Chinese named entity recognition (CNER) is one of the important tasks in natural language processing. Unlike the English, Chinese lacks explicit word boundaries. Therefore, many models were designed to address this issue by incorporating word lexicon information into the CNER. However, lots of irrelevant information may be included when matching the entire lexicon for each character. Inspired by the SoftLexicon method, we propose a multi-head attention based model to simplify the introduced lexicon information to generate word-level attention vector. In this method, a word vector matched for each character is first obtained and further weighted by the relevance with the character-level vector to calculate the word-level attention vector. In this way, only the words existing in the sentence are matched, which reduces the scope of word matching. The effectiveness of this method is verified on multiple Chinese datasets.
{"title":"Chinese Named Entity Recognition Method Based On Multi-head Attention Enhancing Word Information","authors":"Ting Wang, Songze He","doi":"10.1145/3548608.3559300","DOIUrl":"https://doi.org/10.1145/3548608.3559300","url":null,"abstract":"Chinese named entity recognition (CNER) is one of the important tasks in natural language processing. Unlike the English, Chinese lacks explicit word boundaries. Therefore, many models were designed to address this issue by incorporating word lexicon information into the CNER. However, lots of irrelevant information may be included when matching the entire lexicon for each character. Inspired by the SoftLexicon method, we propose a multi-head attention based model to simplify the introduced lexicon information to generate word-level attention vector. In this method, a word vector matched for each character is first obtained and further weighted by the relevance with the character-level vector to calculate the word-level attention vector. In this way, only the words existing in the sentence are matched, which reduces the scope of word matching. The effectiveness of this method is verified on multiple Chinese datasets.","PeriodicalId":201434,"journal":{"name":"Proceedings of the 2022 2nd International Conference on Control and Intelligent Robotics","volume":"57 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2022-06-24","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"127256409","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Zhiguo Wu, Lizhong Shui, Shuguang Liu, Zijian Wang
Czochralski method is the main method to produce silicon single crystal, therefore, its growth process and control method have always been the focus of the silicon industry. Single crystal growth process with an invariable diameter can be are influenced by temperature and pulling speed and rotational speed, crucible's tracking speed and rotational speed, growth physical environment, protective gas flowing speed and temperature, cooling water flowing speed and temperature, etc. while the parameters’ selection such as thermal field distribution in furnace, growth temperature compensation, growth speed is the basic condition of single crystal shape, therefore, a cascade control strategy by combining thermal field with growth speed is adopted in control system. Moreover, due to the lack of accurate mathematical model and the characteristics of non-linearity and variability, an active disturbance rejection control (ADRC) strategy is proposed in this paper. The experimental results show that the above control strategy can effectively control the growth of single crystal with an equal diameter, high control accuracy, strong anti-interference ability, insensitive to parameter changes, and has a certain degree of robustness.
{"title":"An ADRC-based CZ Silicon Single Crystal Growth Control Technology","authors":"Zhiguo Wu, Lizhong Shui, Shuguang Liu, Zijian Wang","doi":"10.1145/3548608.3559253","DOIUrl":"https://doi.org/10.1145/3548608.3559253","url":null,"abstract":"Czochralski method is the main method to produce silicon single crystal, therefore, its growth process and control method have always been the focus of the silicon industry. Single crystal growth process with an invariable diameter can be are influenced by temperature and pulling speed and rotational speed, crucible's tracking speed and rotational speed, growth physical environment, protective gas flowing speed and temperature, cooling water flowing speed and temperature, etc. while the parameters’ selection such as thermal field distribution in furnace, growth temperature compensation, growth speed is the basic condition of single crystal shape, therefore, a cascade control strategy by combining thermal field with growth speed is adopted in control system. Moreover, due to the lack of accurate mathematical model and the characteristics of non-linearity and variability, an active disturbance rejection control (ADRC) strategy is proposed in this paper. The experimental results show that the above control strategy can effectively control the growth of single crystal with an equal diameter, high control accuracy, strong anti-interference ability, insensitive to parameter changes, and has a certain degree of robustness.","PeriodicalId":201434,"journal":{"name":"Proceedings of the 2022 2nd International Conference on Control and Intelligent Robotics","volume":"21 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2022-06-24","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"121539102","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
This paper proposes a linear quadratic controller based on particle swarm algorithm for the rear wheel control of four-wheel steering vehicle. Particle swarm optimization with fitness functions is used to optimize the coefficients of the weight matrix offline. The fuzzy rules following the controller is used if the road condition is terrible. The simulation results show that the LQR control model based on particle swarm optimization makes the trajectory tracking of the vehicle better and the side slip angle of the vehicle lower. It can be proved that the controller has positive effect on handling stability of the vehicle and safety of drivers.
