid development of information technology, information technology has gradually been incorporated into financial management, and more and more enterprises are using financial management in their operations. The financial management level in today’s enterprises is relatively low and cannot meet the needs of the enterprise, so it is necessary to further upgrade financial management. This article studied the application of distributed database systems in financial management, aiming to further improve the level of financial management through distributed management systems. This article tested the efficiency improvement of financial management after using a distributed database system through experiments. The experimental data showed that the efficiency has improved by at least 13% and the highest by 19%. Through this experimental data, it can be proven that distributed database systems can indeed have good results in financial management.

Under the influence of the “dual carbon” strategy, China’s hotel industry has gradually begun to transform into green environmental protection. The research was based on literature and interview research, 13 influencing factors of hotel customers’ green consumption behavior were identified by the Delphi method. The decision laboratory method (DEMATEL) was used to aggregate the 13 forming factors into four-factor gathers: strong cause, weak cause, strong effect, and weak effect. Furthermore, the multilevel hierarchical model of the factors that affect hotel guests’ green consumption habits is adopted to explain the structural model (ISM). To research the elements that influence hotel guests’ green consumption habits and systematically assess the relative relevance of the many influencing factors, it is of great significance for hotels to attract customers to produce green consumption behavior. The results indicate that the green consumption behavior of hotel customers is the outcome of direct factors, intermediate factors and deep factors. Dual carbon strategy, green certification, green fiscal, and corporate green marketing are the primary factors affecting the formation of customers’ green consumption behavior, which should be paid attention to.

Considering the proportion of stylistic skills in language communication, this article combines the characteristics of Chinese and needs to establish a correct positioning in teaching Chinese as a foreign language. The IAST-A (Interaction Audio-Video Share Text Activities) model constructed in this article uses social networks as a medium, and its core element is “activity”. This project requires mastering Chinese listening, speaking, reading and writing skills, but also emphasizes the ability to communicate in Chinese. The Chinese as a Foreign Language Education System based on cloud computing in this article refers to the combination of English education and English education based on cloud technology, making English education more intelligent, conducting listening teaching through cloud platforms, and conducting multi-channel interactive learning. This paper uses fuzzy neural network method to evaluate the quality of teaching, including the evaluation of teachers, students, and teaching managers. In this article, the proportion of students in the first 80–90 score range using the IAST-A model is 8.4%, while the proportion of students in the 80–90 score range after using the model is 43.5%. This reflects the effectiveness of the IAST-A model for teaching Chinese as a foreign language in this article.

Many countries list Russian as their second or third foreign language, and the demand for learning Russian is constantly increasing. People have higher requirements for intelligent information processing, automated management, and digital communication, and they are more enthusiastic about the development of the Russian Massive Open Online Course (MOOC) system. This article used neural network algorithms to study the MOOC system, which was helpful for the teaching of Russian majors. This article analyzed the performance of the Russian MOOC system through experimental design and algorithm comparison. Experimental data showed that the convolutional neural network algorithm had the highest accuracy among the four algorithms, reaching 95%. Therefore, convolutional neural network algorithms can be used to develop Russian MOOC systems.

Tourism industry, as one of the important driving forces for economic growth, can greatly promote regional economic development and accelerate the positive externalities of resource flow among regions. Finance is a vital component of modern economies, playing a critical role in allocating resources and driving economic development. Firstly, the study will conduct data analysis on the current situation of tourism development and financial support in liangshan prefecture. Secondly, an explanation of the vector autoregressive model will be provided; then, the vector autoregression model will be used to evaluate the coordinating relation between tourism development and financial support.

With the development of the times, words play an indispensable role in daily life. Therefore, text recognition detection is very important. In recent years, although there have been many research achievements and progress in text recognition detection, there are still situations of ambiguity and distortion in text recognition. Therefore, it is necessary to further upgrade the text recognition detection algorithm. This paper studied the application of artificial intelligence (AI) technology in text recognition and detection algorithms, aiming to further improve the accuracy of text recognition and detection algorithms through AI technology. This article tested the improvement of text recognition detection accuracy using AI technology through experiments. The experimental data showed an improvement of at least 11% and at most 19%, indicating that AI technology can achieve good results in text recognition and detection algorithms.

Intelligent environmental protection refers to the intelligent form of utilizing environmental data information, including comprehensive collection and analysis, continued development and utilization of environmental data, and achieving the goal of supporting environmental management. It is conducive to reducing pollutant emissions, and promoting environmental legitimacy, thus creating an ecological city that is harmonious with the environment and social economy. The development of networked technology has brought many opportunities for production and human life. With the emergence of environmental problems, they have become a global problem, especially severe in Chinese cities. Therefore, in order to achieve rapid socio-economic development, it is necessary to introduce an intelligent environment that is conducive to building green cities. The research on data mining based on neural network-based smart environmental protection big data was of great help in improving the data mining performance of smart environmental protection platforms.

With the continuous development of livestock and poultry industry, monitoring of livestock and poultry farms has become more and more important. The monitoring of livestock and poultry farms mainly includes instrument monitoring and environmental parameter monitoring. The purpose of this paper is to study the farm intelligent monitoring system based on human-computer interaction and automatic control algorithm. In view of the real situation and problems encountered in the industrial development of China’s livestock and poultry farming industry, the current situation of environmental monitoring of livestock and poultry farms at home and abroad is analysed. Based on the research of wireless sensor network and human-computer interaction technology, a livestock and poultry environmental monitoring system based on automatic control algorithm technology was designed and developed. The automatic control algorithm was tested and the experiment showed that the overall error of the system was small.

