{"title":"The nonlinear Choke Parameter Identification based on the Modified Artificial Neural Network","authors":"I. Orlovskyi, M. Fajfer","doi":"10.1109/PAEP49887.2020.9240877","DOIUrl":null,"url":null,"abstract":"The method of identification of the Nonlinear Choke Parameter (NCP) identification via a Modified artificial Recurrent Neural Network (MRNN) has been proposed and researched. The method makes it possible to use knowledge about the object for the synthesis of an MRNN structure and the computation of their coefficients from measured instantaneous values of currents and voltages in the circuit. NCP in the proposed MRNN were realized by expanding the input signal space of the network, using the normalized signals of polynomial terms. The proposed method can estimate MRNN structure to meets with NCP accuracy requirements. The NCP approximation in the one and a few function forms were also provided. The proposed method of NCP identification is based on the weighted MRNN coefficients. A comparison is made of the accuracy of identification of NCP for different MRNN structures and previously known measurement object data. An accuracy estimation for the models NCP in the form of a MRNN was made on mathematical models and on a real object.","PeriodicalId":240191,"journal":{"name":"2020 IEEE Problems of Automated Electrodrive. Theory and Practice (PAEP)","volume":"29 1","pages":"0"},"PeriodicalIF":0.0000,"publicationDate":"2020-09-21","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":"0","resultStr":null,"platform":"Semanticscholar","paperid":null,"PeriodicalName":"2020 IEEE Problems of Automated Electrodrive. Theory and Practice (PAEP)","FirstCategoryId":"1085","ListUrlMain":"https://doi.org/10.1109/PAEP49887.2020.9240877","RegionNum":0,"RegionCategory":null,"ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":null,"EPubDate":"","PubModel":"","JCR":"","JCRName":"","Score":null,"Total":0}
引用次数: 0
Abstract
The method of identification of the Nonlinear Choke Parameter (NCP) identification via a Modified artificial Recurrent Neural Network (MRNN) has been proposed and researched. The method makes it possible to use knowledge about the object for the synthesis of an MRNN structure and the computation of their coefficients from measured instantaneous values of currents and voltages in the circuit. NCP in the proposed MRNN were realized by expanding the input signal space of the network, using the normalized signals of polynomial terms. The proposed method can estimate MRNN structure to meets with NCP accuracy requirements. The NCP approximation in the one and a few function forms were also provided. The proposed method of NCP identification is based on the weighted MRNN coefficients. A comparison is made of the accuracy of identification of NCP for different MRNN structures and previously known measurement object data. An accuracy estimation for the models NCP in the form of a MRNN was made on mathematical models and on a real object.