{"title":"Low-resource CNN Accelerator Based on FPGA","authors":"Sheng An, Jijie Deng","doi":"10.1109/IMCEC55388.2022.10019956","DOIUrl":"https://doi.org/10.1109/IMCEC55388.2022.10019956","url":null,"abstract":"Convolutional Neural Networks (CNNs) are deployed on the Internet of Things (IoT) edge devices with strict restrictions on resource and power consumption. Most of the work's optimization focuses on inference speed and average precision, ignoring resource and power consumption optimization. This paper proposes the strategies of weight quantization, parameter reordering, block convolution multiplexing, and $text{CONV}+text{POOL}$ ‘ to implement an FPGA-based CNN accelerator with low-resource and low-power. The CNN algorithm is designed in High Level Synthesis tool (HLS), which can realize the FPGA development with high-level language to achieve high development efficiency and flexibility. The experimental results show that our implementation has achieved 2.812W of low power consumption for executing YOLOV2 on the PYNQ-Z2, consuming only 2052Kb of on-chip BRAMs, which is better than the current work in terms of resource consumption. The performance is about 42.31GOP/s, which is much higher than CPU (ARM-A9).","PeriodicalId":423151,"journal":{"name":"2022 IEEE 5th Advanced Information Management, Communicates, Electronic and Automation Control Conference (IMCEC)","volume":"5 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2022-12-16","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"124649428","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}
{"title":"Positive and Negative Association Rules Algorithm Based on Improved Interest Measure","authors":"Suzhi Zhang, Haidi Fan, Yawen Zhang, Zijian Bai, Xiang Lin, Weide Liang","doi":"10.1109/IMCEC55388.2022.10019871","DOIUrl":"https://doi.org/10.1109/IMCEC55388.2022.10019871","url":null,"abstract":"There are issues with the mining of association rules of traditional Chinese medicine prescriptions, such as unclear rule strength, incomplete decision analysis, and low practicability. This paper introduces the concept of interest degree and negative association rules, as well as combining positive association rules to analyze traditional Chinese medicine comprehensively. The strength of the rules is determined by the correlation hidden in the prescription. In this paper, an improved correlation interest degree measure is introduced, which effectively improves the uneven value range of commonly used interest measure measures, lacks zero invariance, and is prone to contradictory rules. Experiments reveal that the algorithm can more precisely differentiate between positive and negative rules as well as strong and weak rules, thereby enhancing the efficacy and applicability of prescription association rules.","PeriodicalId":423151,"journal":{"name":"2022 IEEE 5th Advanced Information Management, Communicates, Electronic and Automation Control Conference (IMCEC)","volume":"88 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2022-12-16","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"124751584","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}
Wang Xiaojin, Sun Shucai, Xie Yehua, Jiang Tao, Liu Hongkun
{"title":"Research on Data Standardization and Unified Data Interface Based on Digital Station System","authors":"Wang Xiaojin, Sun Shucai, Xie Yehua, Jiang Tao, Liu Hongkun","doi":"10.1109/IMCEC55388.2022.10019942","DOIUrl":"https://doi.org/10.1109/IMCEC55388.2022.10019942","url":null,"abstract":"At present, the development of energy and power is faced with a major strategic task of ensuring safe and reliable supply, accelerating the clean and low-carbon transformation, and helping to achieve the “dual carbon” goal. The construction of digital station information system bears the historical mission of energy transformation, has great and far-reaching significance, is an important part of a clean, low-carbon, safe and efficient energy system, and has new features such as clean, low-carbon, safe and controllable, flexible and efficient, intelligent, friendly, open and interactive. In this paper, the equipment model data standardization technology and unified data interface technology based on the digital station information system are studied to solve the problems faced in the construction of the digital station information system, such as multiple islands of information, multiple fault patterns, and low data value. According to the principle of unified collection, storage and utilization of distribution network information, the SG-CIM model specification is used as the basis for modeling to provide standardized data; The unified data interface of digital station information is formed through unified basic software and hardware facilities, unified equipment information model, and unified application and development specifications to achieve equipment centric panoramic data aggregation and sharing, and visual presentation of operation and maintenance business. Realize the coordination and interaction of digital station information system, and provide a strong support way for the country to promote the “double carbon” goal.","PeriodicalId":423151,"journal":{"name":"2022 IEEE 5th Advanced Information Management, Communicates, Electronic and Automation Control Conference (IMCEC)","volume":"78 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2022-12-16","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"130518743","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}