By investigating and analyzing model in language learning and the present situation of English teaching system, we put forward an architecture design of English teaching system based on deep learning model. The system mainly includes three parts: teaching content management, learner behavior analysis and intelligent assisted learning. In the teaching content management, we use the deep learning model to identify, analyze and sort out English teaching resources, and recommend the most suitable learning resources for students. In the aspect of learner behavior analysis, we can analyze and evaluate learners’ learning habits, interests and abilities by collecting and analyzing students’ learning data. Finally, we apply the deep learning model to intelligent assisted learning, and realize adaptive learning and personalized teaching, thus deep learning model has good practicability and popularization value.

In recent years, with the rapid development of computer technology, corpora have played an important role in promoting the research of Chinese, English, and languages around the world. The construction of corpora has also attracted widespread attention at home and abroad. Corpus takes conversational language as the research object and establishes relevant discourse corpora, which helps people express the structural rules of language more formally and computationally. This paper introduces the corpus teaching system based on Internet technology to non English majors to verify the specific effectiveness of data-driven learning method for English vocabulary learning. This article takes 100 students from two classes in a certain university as the subjects. Through experiments, it is found that under the corpus teaching method of the college English teaching system, students spend 34% more time memorizing 10 English words on the same day than ordinary teaching methods. One week later, the number of students who can remember the same words using the corpus teaching method is 42 more than that using ordinary teaching methods. And the corpus data-driven learning method in teaching takes shorter learning time. The experimental results show that the corpus based teaching method of college English teaching system under the Internet environment has a good role in promoting English vocabulary learning.

Machine translation (MT) is an advanced technology that automatically converts the source language into the target language through the use of computers. As communication between countries around the world becomes increasingly close, the need for mutual translation between languages is becoming increasingly evident. In view of the low accuracy and poor vividness of the MT model, this paper uses neural network (NN) to optimize the MT model. The accuracy and vividness of MT models can be improved by using NN. Through the optimization of English MT model based on NN, the accuracy rate has increased from 87.3% to 96%, a full 8.7%. The accuracy of the traditional MT model has increased from 87.5% to 88.1%, an increase of only 0.6%. The experimental results show that the accuracy of MT can be effectively improved by using NN to optimize the translation model.

In this rapidly changing era of information technology, images have always been the main tool for human communication of information. In daily life, a large number of natural scene images, such as road signs, advertising slogans, etc., often contain important textual information, which is of great help to people’s understanding of image content. Text is a type of linguistic information that exists in various aspects of people’s lives. The textual information in it can help people understand, describe, and analyze natural scenes, and has important social value. Neural network technology is of great help for optimizing text detection. This paper discussed the optimization performance of neural network algorithm for text detection, and verified the progressiveness of the algorithm through experiments.

In this paper, a prediction model based on Empirical Mode Decomposition (EMD) combined with Long Short-Term Memory (LSTM) is proposed, which can further improve the accuracy of model learning prediction based on only using LSTM. Before model processing, it needs to find indicators that can evaluate vehicle fuel consumption, and then select the indicators that can be learned by the neural network as the initial input. The initial input is substituted into the EMD model, and then the IMF component obtained through EMD decomposition is substituted into the LSTM model. Taking the data of cars of different brands as practical examples, the test results show that the prediction error of this method is only 12.2%. But the prediction error of only LSTM network model is 15.4%, the average absolute error of the former is 7% higher than that of the latter, and the prediction accuracy is higher.

Greenhouses can ensure the normal growth of crops in extreme environments. With the popularization of social intelligence, how to build an efficient and accurate environmental monitoring system to ensure that crops have suitable growth conditions is a top priority. In order to effectively solve the problem of data missing and large data errors caused by sensor collection in intelligent greenhouses under extreme environments, this paper uses data preprocessing and multi-sensor fusion algorithms to ensure the accuracy of data. The effectiveness of the data fusion algorithm is verified through data transmission experiments.

New media, as an emerging means of communication, has become an indispensable part of people’s lives. It plays an important role in the film and television industry, advertising industry, and cultural research. This article analyzed the internal aspects of online new media. Firstly, it elaborated on the current development status and future trends of new online movies. Secondly, this article also focused on the problems and solutions in the process of intelligent film and television dissemination using Internet Protocol (IP) as the carrier. Finally, a solution and specific implementation methods for building a new media platform using IP technology were proposed, including establishing a complete website system, optimizing website structure, and updating webpage interfaces. The performance of the model was tested. The test results showed that the anti-interference ability of the model ranged from 92% to 95%.

The main idea behind this method is to build a model to describe the relationship between encrypted data and plaintext data, as well as the relationship between the two models. This information can be used to construct an encryption key. In addition, it will help to find out which parts of plaintext are similar to some parts of encrypted data. When an application has a large amount of sensitive information, you may need to protect this information. In this case, it is necessary to encrypt all or part of the sensitive information so that only authorized users can access them. The second point is how to manage encrypted data.

The role and significance of recommendation system in music teaching is very important, but there is a problem of low management level. The recommendation system cannot solve the problem of processing multi-note data in music teaching, and the recommendation accuracy is poor. Therefore, this paper proposes particle swarm optimization to optimize the music recommendation system. Firstly, music teaching standards are used to classify music data, and selected according to the degree of compliance to realize the preprocessing of music data. Then, according to the degree of compliance, a systematic review collection is formed, and the evaluation results are analyzed. MATLAB simulation shows that the particle swarm algorithm has a higher degree of optimization for the music recommendation system and improves the compliance rate of music selection, which is better than the single system method.

