Online Parameter Estimation of Surface‐Mounted Permanent Magnet Synchronous Machine Based on Variable‐Step Neural Network. Issue 2 (20th October 2021)

• Record Type: Journal Article
• Authors: Tang, Chen; Jiang, Quan; Ma, Run
• Is Part Of: IEEJ transactions on electrical and electronic engineering. Volume 17:Issue 2(2022)
• Page Start: 252
• Page End: 260
• Publication Date: 2021-10-20
• Subjects: surface‐mounted permanent magnet synchronous machine -- parameter estimation -- Adaline neural network -- inductance; Electrical engineering -- Periodicals; Electronics -- Periodicals; 621.3
• DOI: 10.1002/tee.23499
• Languages: English
• ISSNs: 1931-4973

Abstract

Abstract : This paper proposes a multiparameter estimation method for surface‐mounted permanent magnet synchronous machine (SPMSM) under constant load torque, of which the variations of dq‐ axis inductances and rotor flux linkage have been taken into consideration. Second, the voltage equations will be transformed to eliminate the distorted voltage caused by the voltage‐source inverter nonlinearity. On this basis, an improved variable‐step Adaline neural network algorithm based on hyperbolic tangent function is proposed to estimate the parameters of SPMSM online, which can increase the convergence speed and reduce the steady‐state errors compared to the traditional algorithm. Finally, in view of the high sensitivity of inductance, another new method that can reduce the fluctuation rate is proposed. In addition, the effectiveness of these methods is verified through experiments on a small‐power SPMSM. © 2021 Institute of Electrical Engineers of Japan. Published by Wiley Periodicals LLC.