Advanced Computational-Time Reduction Technology of Nonlinear Fourier-Based and Magnetic Circuit Hybrid Model

Given the rising demand in modern industry for high-performance permanent magnet (PM) machines and advanced control strategies, more efficient computation tools are an urgent necessity. To address nonlinear characteristics like magnetic saturation in PM machines, hybrid models (HMs) have seen rapid development in recent years. Among these, the nonlinear Fourier-based and magnetic circuit (MC) HMs offer advantages of high accuracy and clear physical significance. However, both the high-dimensional stiffness matrix in the Fourier model and tedious iteration process in the nonlinear MC model contribute to CPU usage issues and unacceptably time consumption. To cope with these issues, the author puts forth a series of efficient techniques aimed at reducing computational burden. These techniques are derived from both individual computations of the two sub-models and iterative calculations between them. They leverage aspects such as PM machine magnetic field distribution, efficient matrix-solving algorithms, nonlinear equation convergence algorithms, and so on. Furthermore, the author outlines tailored capable techniques corresponding to the distinct characteristics of inner and outer rotor structures, as well as surface-mounted and interior PM topologies. Through finite element modeling and prototype experiments, employing these techniques substantially decreases time consumption while maintaining accuracy.