{"title":"LQR Optimal Control of Four-steering Vehicle Based on Particle Swarm Optimization Algorithm","authors":"S. Zhu, Xueyuan Li, Xinyi Qu, Qi Liu, Zirui Li","doi":"10.1145/3548608.3559298","DOIUrl":"https://doi.org/10.1145/3548608.3559298","url":null,"abstract":"This paper proposes a linear quadratic controller based on particle swarm algorithm for the rear wheel control of four-wheel steering vehicle. Particle swarm optimization with fitness functions is used to optimize the coefficients of the weight matrix offline. The fuzzy rules following the controller is used if the road condition is terrible. The simulation results show that the LQR control model based on particle swarm optimization makes the trajectory tracking of the vehicle better and the side slip angle of the vehicle lower. It can be proved that the controller has positive effect on handling stability of the vehicle and safety of drivers.","PeriodicalId":201434,"journal":{"name":"Proceedings of the 2022 2nd International Conference on Control and Intelligent Robotics","volume":"183 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2022-06-24","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"131499497","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
With the rapid development of rail transit network, scientific and effective urban public transportation management is of great significance to maintaining public order and planning transportation operations. This paper aims at the original network structure of YOLOv3, deletes the repeated data of the convolution module in the detection network, and optimizes the convolution method to design a lightweight network structure. Aiming at the problem of the complex posture and background of the crowd in the subway, the Darknet53 network with better performance is selected as the feature extraction network. At the same time, according to the actual speed requirements for subway pedestrian detection, the repeated data and parameters of the repeated convolution module in the deep network are deleted, and the original convolution method is replaced with a smaller depth convolution method. Thereby it reduces time complexity of network and improves the detection speed, and achieves the application of lightweight neural network model in subway crowd recognition.
{"title":"Application of Dense Crowd Detection Method Based on Lightweight Neural Network in Subway Crowd Recognition","authors":"Jimin Liu, Junjie Zhou, Fang Fan, Chuangsen Xie","doi":"10.1145/3548608.3559279","DOIUrl":"https://doi.org/10.1145/3548608.3559279","url":null,"abstract":"With the rapid development of rail transit network, scientific and effective urban public transportation management is of great significance to maintaining public order and planning transportation operations. This paper aims at the original network structure of YOLOv3, deletes the repeated data of the convolution module in the detection network, and optimizes the convolution method to design a lightweight network structure. Aiming at the problem of the complex posture and background of the crowd in the subway, the Darknet53 network with better performance is selected as the feature extraction network. At the same time, according to the actual speed requirements for subway pedestrian detection, the repeated data and parameters of the repeated convolution module in the deep network are deleted, and the original convolution method is replaced with a smaller depth convolution method. Thereby it reduces time complexity of network and improves the detection speed, and achieves the application of lightweight neural network model in subway crowd recognition.","PeriodicalId":201434,"journal":{"name":"Proceedings of the 2022 2nd International Conference on Control and Intelligent Robotics","volume":"1 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2022-06-24","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"129486749","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
This paper develops a professional underwater inspection equipment that aims to shorten the time required for pipeline inspection, solve the problems of low intelligence and poor inspection efficiency of the current traditional underwater inspection equipment, and can quickly and accurately complete the underwater inspection. A local area network is established between robots through LoRa wireless communication technology, and data is uploaded to the management and scheduling platform. Data processing and information display are performed by the platform to improve the efficiency of underwater inspection, and the robot is used to replace manual labor and large-scale equipment to reduce inspection costs. The robot uses deep learning to process camera data to further achieve autonomous search and high-precision identification.
{"title":"Low-cost Underwater Inspection Robot Based on Deep Learning","authors":"Zichen Guo, Changming Zhang","doi":"10.1145/3548608.3559187","DOIUrl":"https://doi.org/10.1145/3548608.3559187","url":null,"abstract":"This paper develops a professional underwater inspection equipment that aims to shorten the time required for pipeline inspection, solve the problems of low intelligence and poor inspection efficiency of the current traditional underwater inspection equipment, and can quickly and accurately complete the underwater inspection. A local area network is established between robots through LoRa wireless communication technology, and data is uploaded to the management and scheduling platform. Data processing and information display are performed by the platform to improve the efficiency of underwater inspection, and the robot is used to replace manual labor and large-scale equipment to reduce inspection costs. The robot uses deep learning to process camera data to further achieve autonomous search and high-precision identification.","PeriodicalId":201434,"journal":{"name":"Proceedings of the 2022 2nd International Conference on Control and Intelligent Robotics","volume":"163 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2022-06-24","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"134498474","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
In order to improve the opening response time of the solenoid valve of a certain marine diesel engine, this paper analyzes the influence on the opening response time of the solenoid valve based on ANSOFT software. The results show that the opening response time of the solenoid valve can be appropriately shortened by increasing the voltage and reducing the working gap, but it still can not meet the 10ms requirement of opening time. However, the opening response time of the solenoid valve can be significantly shortened by optimizing the coil parameters under certain conditions of the solenoid valve structure parameters. Finally, a quarter coil is used to shorten the opening response time of the solenoid valve from 27.4ms to 9.2ms shortened by 66.4%, which meets the requirement of the opening response time of the solenoid of a marine diesel engine.
