Pub Date : 2021-08-01DOI: 10.1109/ICIEA51954.2021.9516261
Peng-fei Liu, Hongyu Long, Mingyao Ma
In order to solve the problem of power network coordination and management caused by the characteristics of distributed power distribution, this paper proposes a multi-objective optimal dispatching method for active distribution network based on cluster division. Firstly, the state estimation is used to eliminate bad data and obtain more accurate system parameters. Secondly, the multi-time scale integrated optimization of active and reactive power coordination, robust correction control optimization considering the statistical uncertainty of power supply and load prediction data and linear control cost optimization based on measurement data were established to dynamically adjust the distributed power cluster to achieve optimal scheduling. The research shows that this method can quickly converge the maximum deviation of data to less than 0.003 under the severe numerical conditions and reduce the network loss and voltage deviation to improve the economic benefit and power quality of the power grid.
{"title":"Research on Multi-objective Optimal Scheduling of Active Distribution Network Based on Cluster Partition","authors":"Peng-fei Liu, Hongyu Long, Mingyao Ma","doi":"10.1109/ICIEA51954.2021.9516261","DOIUrl":"https://doi.org/10.1109/ICIEA51954.2021.9516261","url":null,"abstract":"In order to solve the problem of power network coordination and management caused by the characteristics of distributed power distribution, this paper proposes a multi-objective optimal dispatching method for active distribution network based on cluster division. Firstly, the state estimation is used to eliminate bad data and obtain more accurate system parameters. Secondly, the multi-time scale integrated optimization of active and reactive power coordination, robust correction control optimization considering the statistical uncertainty of power supply and load prediction data and linear control cost optimization based on measurement data were established to dynamically adjust the distributed power cluster to achieve optimal scheduling. The research shows that this method can quickly converge the maximum deviation of data to less than 0.003 under the severe numerical conditions and reduce the network loss and voltage deviation to improve the economic benefit and power quality of the power grid.","PeriodicalId":6809,"journal":{"name":"2021 IEEE 16th Conference on Industrial Electronics and Applications (ICIEA)","volume":"51 1","pages":"1052-1056"},"PeriodicalIF":0.0,"publicationDate":"2021-08-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"74012974","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}
Pub Date : 2021-08-01DOI: 10.1109/ICIEA51954.2021.9516399
Juntao Pan, Rongrong Li, Sujun Wang, Weiwei Zhang
This paper deals with the problem designing fault detection observer for H2AT model in restricted frequency-domain specifications. Both disturbances and fault frequencies are assumed to be known and reside in a middle-frequency range. Different from previous results, the H2AT model is represented in the N-TS fuzzy model. It allows separating the unmeasured premise variables (UPV) in the local nonlinear consequent which provides a more effective way to deal with the UPV. In order to guarantee the best robustness to disturbances and sensitivity to faults, the fault detection observer combines the $H$- / $H$∞ performances. The design conditions are obtained based on Differential Mean Value Theorem (DMVT) property. Simulation result are conducted to demonstrate the viability and validity of the presented method.