{"title":"A Review of Spectral Estimation in Underwater Acoustic Signal Processing","authors":"C. Peng","doi":"10.1109/IMCEC55388.2022.10019921","DOIUrl":"https://doi.org/10.1109/IMCEC55388.2022.10019921","url":null,"abstract":"This paper first introduces the direct method of classical power spectrum estimation, and then proposes two improved classical power spectrum estimation methods: Welch Method and MultiTaper (MTM) Method. Then the Maximum Entropy Spectral Estimation (based on Burg) in parametric spectral estimation is introduced. Then the power spectrum is extended in the spatial domain, and two kinds of spatial spectrum estimation methods, MUSIC Method and ESPRIT method, and their improved algorithms are introduced. All of the above methods have been simulated in Matlab environment. We have compared various spectrum estimation methods in detail, and analyzed and summarized their respective advantages and disadvantages. Finally, the paper points out the shortcomings of current power spectrum estimation and the future development direction.","PeriodicalId":423151,"journal":{"name":"2022 IEEE 5th Advanced Information Management, Communicates, Electronic and Automation Control Conference (IMCEC)","volume":"10 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2022-12-16","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"126952009","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}
{"title":"Improved SSD algorithm for mask wearing detection","authors":"Zhao Jin, Yong Lin, Yian Su, Letian Zhong","doi":"10.1109/IMCEC55388.2022.10020138","DOIUrl":"https://doi.org/10.1109/IMCEC55388.2022.10020138","url":null,"abstract":"Since the outbreak of the epidemic, it is extremely important to wear a mask correctly in the natural environment. Therefore, this paper proposes an improved SSD detection method. This paper takes the original SSD algorithm as the overall framework, and uses the lightweight network MobieNet-V2 to replace the original VGG-16 as the backbone network;In order not to compensate for the loss, the attention mechanism senet is added to improve its accuracy;And K-Means algorithm is used to cluster and generate anchor boxes suitable for data sets;Finally, Focal Loss is used to replace the Softmax of the original network as a new loss function. The experimental results show that compared with the original network, The improved network has better mAP, 8.9% higher than the original model.","PeriodicalId":423151,"journal":{"name":"2022 IEEE 5th Advanced Information Management, Communicates, Electronic and Automation Control Conference (IMCEC)","volume":"9 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2022-12-16","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"126989116","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}
{"title":"An Overview of PAI: Distributed Machine Learning Platform","authors":"Fan Meng, Kunrong Hu, Xiaotao Zhou","doi":"10.1109/IMCEC55388.2022.10020005","DOIUrl":"https://doi.org/10.1109/IMCEC55388.2022.10020005","url":null,"abstract":"Due to the size, complexity, and rapid change of big data, the need for faster training and data processing, and the growing inability of conventional single-machine training methods to match the demands. Recently, some distributed AI platforms have appeared for machine learning. In this article, we describe the OpenPAI distributed machine learning platform's architecture and examine its capabilities and design. We present a summary of the current issues in distributed machine learning and discuss potential future trends.","PeriodicalId":423151,"journal":{"name":"2022 IEEE 5th Advanced Information Management, Communicates, Electronic and Automation Control Conference (IMCEC)","volume":"22 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2022-12-16","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"130580183","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}
{"title":"Research on Verification Method of Logistics Matching Process Optimization Scheme Based on Timed Petri Net","authors":"Zhaoyu Chen, Shuwen Wang, Ying Song","doi":"10.1109/IMCEC55388.2022.10020100","DOIUrl":"https://doi.org/10.1109/IMCEC55388.2022.10020100","url":null,"abstract":"Optimizing the logistics matching process can enhance user satisfaction, enhance order processing capability and improve the degree of logistics automation, which is of great significance to promote economic and social development. After optimizing the logistics matching process, it is difficult to verify the feasibility of the optimization scheme in a short time through practice due to the rapid changes in the market. The Timed Petri Net theory, with the introduction of time parameters, can be used to model and simulate different schemes of the logistics matching process. By comparing and analyzing the simulation results of the optimization scheme and the original scheme, a conclusion on the verification of the optimization scheme is drawn, which provides a new idea for the feasibility verification of the logistics matching process optimization scheme.","PeriodicalId":423151,"journal":{"name":"2022 IEEE 5th Advanced Information Management, Communicates, Electronic and Automation Control Conference (IMCEC)","volume":"24 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2022-12-16","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"124178559","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}