The role of speed multiplexing the design of data transmission system is very important, but there is a problem of unstable data transmission. Single data transfer does not solve the problem of multiple types of data transfer and is less accurate. Therefore, this paper proposes a single-chip microcomputer high-speed multiplexing method to construct a data transmission model. First, classify according to the transmission plan and transmission data, and select the scheme according to the transmission standard to realize the transmission of Standardized data treatment. Then, according to the classification of transmitted data, a collection of transmitted data is formed, and an iterative the transmitted data. MATLAB simulation that under the condition that the transmission plan is unchanged, the integrity of single-channel data transmission by the high-speed multiplexing method of a single-microcomputer microcomputer The transfer time is better than the standard transmission method.

The role of MOOC system in Japanese language teaching is very important, but there is a problem of low level of Japanese assistance. The MOOC system cannot solve the problem of translating grammar and words in Japanese language teaching, and it is less logical. Therefore, this paper proposes a BP neural algorithm to construct an optimization model for MOOC system. First of all, the transfer standard is used to divide the Japanese teaching content, and the Japanese teaching content is carried out according to the translation requirements Pre-treatment for Japanese language teaching. Then, the transfer criteria are divided according to the teaching standards, forming an optimized set of Japanese language teaching, and the Japanese content is carried out Dig deeper. MATLAB simulation shows that the optimization degree and stability of BP neural algorithm are better than those of online translation methods under the condition that the teaching standards are consistent.

The prediction of green credit default on deep learning algorithm is to predict the in the green credit portfolio of enterprises. We use recursive feature elimination (RFE) algorithm to train the DL model, and then apply it to predict the default risk of green credit of listed companies. The RFE algorithm iteratively deletes features from the input data until the performance on the verification set cannot be improved. In our research, we found that the combination of RFE algorithm and logical regression as outlier detector can improve the prediction performance by 2%, and the prediction probability as the input of the risk control system can be used to evaluate the effectiveness of the management’s decision-making ability.

In cloud computing, users may face the problem of resource shortage. In this regard, overcapacity is an important issue. In traditional computer systems, when CPU and memory resources are scarce, it will lead to serious performance degradation or even crash. However, in a cloud computing environment, if the client has no resource constraints (for example, CPU and memory), it can be used to improve the utilization efficiency of other resources such as storage and network bandwidth. This paper proposes a new evaluation method, which is based on the evaluation results of runtime residual capacity obtained from real data and simulation model analysis. The results obtained in each area will indicate whether these computing resources are worth utilizing. In addition, we need to consider other factors, such as cost and safety.

The project is to design and implement a monitoring system based on data mining algorithm, which will be able to monitor the performance of the operations and maintenance (O&M) activities performed by operators to detect problems that may occur during these operations. The goal is also to provide managers, engineers and technicians with effective tools that they will use as support in their daily work. The monitoring system must be able to collect relevant information about O&M activities, analyze them using appropriate algorithms and provide suggestions for improvement. In addition, it should have a graphical interface so that users can easily access all necessary information.

The role of transportation route optimization in civil aviation logistics is significant, but there is a problem with poor optimization. The route planning method cannot solve the problem of multi-route selection in civil aviation transportation, and the rationality. Therefore, this paper proposes an to construct a logistics route optimization model. Firstly, transportation knowledge is used to classify the transportation route, and the transportation route is selected according to the length of the route to realize the standardized processing of time. The transportation knowledge then classifies the time into a path optimization collection and iteratively analyzes the scored content. MATLAB simulation shows that the A+ algorithm's optimization degree and optimization time of A+ algorithm are better than that of the path planning method under a certain area range.

The role of the translation system in interactive Japanese translation is very important, but there is a problem with the low translation levels. The translation system cannot solve the problem of translating grammar and words in interactive Japanese translation, and the Chinese logic is low. Therefore, this paper proposes a feature extraction algorithm and designs a translation system optimization. First, the Japanese standard is used to divide the interactive Japanese translation content, and the interactive Japanese translation content is carried out according to the translation requirements to realize the preprocessing of interactive Japanese translation. Then, the Japanese standards are divided according to the teaching standards to form an optimized collection of interactive Japanese translations, and the Japanese content is dug deeper. MATLAB simulation shows that the optimization degree and logic of the feature extraction traditional translation under the condition that the teaching standards are consistent.

With the acceleration of global city, low-carbon urban planning has become increasingly important. Traditional urban planning methods have encountered certain problems in facing this challenge. Therefore, from the perspective of clustering optimization algorithms, this article proposes a low-carbon urban planning method based on clustering optimization algorithms. Plan and adjust clustering optimization algorithms to comprehensively identify low-carbon city data and shorten the time of clustering optimization algorithms. Structuring urban areas using clustering algorithms to reduce carbon emissions and energy consumption. In the experiment, this method was applied to low-carbon planning in a certain city and achieved good results. The results show that the low-carbon urban planning method using clustering optimization algorithm can reduce carbon emissions, save energy, improve regional quality, and provide new solutions for urban planning.

The role of object detection in embedded systems is vital, but there is a problem with false detection. Manual detection cannot solve the problem of large amounts of data in embedded systems, and the processing efficiency is low. Therefore, this paper proposes an object detection algorithm for embedded system analysis. Firstly, the detection theory is used to evaluate the system, divided according to the detection requirements to reduce the interference factors in the detection. Then, the detection theory analyzes the embedded system, forms a detection scheme, and comprehensively analyzes the detection requirements. MATLAB simulation shows that under specific detection standards, the detection accuracy and object detection of manual detection methods.

The immersive scene design of a kind of art design type, which is based on the theory of positive psychology and flow theory, and uses digital media as the main technology and tool to build a certain scene, and through stimulating the feeling and perception of the experiencer, it can achieve the state of immersion and forgetting other things. The mainly studying the application. The main purpose of this research is to develop an effective method to design and implement digital media systems using computer virtual reality technology. This importance, components, functions of digital media system design, and how to realize it through.