{"title":"Fuzzy Fault Detection Observer Design for Head-Two-Arms-Trunk System","authors":"Juntao Pan, Rongrong Li, Sujun Wang, Weiwei Zhang","doi":"10.1109/ICIEA51954.2021.9516399","DOIUrl":"https://doi.org/10.1109/ICIEA51954.2021.9516399","url":null,"abstract":"This paper deals with the problem designing fault detection observer for H2AT model in restricted frequency-domain specifications. Both disturbances and fault frequencies are assumed to be known and reside in a middle-frequency range. Different from previous results, the H2AT model is represented in the N-TS fuzzy model. It allows separating the unmeasured premise variables (UPV) in the local nonlinear consequent which provides a more effective way to deal with the UPV. In order to guarantee the best robustness to disturbances and sensitivity to faults, the fault detection observer combines the $H$- / $H$∞ performances. The design conditions are obtained based on Differential Mean Value Theorem (DMVT) property. Simulation result are conducted to demonstrate the viability and validity of the presented method.","PeriodicalId":6809,"journal":{"name":"2021 IEEE 16th Conference on Industrial Electronics and Applications (ICIEA)","volume":"40 1","pages":"261-266"},"PeriodicalIF":0.0,"publicationDate":"2021-08-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"74475717","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}
Aiming at the subsynchronous oscillation problem of direct drive wind farm in low operating conditions, the eigenvalue analysis method and impedance analysis method are used to analyze two kinds of subsynchronous oscillation problems caused by different current loop parameters of direct drive wind turbine group dominated by current inner loop control. The results show that: for the first kind of subsynchronous oscillation, the dominant influencing factors are the power grid strength and the current inner loop control parameters; with the decrease of the power grid strength, the current inner loop control parameters decrease, weakening the system damping, and the oscillation frequency gradually decreases; for the second kind of subsynchronous oscillation, the dominant influencing factors are the current inner loop control parameters, with the decrease of the power grid strength The damping of the system is weakened and the oscillation frequency increases slightly. In EMTDC/PSCAD, the time domain simulation model of direct drive wind turbine group connected to AC power grid is built, which verifies the correctness and effectiveness of the theoretical analysis.
{"title":"Comparative analysis of two kinds of subsynchronous oscillation of direct drive PMSG based wind farm dominated by inner current loop","authors":"Wenchao Zhai, Qi Jia, Gangui Yan, Junxi Wang, Mohan Shen","doi":"10.1109/ICIEA51954.2021.9516184","DOIUrl":"https://doi.org/10.1109/ICIEA51954.2021.9516184","url":null,"abstract":"Aiming at the subsynchronous oscillation problem of direct drive wind farm in low operating conditions, the eigenvalue analysis method and impedance analysis method are used to analyze two kinds of subsynchronous oscillation problems caused by different current loop parameters of direct drive wind turbine group dominated by current inner loop control. The results show that: for the first kind of subsynchronous oscillation, the dominant influencing factors are the power grid strength and the current inner loop control parameters; with the decrease of the power grid strength, the current inner loop control parameters decrease, weakening the system damping, and the oscillation frequency gradually decreases; for the second kind of subsynchronous oscillation, the dominant influencing factors are the current inner loop control parameters, with the decrease of the power grid strength The damping of the system is weakened and the oscillation frequency increases slightly. In EMTDC/PSCAD, the time domain simulation model of direct drive wind turbine group connected to AC power grid is built, which verifies the correctness and effectiveness of the theoretical analysis.","PeriodicalId":6809,"journal":{"name":"2021 IEEE 16th Conference on Industrial Electronics and Applications (ICIEA)","volume":"29 1","pages":"1585-1590"},"PeriodicalIF":0.0,"publicationDate":"2021-08-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"72603782","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}
Pub Date : 2021-08-01DOI: 10.1109/ICIEA51954.2021.9516357
Hanting Yin, Quan Jiang, Chen Tang
This paper introduces the remote current and voltage monitoring system based on cloud storage and 5G wireless communication technology, including acquisition module, data processing module, Ethernet module and terminal; Acquisition module acquisition device of three phase current and three-phase voltage being measured, by matching the operational amplifier to convert signals AD input circuit, then the measured data of voltage and current fast Fourier transform, then the main frequency component amplitude and phase of signal transmission to the Ethernet module, realize the transmission of high compression test data. The data of voltage and current can be read and recorded by accessing Ethernet module through terminal and Internet of Things platform. Thus, a remote automatic monitoring system for voltage and current of electromechanical equipment and their operating conditions is proposed.