{"title":"Semantic Segmentation based network for 6D pose estimation","authors":"Dan Wang, Nan Xu, Xiaobo Lin, Zijiang Wang","doi":"10.1109/IMCEC55388.2022.10020047","DOIUrl":"https://doi.org/10.1109/IMCEC55388.2022.10020047","url":null,"abstract":"Recognizing the position and pose of targets is important for robots to perceive the environment and is an important prerequisite for robots to complete intelligent grasping. In order to reduce the search scope of 6D space and increase the accuracy in the case of shade or truncation, we also use the pixel-level prediction, predicting confidence scores. Besides that, a parallel branch network is proposed which includes a task flow processing semantic object segmentation label for each point in the forecast. which is very important for dealing with objects in complex scenes. Experimental results show that the proposed method is effective in predicting 6D pose of occluded objects.","PeriodicalId":423151,"journal":{"name":"2022 IEEE 5th Advanced Information Management, Communicates, Electronic and Automation Control Conference (IMCEC)","volume":"107 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2022-12-16","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"121138504","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}
Yu Shen, Siyu Yang, Zhensong Zeng, Dingfei Que, Yu Wang
{"title":"Frequency Control Strategy of the Basic Round of Adjustable Energy Storage Participating in Power System","authors":"Yu Shen, Siyu Yang, Zhensong Zeng, Dingfei Que, Yu Wang","doi":"10.1109/IMCEC55388.2022.10020109","DOIUrl":"https://doi.org/10.1109/IMCEC55388.2022.10020109","url":null,"abstract":"With the large-scale development of energy storage and electric vehicles, its installed capacity is increasing. Energy storage quickly adjusts the output power to participate in frequency control and improve the dynamic frequency characteristics of the power system. Based on the active power deficiency caused by DC blocking under different operation modes, the emergency control strategy of adjustable energy storage participating in frequency control in the basic round is proposed. Then, the configuration model of the capacity and action frequency threshold value of adjustable energy storage participating in frequency control is proposed. The configuration model is solved by the dichotomy method. Next, the energy storage power control margin index is defined to characterize the security margin and whether the under-frequency load shedding is triggered when energy storage power control is adopted between the blank interval of the second defense line and the third defense line. Finally, the effectiveness and adaptability of the proposed method are verified by Fujian provincial power grid.","PeriodicalId":423151,"journal":{"name":"2022 IEEE 5th Advanced Information Management, Communicates, Electronic and Automation Control Conference (IMCEC)","volume":"30 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2022-12-16","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"114208423","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}
Chuming Zhang, Guojun Wei, Xuan Wang, Yu Cao, Mei-lin Xie
{"title":"Analysis and Design of Fault Prediction and Health Assessment System for the optoelectronic Equipment","authors":"Chuming Zhang, Guojun Wei, Xuan Wang, Yu Cao, Mei-lin Xie","doi":"10.1109/IMCEC55388.2022.10019826","DOIUrl":"https://doi.org/10.1109/IMCEC55388.2022.10019826","url":null,"abstract":"In recent years, due to the rise of emerging technologies such as integrated circuits, artificial intelligence and big data, Optical measurement equipment performance is also constantly improving, more and more new technology, new processes, new materials continue to appear. Therefore, the maintenance of the optoelectronic equipment is also much more difficult. Based on these problems, this paper investigates the application of fault prediction and health management in optoelectronic devices. The main analysis and design of the system failure prediction and health assessment, and in response to the needs of optoelectronic devices for the life analysis, health status determination, fault prediction alarms, etc, designed a failure prediction and health assessment system for optoelectronic devices.","PeriodicalId":423151,"journal":{"name":"2022 IEEE 5th Advanced Information Management, Communicates, Electronic and Automation Control Conference (IMCEC)","volume":"83 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2022-12-16","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"124115731","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}