Cloud data center and cloud management platform are the two most important technology applications of cloud computing. Cloud data center is a virtual data center built on the Internet. Users can use this virtual data center to store their information, conduct their business transactions, etc., without building a physical data center. In contrast, the cloud management platform is a service that allows users to manage and control their own resources in the cloud through application program interfaces (APIs). Introduced due to technological progress. Cloud computing can be defined as “providing computing services on demand, regardless of the underlying physical infrastructure”. The cloud involves not only virtualized data storage, but also other services, such as application deployment and management. These services are provided by third-party providers through server networks located in different locations around the world.

The role of the clustering algorithm in the automatic scoring of the English translation system is very important, but there is a problem of poor scoring accuracy. Simple English scoring methods do not solve the correlation problem of sentence translations in automatic scoring, and there is less correlation. Therefore, this paper proposes a clustering algorithm to construct an automatic scoring model. Firstly, the semantic knowledge is used to classify the English translation content, and the English content is divided according to the scoring criteria to realize the standardized processing of sentences. Semantic knowledge then classifies sentences into English translation collections and iteratively analyzes the scored content. MATLAB simulation shows that under the condition of a certain number of words, the clustering algorithm's scoring accuracy and translation time are better than the simple English scoring method.

Based on the problems and bottlenecks encountered in data management of power grid enterprises, this paper studies the construction and application of data accountability system. It is mainly divided into two parts: the first part studies the construction of data accountability system from the aspects of background, process and methods of accountability system construction; The second part describes the pilot application of data accountability in power grid enterprises. Through the research on the data accountability system of power grid enterprises, the data accountability management has been implemented in the business system, which has promoted the marketing data asset management of power grid enterprises.

The role of informatization in secondary vocational colleges and universities is very important, but there is a problem that the degree of informatization is not high. The existing management system cannot solve the problem of managing multiple types of information in teaching, and the concentration of information is not high. Therefore, this paper proposes a data mining algorithm and constructs an information guidance model. First of all, according to the requirements of informatization, the information of secondary vocational secondary schools is collected, and the information is summarized according to the degree of importance to realize the preprocessing of information. Then, an information collection is formed, and the data is self-learning and analyzed. MATLAB simulation shows that under the condition of specific processing standards, the guidance degree and informatization degree of the data mining algorithm are better than those of cluster analysis method.

The embedded network equipment fault remote monitoring is an embedded system that uses the Internet to connect and communicate with other equipment. It can be used for industrial automation, medical treatment, home security, etc. The monitoring platform mainly includes three parts: data receiving and transmitting module, database and client. The data receiving and sending module is not only responsible for receiving data and storing it in the database, but also responsible for forwarding the data to the client. The database is responsible for storing and maintaining data, and providing links to clients for querying historical data. The client obtains the real-time data of remote sensors and controllers through the data receiving and transmitting module connected to the server, performs status display and remote monitoring, and accesses the database for historical data query and analysis. Aiming at the two key technologies of signal acquisition and information transmission in the monitoring system, this paper studies and designs a monitoring system applied to the remote status of equipment failures. This system mainly focuses on the function realization of the lower computer, which can complete the accurate acquisition of equipment status signals and reliable transmission through LoRa wireless technology.

The role of dynamic monitoring in the behavior of students in college physical education is significant, but there is a problem that the monitoring accuracy is not high. Previous statistical monitoring methods could not solve the problem of accurate monitoring in behavioral monitoring, and there were few monitoring indicators. Therefore, a dynamic monitoring method is proposed to construct a behavior monitoring model. Firstly, the big data mining theory is used to plan behavior monitoring data, and the data collection and division are carried out according to student behavior to reduce the subjective factors in monitoring. Then, the big data mining theory plans the behavior monitoring, forms a data collection of monitoring results, and continuously monitors the data. MATLAB simulation shows that the dynamic monitoring method’s evaluation accuracy and monitoring time are better than the previous statistical monitoring methods under the condition of certain monitoring data.

The role of the ARM program in automatic shelling data collection is very important, but there is a problem with the low accuracy of collecting shelling data. The ant colony algorithm cannot solve the problem of collecting multiple types of automatic hulling data, and the recognition rate is low. Therefore, this paper proposes an ARM program to build an automatic shelling data optimization model. Firstly, the dehulling data collected is classified by the dehulling standard, and the automatic dehulling scheme is selected according to the particle diameter Implement preprocessing of shelled data. Then, according to the degree of shelling, a collection of shelling data is formed, and the parameters are iteratively optimized. MATLAB simulation shows that the ARM program can improve the shelling depth and shorten the shelling time when the shelling diameter is consistent, and the relevant results are better than the ant colony algorithm.

The role of system information in vocational colleges is significant, but there is a problem of a low level of information processing. The management system cannot solve the problem of managing multiple types of information in the vocational college system, and the rationality is low. Therefore, this paper proposes a decision tree ID3 algorithm to construct a system information optimization model. First of all, the management standards of colleges and universities are used to classify the information of vocational colleges, and the information of vocational colleges is selected according to the degree of importance to realize the preprocessing of information. Then, the university management standards are classified according to importance, form an information optimization collection, and conduct a self-learning analysis of the scoring content. MATLAB simulation shows that the optimization degree and optimization stability of the decision tree ID3 algorithm are better than those of the ant colony algorithm when the system is fixed.

