Fusion Predictive Control Based on Uncertain Algorithm for PMSM of Brake-by-Wire System

Abstract

Brake-by-wire (BBW) system is an important safe component of intelligent vehicles. The fast and stable current control of permanent magnet synchronous motor (PMSM) is a key factor that ensures the dynamic response performance of BBW. Hence, a fusion predictive control (FPC) method based on uncertain algorithm is proposed, which solves the parameter sensitivity problem in traditional deadbeat current control of PMSM. It contains a fusion predictive controller, a flux observer, and an uncertain controller. The FPC integrates forward linear prediction (FLP) and deadbeat prediction (DP) to improve the current response speed. Considering the model parameter mismatch of the controller, the uncertain controller can adjust the reliability of FLP and DP in the FPC in real time through the uncertainty measurement. On the basis of the whole system, a flux observer is designed, which can dynamically adjust the parameters of flux to satisfy the current loop. Both simulation and experimental results show that this method reduces the sensitivity of system parameters and improves the dynamic performance of the current loop (up to 12.56%).