{"title":"Analysis of Influence factors of Opening Response Time of High-pressure solenoid valve","authors":"Guofei Du, Yuan Li, Jian Zhao, B. Zhang","doi":"10.1145/3548608.3559258","DOIUrl":"https://doi.org/10.1145/3548608.3559258","url":null,"abstract":"In order to improve the opening response time of the solenoid valve of a certain marine diesel engine, this paper analyzes the influence on the opening response time of the solenoid valve based on ANSOFT software. The results show that the opening response time of the solenoid valve can be appropriately shortened by increasing the voltage and reducing the working gap, but it still can not meet the 10ms requirement of opening time. However, the opening response time of the solenoid valve can be significantly shortened by optimizing the coil parameters under certain conditions of the solenoid valve structure parameters. Finally, a quarter coil is used to shorten the opening response time of the solenoid valve from 27.4ms to 9.2ms shortened by 66.4%, which meets the requirement of the opening response time of the solenoid of a marine diesel engine.","PeriodicalId":201434,"journal":{"name":"Proceedings of the 2022 2nd International Conference on Control and Intelligent Robotics","volume":"20 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2022-06-24","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"134364758","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
This article is an applied research that combines the technology and methods of big data with student work. This study used the canteen consumption data of freshman to junior students in a university in Guangdong from September 2018 to December 2020. And found the suspected lonely in the student group by constructing a co-occurrence network. The results showed that: (1) the frequency of co-occurrence among freshmen to junior students shows a decreasing trend, and the frequency of co-occurrence of freshmen friends was significantly higher than that of sophomores and juniors; (2) The number of college students who are suspected of being "lonely" increases with grades; (3) The average consumption level of students who are suspected of being "lonely" is significantly lower than that of students who are "non-lonely"; (4) Students who are suspected of being "lonely" are more inclined to use meal cards for consumption rather than WeChat/Alipay; (5) The degree of intimacy among freshman students is significantly higher than that of sophomores and juniors.
{"title":"An Applied Research of Big Data in the Work of Higher Education Students: Taking the Analysis of Finding Loneliness and Their Characteristics as an Example","authors":"Weiqing Li, Jingyi Zhang, Yaping Qiu, Chengxu Cao","doi":"10.1145/3548608.3559308","DOIUrl":"https://doi.org/10.1145/3548608.3559308","url":null,"abstract":"This article is an applied research that combines the technology and methods of big data with student work. This study used the canteen consumption data of freshman to junior students in a university in Guangdong from September 2018 to December 2020. And found the suspected lonely in the student group by constructing a co-occurrence network. The results showed that: (1) the frequency of co-occurrence among freshmen to junior students shows a decreasing trend, and the frequency of co-occurrence of freshmen friends was significantly higher than that of sophomores and juniors; (2) The number of college students who are suspected of being \"lonely\" increases with grades; (3) The average consumption level of students who are suspected of being \"lonely\" is significantly lower than that of students who are \"non-lonely\"; (4) Students who are suspected of being \"lonely\" are more inclined to use meal cards for consumption rather than WeChat/Alipay; (5) The degree of intimacy among freshman students is significantly higher than that of sophomores and juniors.","PeriodicalId":201434,"journal":{"name":"Proceedings of the 2022 2nd International Conference on Control and Intelligent Robotics","volume":"28 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2022-06-24","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"133326519","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Unmanned Ground Vehicle (UGV) is currently the key research direction in the world. As an interdisciplinary complex, it covers machinery, computers, automation, ergonomics and other aspects. The application of UGV will liberate people from driving, reducing traffic accidents, and saving people's time and cost. As an important part of the UGV software system, trajectory prediction is very important to improve the security of UGV. This article classifies and summarizes various current prediction algorithms, puts forward the current problems of various algorithms, and looks forward to the future development direction.