{"title":"Remote current and voltage monitoring system based on cloud storage and wireless communication technology","authors":"Hanting Yin, Quan Jiang, Chen Tang","doi":"10.1109/ICIEA51954.2021.9516357","DOIUrl":"https://doi.org/10.1109/ICIEA51954.2021.9516357","url":null,"abstract":"This paper introduces the remote current and voltage monitoring system based on cloud storage and 5G wireless communication technology, including acquisition module, data processing module, Ethernet module and terminal; Acquisition module acquisition device of three phase current and three-phase voltage being measured, by matching the operational amplifier to convert signals AD input circuit, then the measured data of voltage and current fast Fourier transform, then the main frequency component amplitude and phase of signal transmission to the Ethernet module, realize the transmission of high compression test data. The data of voltage and current can be read and recorded by accessing Ethernet module through terminal and Internet of Things platform. Thus, a remote automatic monitoring system for voltage and current of electromechanical equipment and their operating conditions is proposed.","PeriodicalId":6809,"journal":{"name":"2021 IEEE 16th Conference on Industrial Electronics and Applications (ICIEA)","volume":"315 1","pages":"189-194"},"PeriodicalIF":0.0,"publicationDate":"2021-08-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"74405202","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}
Pub Date : 2021-08-01DOI: 10.1109/ICIEA51954.2021.9516332
Y. S. Hagh, A. Fekih, H. Handroos
This paper proposes a novel robust finite synergetic control approach for nonlinear systems subject to time-varying mismatched disturbances. Its main objective is to guarantee the finite-time convergence of the states while eliminating the singularity problem and providing a chattering-free response despite the disturbances. The controller is formulated based on a novel non-singular terminal sliding manifold and the disturbances are estimated using a nonlinear finite time disturbance observer. By estimating the values of the mismatched disturbances and uncertainties, a novel synergetic manifold is introduced which compensates for an estimate of the disturbances. This yields a robust Finite Non-singular Terminal Synergistic Control (FNTSC) that is capable of counteracting the effects of the time-varying mismatched disturbances. System stability is established using the Lyapunov stability theory. The effectiveness and performance of the proposed approach is assessed using a four-bar linkage mechanism as a study case. The obtained results confirmed the robustness, finite time convergence and chattering free dynamics of the proposed controller.
{"title":"Robust finite non-singular terminal synergetic control for second order nonlinear systems subject to time-varying mismatched disturbances","authors":"Y. S. Hagh, A. Fekih, H. Handroos","doi":"10.1109/ICIEA51954.2021.9516332","DOIUrl":"https://doi.org/10.1109/ICIEA51954.2021.9516332","url":null,"abstract":"This paper proposes a novel robust finite synergetic control approach for nonlinear systems subject to time-varying mismatched disturbances. Its main objective is to guarantee the finite-time convergence of the states while eliminating the singularity problem and providing a chattering-free response despite the disturbances. The controller is formulated based on a novel non-singular terminal sliding manifold and the disturbances are estimated using a nonlinear finite time disturbance observer. By estimating the values of the mismatched disturbances and uncertainties, a novel synergetic manifold is introduced which compensates for an estimate of the disturbances. This yields a robust Finite Non-singular Terminal Synergistic Control (FNTSC) that is capable of counteracting the effects of the time-varying mismatched disturbances. System stability is established using the Lyapunov stability theory. The effectiveness and performance of the proposed approach is assessed using a four-bar linkage mechanism as a study case. The obtained results confirmed the robustness, finite time convergence and chattering free dynamics of the proposed controller.","PeriodicalId":6809,"journal":{"name":"2021 IEEE 16th Conference on Industrial Electronics and Applications (ICIEA)","volume":"32 1","pages":"228-233"},"PeriodicalIF":0.0,"publicationDate":"2021-08-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"76578117","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 high-power applications, when the current capacity needs to be further expanded, the method of using insulated gate bipolar transistors in parallel is usually used. However, due to the influence of parasitic parameters, the current of each branch of the paralleled IGBTs may be imbalanced. The imbalance of dynamic and static currents can cause overheating and overcurrent damage to the device. This paper studies a driving method that uses differential gate voltage and synchronous pulse to improve the current imbalance during the turn-on process of paralleled IGBTs. The effectiveness of the strategy is verified by simulation. The experimental results show that the voltage differential and pulse synchronous driving method can effectively reduce the dynamic current imbalance between paralleled IGBTs.
