A Novel Online Parameter Identification Algorithm Designed for Deadbeat Current Control of the Permanent-Magnet Synchronous Motor
Zitan Wang, Jianyun Chai, Xuewei Xiang, Xudong Sun, Haifeng Lu
Abstract
The deadbeat control is one of the widely concerned control methods for the permanent-magnet synchronous motor (PMSM) due to its fast dynamic performance. However, its control performance relies heavily on the accuracy of the PMSM model parameters, which may vary with the operation cases. In this article, a novel online parameter identification algorithm is proposed for the PMSM deadbeat control. First, an identification model is established to estimate the parameter errors from the offsets of the deadbeat control, in which the nonlinearity of the voltage source inverter is fully concerned. Then, a novel “parameter perturbation method” is proposed for gathering the essential data to solve the rank deficient problem in the parameter identification process. It does not need to inject additional instructions in the control, resulting in no impact on the normal operation of the PMSM. The Adaline neural network is employed to achieve the online acquisition of parameter identification results. Finally, the effectiveness and superiority of the proposed algorithm are verified by experimental results on a PMSM platform.