Frequency Control System Electromagnetic Compatibility And Electromagnetic Optimization Algorithm

Thanks to the support of the National Natural Science Foundation of China under Grand No. 50277036 for "The Pulse Electromagnetic Fields and System-Level EM Radiation of Electric Vehicles", this thesis concentrated on the design of the variable voltage and variable frequency (VVVF) system and the development of a genetic algorithm for vector optimal design problems. The work includes mainly four parts: A brief introduction of the frequency-converter, the EMC design of the VVVF system and the national standard of the electromagnetic interference of the VVVF system, the study of global search optimization algorithms, and an improved genetic algorithm called emigration genetic algorithm for multi-objective optimal problems.In the design study of variable voltage and variable frequency systems, the thesis focused on the exploration of the main sources of the electromagnetic interference, the main path of the electromagnetic interference and the sensing unit in the VVVF system. Based on these works, approaches such as such as the earthing, masking and isolation for eliminating the electromagnetic interference are investigated. For self-contained, the national standards of the EMC requirements of VVVF systems are also included.An improved Evolutionary Algorithm is proposed in this paper for the multi-objective design problems of electromagnetic devices including VVVF ones. An emigration operator is designed to preserve the diversity and to smooth the Pareto front. The approaches such as the fitness sharing, the clustering method, and the elitism strategy are also improved and extended in this paper. Two multi-objective optimal problems are solved by using the proposed algorithm. The numerical results demonstrated that the proposed algorithm is a robust vector optimizer for engineering multi-objective design problems.