{"title":"Summary of Trajectory Prediction Method for Unmanned Ground Vehicle","authors":"Xuejie Du, Aoran Li, Xueyuan Li, Qi Liu","doi":"10.1145/3548608.3559219","DOIUrl":"https://doi.org/10.1145/3548608.3559219","url":null,"abstract":"Unmanned Ground Vehicle (UGV) is currently the key research direction in the world. As an interdisciplinary complex, it covers machinery, computers, automation, ergonomics and other aspects. The application of UGV will liberate people from driving, reducing traffic accidents, and saving people's time and cost. As an important part of the UGV software system, trajectory prediction is very important to improve the security of UGV. This article classifies and summarizes various current prediction algorithms, puts forward the current problems of various algorithms, and looks forward to the future development direction.","PeriodicalId":201434,"journal":{"name":"Proceedings of the 2022 2nd International Conference on Control and Intelligent Robotics","volume":"83 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2022-06-24","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"132276327","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
There are some problems in urban cold chain logistics such as low timeliness and low customer value. This paper establishes the fresh cold chain logistics delivery routing optimization model with the minimum total cost and maximum potential customer value as the objective function, based on the perishable characteristics of fresh products and the customer value theory. The model considers vehicle load, mileage limit, and customer time window constraints. Then a hybrid genetic-simulated annealing algorithm is designed to solve the model, and an example is used to verify the effectiveness of the algorithm. In addition, the comparative analysis of the results obtained by the genetic algorithm and the genetic-simulated annealing algorithm shows that the latter can search for a better solution. The method proposed in this paper can maximize the value of customers and reduce the cost of delivery.
{"title":"Optimization of Routing Problem for Cold Chain Logistics Considering Customer Value","authors":"Feng Deng","doi":"10.1145/3548608.3559223","DOIUrl":"https://doi.org/10.1145/3548608.3559223","url":null,"abstract":"There are some problems in urban cold chain logistics such as low timeliness and low customer value. This paper establishes the fresh cold chain logistics delivery routing optimization model with the minimum total cost and maximum potential customer value as the objective function, based on the perishable characteristics of fresh products and the customer value theory. The model considers vehicle load, mileage limit, and customer time window constraints. Then a hybrid genetic-simulated annealing algorithm is designed to solve the model, and an example is used to verify the effectiveness of the algorithm. In addition, the comparative analysis of the results obtained by the genetic algorithm and the genetic-simulated annealing algorithm shows that the latter can search for a better solution. The method proposed in this paper can maximize the value of customers and reduce the cost of delivery.","PeriodicalId":201434,"journal":{"name":"Proceedings of the 2022 2nd International Conference on Control and Intelligent Robotics","volume":"53 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2022-06-24","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"133400602","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
In order to obtain the three-dimensional position of the target, and then measure the trajectory of the moving target, the problem of target recognition and positioning based on binocular stereo vision is studied. A binocular stereo vision system is built and MATLAB software is used to complete the calibration of the system. Through the checkerboard calibration method, the internal and external para-meters, relative attitude and distortion parameters of the left and right cameras are obtained. In the stereo matching stage, the image is preprocessed first, the RANSAC algorithm is used to eliminate the mismatch, and then the SIFT and SURF feature recognition algorithms are used to match the left and right images, and the pixel coordinates of the corresponding points are obtained, and then the three-dimensional space coordinates of the target point are obtained. and compared the measurement accuracy of the two feature recognition algorithms. The experimental results show that the method can accurately locate the target point and has high precision.
{"title":"Object Recognition and Localization Based on Binocular Stereo Vision","authors":"Fuqiang Wang, Libo Ding","doi":"10.1145/3548608.3559192","DOIUrl":"https://doi.org/10.1145/3548608.3559192","url":null,"abstract":"In order to obtain the three-dimensional position of the target, and then measure the trajectory of the moving target, the problem of target recognition and positioning based on binocular stereo vision is studied. A binocular stereo vision system is built and MATLAB software is used to complete the calibration of the system. Through the checkerboard calibration method, the internal and external para-meters, relative attitude and distortion parameters of the left and right cameras are obtained. In the stereo matching stage, the image is preprocessed first, the RANSAC algorithm is used to eliminate the mismatch, and then the SIFT and SURF feature recognition algorithms are used to match the left and right images, and the pixel coordinates of the corresponding points are obtained, and then the three-dimensional space coordinates of the target point are obtained. and compared the measurement accuracy of the two feature recognition algorithms. The experimental results show that the method can accurately locate the target point and has high precision.","PeriodicalId":201434,"journal":{"name":"Proceedings of the 2022 2nd International Conference on Control and Intelligent Robotics","volume":"3 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2022-06-24","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"131346973","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
京公网安备 11010802042870号
京ICP备2023020795号-1
扫码关注我们
求助内容：
应助结果提醒方式：