Improved Two-Vector-Based Model Predictive Current Control with Online Parameter Identification for Doubly Salient Electromagnetic Machine

• Published in: 2023 26th International Conference on Electrical Machines and Systems (ICEMS) pp. 3048 - 3053
• Main Authors: Zhou, Xingwei, Zhan, Minhui, Guo, Yaowu, Niu, Shuangxia, Dai, Shangjian, Zhang, Li
• Format: Conference Proceeding
• Language: English
• Published: IEEE 05.11.2023
• Subjects: Current control; doubly salient electromagnetic machine (DSEM); model predictive control (MPC); online parameter identification; Parameter estimation; Prediction algorithms; Predictive models; Sensitivity analysis; Steady-state; Windings
• ISSN: 2642-5513
• DOI: 10.1109/ICEMS59686.2023.10344470

The traditional finite control set model predictive control (MPC) suffers from the problems of large current ripple and high system parameter dependence. To overcome these, an improved two-vector-based model predictive current control (TV-MPCC) with online parameters identification is proposed for doubly salient electromagnetic machine (DSEM). At first, the voltage vector combinations are localized and shortlisted based on the principle of differential-free beat current control, thus avoiding the complex computation of traditional traversal algorithms. Then, the impact of parameters mismatch is studied in depth, and the sensitivity of different parameters is obtained. Afterwards, an online identification based on MRAS is put forward to eliminate the influence of key parameters mismatch of self-inductance as well as the mutual inductance between armature and excitation windings. Finally, the validity and feasibility of the proposed strategies are verified by simulations under multiple operating conditions.