{"title":"A Voltage Differential and Pulse Synchronous Driving Control of Paralleled IGBTs for Current Balance Improving","authors":"Yuhan Wu, Guangang Gao, Yixin Liu, Feng Mu, Xianjin Huang","doi":"10.1109/ICIEA51954.2021.9516174","DOIUrl":"https://doi.org/10.1109/ICIEA51954.2021.9516174","url":null,"abstract":"In high-power applications, when the current capacity needs to be further expanded, the method of using insulated gate bipolar transistors in parallel is usually used. However, due to the influence of parasitic parameters, the current of each branch of the paralleled IGBTs may be imbalanced. The imbalance of dynamic and static currents can cause overheating and overcurrent damage to the device. This paper studies a driving method that uses differential gate voltage and synchronous pulse to improve the current imbalance during the turn-on process of paralleled IGBTs. The effectiveness of the strategy is verified by simulation. The experimental results show that the voltage differential and pulse synchronous driving method can effectively reduce the dynamic current imbalance between paralleled IGBTs.","PeriodicalId":6809,"journal":{"name":"2021 IEEE 16th Conference on Industrial Electronics and Applications (ICIEA)","volume":"35 6 1","pages":"837-842"},"PeriodicalIF":0.0,"publicationDate":"2021-08-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"77687838","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}
Pub Date : 2021-08-01DOI: 10.1109/ICIEA51954.2021.9516402
Tao Wang, Dong Liu, Yang Liu, Zhenwei Li, Le Xie, Qilong Jiang
In view of the problem that the characteristic extraction of subway plug door fault diagnosis is too high, which leads to low diagnostic accuracy, a mixed feature selection method based on ReliefF algorithm and BGWO (binary grey wolf optimizer, BGWO) was proposed. Firstly, multiple domains feature extraction were carried out on the collected current signal of the subway plug door motor, and an original fault feature set describing the subway plug door fault was obtained. Afterwards, the ReliefF algorithm was used to evaluate the extracted original fault feature weights and screened out the less relevant features. Finally, the classification error rate of GWO (grey wolf optimizer, GWO)-SVM (support vector machine, SVM) is used as the fitness value, and BGWO is used as the feature selection algorithm to perform feature selection on the feature subset obtained by the ReilefF algorithm. The data collected in a Metro Depot in Jiangsu Province is used as the original data set for verification. The experimental results show that the method can screen out low dimensional fault feature sets with high correlation, low redundancy and high fault identification, it can effectively improve the accuracy of subway plug door fault diagnosis.