The development of digital media has provided people with more diverse and diverse ways of entertainment and learning. In this context, dynamic has gradually become an indispensable part of digital media. In digital media, can express the characteristics and advantages of products through animation, videos, and other means, enhance users’ impression and experience of the product, and improve user satisfaction. For example, on an e-commerce website, using dynamic visual communication design can present the characteristics of a product to users through visual and voice means, thereby increasing their purchasing interest. This paper mainly discusses the application of dynamic, and forecasts the future development direction of dynamic visual. Visualization and animation are the two most important technologies used in this field. In visualization, we try to create an image or video so that people can see it without any problems. Animation is basically used to move images from one position to another. For example, if you want your company logo or product name to appear on the screen for a few seconds, it will be animated so that the audience can clearly see what you are trying to convey through it.

The role of computer-assisted translation in Korean translation practice is very important, but there is a problem of poor translation accuracy. The online Korean translation method cannot solve the problem of the association of sentence translation in practice, and the logic is poor. Therefore, this paper proposes computer-aided construction of Korean translation relationships. First, the difficulty of translation is classified using grammar knowledge, and the Korean content is divided according to the translation standard to simplify the Korean content Processing. Then, grammatical knowledge classifies sentences, forms a Korean translation mapping table, and performs revision analysis of the translated content. MATLAB simulation shows that the accuracy and time of computer-aided translation are better than online Korean translation methods under the condition of certain translation difficulties.

The application of virtual reality service in call center makes the voice evaluation quality of call center a research hotspot. Previous voice quality assessment methods could not solve the problem of stacked encoding in call centers, and the voice quality assessment capability was low. Therefore, this paper proposes a virtual reality method to construct a voice quality evaluation system for call centers. First, stacked autoencoders are used to classify the data of the call center and divide the collection according to the department, which reduces the complexity of data processing. The autoencoder then categorizes the call center into a collection of voice quality, and the data in the collection is automatically evaluated. In MATLAB simulation, the evaluation accuracy and calculation time of the virtual reality method are better than the previous speech evaluation methods under the condition that the amount of call data is fixed.

The role of the Octopus optimization algorithm in the empirical demonstration of the reading needs of library readers is significant, but there is a problem with the low accuracy of demand analysis. Hierarchical needs analysis cannot solve the problem of the needs of readers in multiple types of libraries, and the demand judgment results are poor. Therefore, this paper proposes an Octopus optimization algorithm to demand a library reader judgment model. Firstly, the reading needs are used to classify the markets, and the demand methods are selected according to the requirements of library readers to realize the preprocessing of needs analysis. Then, according to the degree of Demand, the requirements analysis set is formed, and the parameters are iteratively judged. MATLAB simulations show that the Octopus optimization algorithm among library readers can increase the scale and shorten the Demand The requirements analysis time and the relevant results are better than the hierarchical demand analysis method.

The role of data mining algorithms in developing the tourism economy is significant, but there is the problem of inaccurate economic forecasting. The historical economic analysis method cannot solve the problem of randomness in the tourism economy, and there are few indicators for tourism economic analysis. Therefore, this paper proposes a data mining algorithm method to construct a tourism economic development model. First of all, under the normalization of the epidemic, the crawler method to classify the data and carry out according to the stage of economic development. Collective division of economic data to reduce the influence of random factors. Then, the crawler method divides the tourism economic development score into grades, forms an economic data collection, and continuously analyzes the economic data. MATLAB simulation shows that under the normalization of the epidemic, the accuracy and analysis time of the data mining algorithm method are better than the historical economic analysis method.

Bayesian networks play an important role in financial risk data analysis, but there is a problem of low accuracy of risk prediction. Financial statistical analysis cannot solve the problem of accurate early warning in financial risks, and there are few early warnings. Therefore, this paper proposes a Bayesian network to construct. Firstly, the big data mining theory is used to grade the massive information, and the massive information is carried out according to the risk standards Set division to reduce ambiguity in early warning. Then, the big data mining theory grads the financial risk early warning forms a collection of early warning results and continuously warns the massive information. MATLAB simulation shows that under a certain amount of financial data, the Bayesian network’s early warning accuracy and warning time are superior to financial statistical analysis.

The role of IoT technology in lane detection is significant, but there is a problem with low detection accuracy. Manual route planning cannot solve the problem of lane detection in route selection, and there are fewer lane detection indicators. Therefore, this paper proposes an Internet of Things technology to build a lane intelligent detection system. Firstly, the operation research theory is used to draw the route of the lane data, and the lane collection is divided according to the traffic flow standard to reduce the uncertainties in lane detection. Then, according to the operation research theory, the planning path is drawn to form a lane planning set. MATLAB simulation shows that under the condition of a certain traffic flow, the detection accuracy and detection time of IoT technology are better than that of manual route planning method.

Because the financial risk data of domestic listed companies are unbalanced, and the traditional supervised model is difficult to establish a risk discrimination model through large sample learning. Due to the lack of data, the training of supervised models is usually based on manually filtered balanced data sets, which will affect the applicability of models in the real world. This research involves financial risk identification based on fish school algorithm. The this study is to compare the results of two algorithms (i.e. Fish School and Fuzzy Logic) to distinguish companies with different levels of financial risk. The comprehensively analyze the performance of two different methods in screening in PSX.

The role of teaching management in teaching management is very important, but there is a problem of low management level. The teaching management system cannot solve the problem of multi-type lesson plan management in the teaching management system, and the matching is low. Therefore, this decision tree a teaching management. First, the teaching plan is classified according to the teaching plan and content, and the teaching plan is selected according to the teaching plan to realize the standardized processing of teaching information. Then, according to the classification of teaching content, a collection of teaching information is formed, and the teaching content is iteratively analyzed. MATLAB simulation shows that the decision tree the standard management method in optimizing the teaching management system in terms of optimization degree and optimization time under the condition that the teaching plan remains unchanged.

