Dimensionality-Reduced Line-Constrained EMPC With Portability Applied in the Current Loop of IPMSMs

Line-constrained explicit model predictive control (EMPC) offers enhanced algorithmic efficiency, addressing the challenges posed by the deployment of conventional EMPC strategies on low-cost devices. Nevertheless, the memory consumption and computational time of this method considerably surpass most control algorithms. A further limitation is the need for offline recalculations when the controlled object changes, impacting the algorithm's adaptability. This article introduces a dimensionality reduced line-constrained EMPC (DRLC EMPC) tailored for the current loop control of interior permanent magnet synchronous motors. By innovatively refashioning the predictive model, the system's dimensionality undergoes further reduction. Importantly, parameters pertinent to the controlled object are externalized from the offline computations, obviating the necessity for matrix regeneration during system alterations, thereby enhancing the algorithm's portability. In addition, to achieve enhanced control performance, the proposed method employs an extended Kalman filter for observing the disturbance terms. The frequency-domain attributes of the proposed method are also explored to elucidate its advantages in the electro-drive realm. Experimental analyses demonstrate the superior efficacy of the DRLC EMPC.