EMI Filter Parameter Design Optimization Based on Improved Particle Swarm Algorithm
Jian Zheng,
Fang Liu,
Shuangyin Cheng,
Zhenwei Zhang and
Zhijie Chen
When designing the parameters of electromagnetic interference filters in BUCK rectifiers using a particle swarm optimization algorithm, the resulting solution set tends to fall into local optima, resulting in a lower cost-performance ratio for the filters. Therefore, an optimization scheme for the parameter design is proposed. First, a multi-objective optimization model is established, with common-mode capacitance and inductance, as well as differential-mode capacitance as decision variables, conducted noise and cost as the objective function, and conducted noise limits and leakage current as constraints. Next, three improvements are made to the conventional algorithm, classifying initialization for the particle population, dividing the total number of iterations into stages with differentiated handling, and implementing an adaptive early termination for iterations, thereby an improved algorithm is obtained. Finally, both the improved and conventional algorithms are applied to solve the optimization model, yielding two types of optimal solutions: performance-prioritized and cost-prioritized solutions. The comparison results show that, between the two optimal solutions, the performance of the improved algorithm is similar to that of the conventional algorithm, but its cost is reduced by 28.37% and 53.14%, respectively. Meanwhile, the Pareto solution set obtained by the improved algorithm is more widely distributed, avoiding local optima, and the iterative convergence efficiency of the improved algorithm was improved by 81%.