The evaluation of quality education credits is to use the decision tree to judge whether students should obtain educational credits. Decision trees are used because they can easily be built from existing data sets collected for other purposes. The analysis here will be completed with the help of data from previous years, which will be used as input to this prediction model. The main purpose behind the use of this technology is to find out whether correlation between some factors such as age, gender, class and high-quality education credits. The purpose of such analysis is to determine students who may be eligible for educational credits based on their academic achievements, but their academic achievements do not meet the necessary criteria for obtaining educational credits. This type of analysis will help schools decide which students should receive credit and which students do not need credit.

Multi sensor fusion state estimation of autonomous navigation vehicle based on RNN model. Multi sensor fusion is a process of combining information from different sensors to estimate a single value. In this paper, we propose a recursive neural network (RNN) model to estimate the state of multiple sensors. This method uses RNN architecture with two hidden layers and one output layer to combine sensor data. The input data is first divided into a sequence and then input into the RNN model to predict the output. We use three types of sensors: IMU, LIDAR and GPS. First, the RNN model is trained using three sensors for the navigation system. Then, we use this trained model to estimate the current position using only one sensor data. Finally, we combine the data of all three sensors to obtain the final estimated position of the vehicle at any given time step.

A novel image clarity processing algorithm based on improved particle filter is proposed to address the issues of blurring and fogging in inscriptions and calligraphy images. This algorithm introduces feature parameters such as peak signal-to-noise ratio and ambiguity, and uses an improved particle filter method to process and restore the clarity of inscription calligraphy images. The specific implementation process is as follows: Firstly, the inscription calligraphy image is divided into a series of small blocks, and the peak signal-to-noise ratio and ambiguity feature parameters of each block are calculated. Then, an improved particle filter algorithm is used to process and restore the small blocks with the best peak signal-to-noise ratio and ambiguity. Finally, concatenate all the processed small pieces into a complete inscription calligraphy image. The improved particle filter algorithm adopts adaptive observation noise covariance matrix and resampling strategy, which improves the accuracy and robustness of image processing. In the experiment, we used multiple actual inscriptions and calligraphy images for testing and comparison, and the results showed that this algorithm not only effectively improves the clarity and restoration effect of inscriptions and calligraphy images, but also has good stability and practicality.

The a method to monitor the voltage. This method can be applied to high-voltage circuit breakers, with intelligent control function, and can be connected to the Internet. In case of over-current or short circuit and other abnormal conditions, the method will automatically detect these abnormal conditions, so that the control system can take appropriate measures immediately. Therefore, compared with traditional methods, this method has the advantages of fast response time and low cost. Vacuum monitoring method is a kind of operation monitoring method Things (IoT). The pressure in the vacuum chamber and sends the data to the remote server. This information can be used to detect abnormal operations, such as when there is no air in the vacuum chamber or there are too many particles blocking the gas flow. The vacuum monitoring method uses sensors to measure the pressure and temperature in the vacuum chamber.

The research will focus on the application of computer digital media and virtual reality technology to support, enhance, improve and promote the learning process in various fields. The project is expected to provide a new way for students to learn through the use of computer digital media and virtual reality technology. In addition, it will contribute to the development of Malaysia's education system through its potential as an effective teaching tool.

The role of encryption algorithms in wireless sensor network data is very important, but there is a problem with insecure data transmission. The ordinary transmission method cannot problem of multi-dimensional, and the security is low. Therefore, this paper proposes a a network data transmission model. First, according to the wireless transmission plan, wireless transmission data classification, according to the wireless transmission standard for transmission data selection. Standardize the processing of data transmitted wirelessly. Then, according to the classification of wireless transmission data, a collection of wireless transmission data is formed, and the wireless transmission data is iteratively analyzed. MATLAB simulation shows that under the condition that the wireless transmission plan is unchanged, the cloud storage method is effective for the security of transmitted data. The warning time is better than the standard wireless transmission method.

As an efficient still image data compression algorithm, JPEG compression algorithm is widely used in many embedded multimedia products. The speed of image data encoding and decoding has become the most important aspect that restricts the improvement of product performance. In a single CPU embedded environment, JPEG encoding and decoding speed has almost reached the limit, and it is difficult to improve the space. However, the emergence of the multi-core embedded system represented by Fujitsu FR1000 in recent two years has made it possible to significantly improve the performance of multimedia products. In this paper, we JPEG decoding. The proposed and idea of selective optimization. It uses partial redundancy elimination (PREL) to avoid some redundant blocks in the JPEG decoder, and selects appropriate blocks for further processing according to the current image data. The whole process can be summarized as follows: 1) Initialize PREL; 2) Read all pixel values from the input buffer; 3) Selecting one or more blocks according to their contents for further processing; 4) Perform PREL on the selected block; 5) Process the remaining blocks to get the final result.

With the continuous development of agricultural farming technology, the way of human carrying pesticide spraying gradually disappeared, and unmanned spraying operation gradually appeared. Color is a very important aspect of our life. From the clothes we wear to the way we decorate our house, we can see it in almost everything we do. We use color to make us feel better. It will affect other people's view of us. In this article, I will discuss the importance of color in agricultural UAVs and how to use it to improve their appearance. Therefore, this paper introduces the application and characteristics of the modeling and power technology in the development of agricultural equipment, analyzes the connection and deficiency of its appearance modeling in operational performance through the comparison of various types of UAVs, expounds the agricultural development trend of various types of agricultural, and puts forward the prospect of agricultural plant protection UAV in agricultural development in China in the future. Based on Rhino5.0 modeling software, the appearance design and overall structure 3D modeling are completed, and the user operation interface matching the overall control system style is designed. Then, the appearance design of plant protection UAV is verified and analyzed through ergonomics, perceptual engineering and bionics; Finally, BP neural network is used to evaluate the design scheme.

