Multi-Objective Optimization Design of Electric Vehicle Converters Based on $\varepsilon-\gamma-\sigma$

Abstract

Converter is a key component of electric vehicles (EVs) driver system. With the increasing fierce competition in the market, maximizing the efficiency and reliability, and reducing the costs of EV converters are more and more critical for EVs manufactures. To solve this problem, a novel customized design method of EV converters to achieve trade-offs between power losses, lifetime and cost is proposed. In this paper, first, a multidimensional function with the objectives of specific energy consumption ε, mean time to first failure γ and specific cost σ is established under the EV driving cycles. Then the multiobjective optimization problem with variable of IGBT switching frequency is analyzed to seek the Pareto optimal solutions. On this basis, the best trade-off between loss, reliability and cost of the EV converters, which provides a new routine for the design of the converters, is obtained.