Preference-Based Multiobjective Optimization of Nonresonant Wireless Charging System

Desensitizing the misalignment and achieving low loss can improve the performance of wireless charging systems (WCS). A preference-based multiobjective optimization and resistance estimation algorithm based on the orthogonal table and multiple linear regression is used to achieve this. First, the working principle of a nonresonant WCS is analyzed to illustrate the relationship between WCS performance and mutual inductance along with coil resistance. Then, based on the orthogonal method and F-test, the parameters affecting the coil resistance are determined, and an estimation model for the coil resistance is established. Third, preference region and Latin hypercube sampling (LHS) are added to the nondominated sorting genetic algorithm II (NSGAII) method to enhance individual density within the preferred region without increasing the number of individuals and optimize the antimisalignment ability and coil resistance based on the model above of nested coils as an example using the proposed method. Compared to the NSGAII method, the mutual inductance fluctuation within 120 mm is reduced by 58% under similar conditions. Finally, the simulation and experimental results demonstrate that the proposed resistance estimation method has acceptable accuracy, and the proposed optimization method can optimize the horizontal antimisalignment ability of nested coils, and thus constant voltage/constant current charging is achieved.