With the rapid development of 5G, Internet other technologies. In the power industry, the State Grid and China Southern Power Grid put forward the development requirements of “industrial digital new infrastructure”. In this context, it has had a huge impact on the major enterprises in the power industry, which requires that all major enterprises in the power industry should actively carry out business transformation. The power engineering cost calculation data platform is a new digital infrastructure that can be used to calculate engineering. This platform promote the process of calculating and monitoring projects in an effective way. It provides a user friendly interface, so that anyone can easily use it without any technical knowledge. It also human error and improve the accuracy of calculation. The power engineering cost calculation data platform is an online tool that helps calculate various costs related to power generation, transmission and distribution projects by using advanced mathematical models based on real-time data.

The architecture of distribution network operation analysis system based on mobile information extraction method is a computer program used to analyze distribution network operation. It is designed to analyze the operation of the distribution network by using mobile information (such as image and video data) captured from different locations in the network. The main function of the software is to extract the relevant information of each node or device in the network, and then use it for further analysis. The software allows users to identify all nodes within a specific range (expressed in distance) from any location of the capture node. As a typical product, active has realized the scale access and demand response of renewable energy. In order to maximize the operation control and energy management of active need the support of intelligent electronic equipment and information and communication technology. Therefore, when analyzing, information space as well as the component fault in power physical space.

The three main goals of power production are stability, high efficiency and economic operation. The damage rate of lines can reflect the comprehensive level of economic benefits and management level of the power industry. Seen from the waste of power consumption in China, there is still a large space for loss reduction. It is particularly important good job of actual line loss estimation and in-depth study. In order to investigate the power grid line loss in real time, I used the Is designed as a real-time data acquisition and analysis tool for power system. It is an open source platform that can collect, store, process and display large amounts of data from analog or digital devices and sensors connected to the grid or substation network (PI website). It can be used by plant operators to detect problems on their systems, such as low voltage conditions that may require immediate action by utilities. It also provides advanced analysis modules for event detection and prediction.

Pesticide residues are unwanted chemical substances left after pesticides are applied to plants. Insecticides can be sprayed on crops or used as liquid sprays, but some of these chemicals may eventually enter the ground and water systems, which may cause health problems for many animals, such as humans and other animals. Based on the modification and improvement of the classical photoelectric method, a mobile photoelectric. The method is based on a set of parameters that determine the amount of light absorbed by any sample. When the spectrophotometer is used to measure the amount of light emitted by any sample according to certain wavebands, the absorption spectrum can be found. Pesticide residues are unwanted chemical substances left after pesticides are applied to plants. Insecticides can be sprayed on crops or used as liquid sprays, but some of these chemicals may eventually enter the ground and water systems, which may cause health problems for many animals, such as humans and other animals. Based on the modification and improvement of the classical photoelectric method, a mobile photoelectric method of pesticide. The method is based on a set of parameters that determine the amount of light absorbed by any sample. When the spectrophotometer is used to measure the amount of light emitted by any sample according to certain wavebands, the absorption spectrum can be found.

The information technology makes education informatization possible. The teaching management database stores. How to find the hidden in the data, and provide decision-making support for school teaching management has gradually become a new direction for educators to explore. The association analysis in university teaching management is Professor and Head, Madras Dr. Nandakumar's new book. The book was published by Springer India in 2017 and is entitled “Application of Decision Trees and Correlation Analysis”. This book is based on S The original research work carried out by Dr. Nandakumar during Dr. IIT Madras’ research covers topics such as decision tree learning, association rule mining, clustering technology, etc., which is used to use data from different fields for teaching management.

Metauniverse is a term used to describe the universe of all things. This is a universe that has no beginning or end, but it has infinite dimensions. The word metauniverse was first coined by Albert Einstein in his general theory of relativity. The metauniverse can be described as a space-time continuum containing all things at the same time, and it is also considered as the source of all things. This concept was further developed by scientists such as Stephen Hawking and David Finkelstein. They believe that everything that exists can be explained by quantum physics. Metacosmic algorithm is a data optimization method. It is an algorithm that can be used to problem. The main goal of the algorithm is to considering all possible solutions, and then selecting the best solution according to various criteria such as time and memory. This article will discuss how to optimize search engine results in terms of time and memory consumption by using meta - universe algorithm.

Computer has become an indispensable part of our life. They are used in almost all fields of human activities, from banking to medicine. Computers can perform various tasks, which were once the only domain of human beings. Computers can also be used as tools for education and learning. In fact, computers have long been used as calculators and word processors for students to write articles and prepare reports on paper, and computers have been used for education. However, with the emergence of personal computer (PC), education has entered a new era, and students can learn at their own pace. This paper briefly describes the concept and design idea of STEM, and takes a specific design project as an example to illustrate the practice and application of STEM education in the training process of computer thinking from the aspects of data extraction, abstract model, optimization algorithm, simulation programming, etc.

With the growth of social demand and the continuous breakthroughs in the field of unmanned intelligence, mobile robots have gradually entered the public's vision. They have application backgrounds in the fields of daily household, industrial production, unmanned exploration and even national defense. Autonomous navigation, as its core technology, has always been the focus of research in this field. Considering the advantages of visual sensors such as low cost and strong perception ability, this paper studies visual SLAM and path planning with visual navigation as the goal. Machine learning is a branch of artificial intelligence, which deals with the use of computers for learning without explicit programming. This is a process of computer analyzing data and making predictions based on previous experience. In this paper, we will discuss the path planning and machine learning in the vision navigation system of indoor mobile robot. Path planning is an important part of any autonomous system, because it determines how the robot should move to reach its destination or target. In this article, we will discuss path planning using different techniques such as heuristic search.