针对地铁塞门故障诊断中特征提取量过高导致诊断准确率不高的问题，提出了一种基于ReliefF算法和BGWO (binary grey wolf optimizer, BGWO)的混合特征选择方法。首先，对采集到的地铁塞门电机电流信号进行多域特征提取，得到描述地铁塞门故障的原始故障特征集;然后，利用ReliefF算法对提取的原始故障特征权重进行评估，筛选出相关度较低的特征。最后，以GWO(灰狼优化器，灰狼优化器)-SVM(支持向量机，支持向量机)的分类错误率作为适应度值，以BGWO作为特征选择算法，对ReilefF算法得到的特征子集进行特征选择。以江苏省某地铁车辆段的数据作为原始数据集进行验证。实验结果表明，该方法能够筛选出具有高相关性、低冗余度和高故障识别率的低维故障特征集，有效提高了地铁塞门故障诊断的准确率。
{"title":"Fault feature selection of Subway Plug Door Based on ReliefF and BGWO","authors":"Tao Wang, Dong Liu, Yang Liu, Zhenwei Li, Le Xie, Qilong Jiang","doi":"10.1109/ICIEA51954.2021.9516402","DOIUrl":"https://doi.org/10.1109/ICIEA51954.2021.9516402","url":null,"abstract":"In view of the problem that the characteristic extraction of subway plug door fault diagnosis is too high, which leads to low diagnostic accuracy, a mixed feature selection method based on ReliefF algorithm and BGWO (binary grey wolf optimizer, BGWO) was proposed. Firstly, multiple domains feature extraction were carried out on the collected current signal of the subway plug door motor, and an original fault feature set describing the subway plug door fault was obtained. Afterwards, the ReliefF algorithm was used to evaluate the extracted original fault feature weights and screened out the less relevant features. Finally, the classification error rate of GWO (grey wolf optimizer, GWO)-SVM (support vector machine, SVM) is used as the fitness value, and BGWO is used as the feature selection algorithm to perform feature selection on the feature subset obtained by the ReilefF algorithm. The data collected in a Metro Depot in Jiangsu Province is used as the original data set for verification. The experimental results show that the method can screen out low dimensional fault feature sets with high correlation, low redundancy and high fault identification, it can effectively improve the accuracy of subway plug door fault diagnosis.","PeriodicalId":6809,"journal":{"name":"2021 IEEE 16th Conference on Industrial Electronics and Applications (ICIEA)","volume":"36 1","pages":"1346-1351"},"PeriodicalIF":0.0,"publicationDate":"2021-08-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"79807581","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}
Pub Date : 2021-08-01DOI: 10.1109/ICIEA51954.2021.9516053
Nalini Rizkyta Nusantika, Xiaoguang Hu, Jin Xiao
Overhead power line icing disaster is one of the most noteworthy problem which highly damage the safety of power grid, so it is needed for monitoring in the area. Recently, computer vision and Unmanned Aerial Vehicle (UAV) is developing rapidly, which is suitable for the monitoring of transmission lines icing. Edge detection method is the main priority for automatically measurement of monitoring system based on UAV. There are numerous edge detection methods, but Canny edge detector is considered to be more reliable than some traditional methods. In this paper, a method of improvement Canny using hybrid technique is proposed for more accurate measurements. Traditional Canny operator and improvement Canny operator are compared and analyzed in the simulation. For the comparison the result used are mean square error (MSE), root mean square error (RMSE) and peak signal to noise ratio (PSNR). The simulation results show that the improvement Canny operator using the hybrid technique better than traditional Canny.
{"title":"Improvement Canny Edge Detection for the UAV Icing Monitoring of Transmission Line Icing","authors":"Nalini Rizkyta Nusantika, Xiaoguang Hu, Jin Xiao","doi":"10.1109/ICIEA51954.2021.9516053","DOIUrl":"https://doi.org/10.1109/ICIEA51954.2021.9516053","url":null,"abstract":"Overhead power line icing disaster is one of the most noteworthy problem which highly damage the safety of power grid, so it is needed for monitoring in the area. Recently, computer vision and Unmanned Aerial Vehicle (UAV) is developing rapidly, which is suitable for the monitoring of transmission lines icing. Edge detection method is the main priority for automatically measurement of monitoring system based on UAV. There are numerous edge detection methods, but Canny edge detector is considered to be more reliable than some traditional methods. In this paper, a method of improvement Canny using hybrid technique is proposed for more accurate measurements. Traditional Canny operator and improvement Canny operator are compared and analyzed in the simulation. For the comparison the result used are mean square error (MSE), root mean square error (RMSE) and peak signal to noise ratio (PSNR). The simulation results show that the improvement Canny operator using the hybrid technique better than traditional Canny.","PeriodicalId":6809,"journal":{"name":"2021 IEEE 16th Conference on Industrial Electronics and Applications (ICIEA)","volume":"318 1","pages":"1838-1843"},"PeriodicalIF":0.0,"publicationDate":"2021-08-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"80165515","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}
The paper proposes a current-sensorless finite-control-set model predictive control (FCS-MPC) method for three-level voltage source inverter (3L-VSI). Firstly, based on the proposed control model, the predictive variables of traditional FCS-MPC method are converted into capacitor voltage and current. Then, in order to eliminate current sensor completely, a capacitor current observer is utilized to estimate the capacitor current, which could reduce system cost and enhance reliability. Besides, the estimated current is brought into the prediction formula, and the optimal voltage vector is selected by minimizing cost function and applied to the control of 3L-VSI. Finally, simulation comparison results of traditional and current-sensorless FCS-MPC demonstrate the effectiveness of the proposed control scheme.
