篇名 | MRAS Speed Observer for Permanent Magnet Linear Synchronous Motor Based on RBF Neural Networks |
---|---|
卷期 | 31:2 |
作者 | Qi-Yong Chen 、 Rong-Kun Wang 、 Fei Man 、 Kai Peng 、 Bing-Tao Hu 、 Dong Yan |
頁次 | 001-011 |
關鍵字 | model reference adaptive system 、 permanent magnet linear synchronous motor 、 radial basis function neural network 、 speed sensorless control 、 EI 、 MEDLINE 、 Scopus |
出刊日期 | 202004 |
DOI | 10.3966/199115992020043102001 |
In order to realize speed sensorless control of permanent magnet synchronous linear motors (PMLSMs) in the whole speed range, a model reference adaptive system (MRAS) speed observer based on radial basis function neural network (RBFNN) is proposed to observe the speed information of PMLSMs in this paper. The speed observer uses a RBFNN as its adaptive mechanism and achieves the speed accurate estimation by taking advantage of strong nonlinear approximation ability of the RBFNN. In addition, a RBFNN identifier is designed to provide gradient information for the RBFNN in the speed observer. Simulation and experiment indicate that the speed observer can accurately observe the speed in the whole speed range for PMLSMs and has good dynamic response characteristics.