The research and design of cloud storage server based on virtualization technology is a new method in IT field. It helps to provide better performance, scalability, reliability and security for cloud storage systems. This approach reduces costs by using virtualization technology. The reliable cloud storage server, which can be used in different scenarios such as private cloud, public cloud or hybrid cloud. To achieve these goals, it uses hardware virtualization technology, which provides more flexibility than software based approaches such as Xen or KVM technology.

In view of the practical application of Apriori algorithm in addictive behavior and mental health, the students of grade 2020 in a certain university are taken as the research object. We first outline Apriori algorithm, points out the defects of traditional Apriori algorithm, and second analyze four improved methods of Apriori algorithm. Finally, the practical application of the improved Apriori algorithm in the study of addictive behavior and mental health is explored.

This method has three important characteristics: (1) using large-scale and complex phenomena to obtain the main objective function; (2) Its design is based on the combination of fuzzy reasoning and evolutionary algorithm to optimize the value of the interdependent objective function; (3) The optimization mechanism focuses on “fuzzy trust” rather than simple “trust”. In addition, a more detailed study was carried out by comparing with other results in related fields. Financial risk control is one of the most important problems that modern organizations must solve. Therefore, it is necessary to find effective risk control approaches and tools. Data can play an important role in this process because they are many different sources of information about business activities, and they can also help us make informed decisions when we use the appropriate methods.

With the development of cross-border e-commerce, cross-border e-commerce for agricultural products has become an important market and application field. To ensure the sustainable development of cross-border e-commerce for agricultural products, it is necessary to cultivate qualified talents. This article is based on BP neural network and designs a talent cultivation plan for cross-border e-commerce of agricultural products, which has been verified in practical applications. The results indicate that the talents trained in this program can adapt to the current demand of the cross-border e-commerce market for agricultural products and have achieved good results.

Deep neural networks are increasingly important tools for machine learning and artificial intelligence research, because they can learn very complex tasks without explicit programming. They are also flexible because they can be easily adapted to solve new problems. Character recognition involves the post perception process and is affected by many factors, such as the amount of visual information available and cognitive impairment. This method assumes that there are two types of representations in character memory: abstract forms, which are determined by font attributes, may vary with each letter, and specific forms or “embodiment” representations. The detection part adopts the horizontal text line detection method to detect the text, and in the text part, the system is analyzed and introduced in detail from the production of the model to the design and implementation of the neural network.

With the popularization and development of the Internet, e-commerce has become an indispensable part of modern business models. Faced with the increasing demand from consumers, e-commerce platforms require a large amount of agricultural product information to meet their needs. At the same time, e-commerce platforms based on big data algorithms can also provide more refined services, optimize marketing strategies, improve shopping experience, and so on. This article focuses on the review and analysis of e-commerce agricultural products based on big data algorithms. It introduces the process, model design, experimental results, and analysis of agricultural product data construction based on big data algorithms on current e-commerce platforms, and ultimately concludes that e-commerce agricultural products based on big data algorithms can better meet consumer needs, optimize agricultural product procurement, increase sales volume, and improve the quality of agricultural product supply chain.The review and analysis is a method that can be used to analyze the comments and comments left by customers in the product review part of online stores. This technology is not new, but it has improved over time. It was first developed for Amazon in 2009. Today, many companies use this technology as part of their marketing strategies to increase sales and improve customer satisfaction. The determine how consumers like certain products, so as to more effectively advertise on websites or social media platforms (such as Twitter and Facebook).

It is of great practical value to establish an accurate financial forecasting model for financial product management and risk control. In view of the characteristics of short launch cycle of financial products and less modeling data in the new era, a grey linear regression combination financial forecasting model with less data modeling is constructed. Grey linear regression is a technique for financial forecasting, which is used to estimate the future value of variables at different time points. It is also known as the method of “combining” two models or datasets to predict the future value of a variable from the past and current values of other variables. if we know the relationship between two variables, we can use it to predict the value of another variable according to its relationship with the two variables. Simply put, grey linear regression combines two models (or datasets) and predicts one variable. Finally, the paper empirically analyzes the effectiveness of the grey linear combination financial forecasting model for the few data modeling, and the empirical results show that the combination financial forecasting model has a high prediction accuracy.

This article studies the application of k-means clustering algorithm in automatic machine learning. By selecting the Iris dataset from the UCI dataset for experiments, we found that the k-means algorithm can accurately classify data. At the same time, we compared the traditional k-means clustering algorithm with the optimized K-Means++ and Spherical K-Means clustering algorithms in terms of clustering accuracy and speed, and found that the optimized algorithm has improved clustering accuracy and speed compared to traditional algorithms. Therefore, applying these two optimized k-means clustering algorithms in AutoML can improve the clustering and generalization abilities of machine learning models.

With the popularity of mobile networks, e-commerce exchanges, mobile terminals and social networks have largely expanded the scope of network applications. Nowadays, network data almost covers all areas of society, and it has brought great changes to national politics, culture and economy, as well as the lives of citizens. Data mining is a technology that can be used to extract information from large databases. Data mining uses a variety of techniques, such as association rules, classification, clustering, and sequential pattern analysis. Data mining has become an important technology in the field of business intelligence, which helps enterprises make better decisions based on data analysis. This paper focuses mining algorithms such as association rules, classification and sequential pattern analysis. We will also discuss how to use the WEKA toolkit software library to implement these algorithms in JAVA language.