{"title":"Current-sensorless Finite-Control-Set Model Predictive Control for Three-level Voltage Source Inverter","authors":"Kaixin Wang, Shen-Yaur Chen, Yuhang Tang, Yong Yang, Mingdi Fan, Menxi Xie","doi":"10.1109/ICIEA51954.2021.9516408","DOIUrl":"https://doi.org/10.1109/ICIEA51954.2021.9516408","url":null,"abstract":"The paper proposes a current-sensorless finite-control-set model predictive control (FCS-MPC) method for three-level voltage source inverter (3L-VSI). Firstly, based on the proposed control model, the predictive variables of traditional FCS-MPC method are converted into capacitor voltage and current. Then, in order to eliminate current sensor completely, a capacitor current observer is utilized to estimate the capacitor current, which could reduce system cost and enhance reliability. Besides, the estimated current is brought into the prediction formula, and the optimal voltage vector is selected by minimizing cost function and applied to the control of 3L-VSI. Finally, simulation comparison results of traditional and current-sensorless FCS-MPC demonstrate the effectiveness of the proposed control scheme.","PeriodicalId":6809,"journal":{"name":"2021 IEEE 16th Conference on Industrial Electronics and Applications (ICIEA)","volume":"181 1","pages":"1739-1744"},"PeriodicalIF":0.0,"publicationDate":"2021-08-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"80225043","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}
The inverter fault diagnosis based on BP neural network can fall into local minimum and overfitting. To solve these problems, we propose a fault diagnosis method based on BP neural network optimized by cross entropy and L2 regularization. In this proposed method, the quadratic cost function is replaced by the cross entropy cost function, which avoids the influence of the partial derivative of the activation function. L2 regularization is used to adjust network toward the small weight distribution. This method reduces the possibility of falling into local minimum and overfitting. The experimental results show that the optimized neural network can improve the accuracy of inverter fault diagnosis.
{"title":"Inverter Fault Diagnosis Based on Optimized BP Neural Network","authors":"Xing Liu, Mingyao Ma, Weisheng Guo, Xuesong Meng, Pengbo Xiong","doi":"10.1109/ICIEA51954.2021.9516072","DOIUrl":"https://doi.org/10.1109/ICIEA51954.2021.9516072","url":null,"abstract":"The inverter fault diagnosis based on BP neural network can fall into local minimum and overfitting. To solve these problems, we propose a fault diagnosis method based on BP neural network optimized by cross entropy and L2 regularization. In this proposed method, the quadratic cost function is replaced by the cross entropy cost function, which avoids the influence of the partial derivative of the activation function. L2 regularization is used to adjust network toward the small weight distribution. This method reduces the possibility of falling into local minimum and overfitting. The experimental results show that the optimized neural network can improve the accuracy of inverter fault diagnosis.","PeriodicalId":6809,"journal":{"name":"2021 IEEE 16th Conference on Industrial Electronics and Applications (ICIEA)","volume":"21 1","pages":"803-808"},"PeriodicalIF":0.0,"publicationDate":"2021-08-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"81732543","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：