Multi-objective optimization of capacitor bank considering the parasitic parameters of capacitors

For the applications where a single capacitor is incapable to meet the needs, different types, behaviors and specifications of capacitors can be packaged as a capacitor bank to fulfill the requirements of the systems. Even though some commercial capacitor bank products have already existed in the market, most of them are the experienced based design which uses same capacitors connected in parallel or series to fulfill the capacitance and voltage rating requirements, so that the volume, cost, reliability and other performance factors are not optimized. To the best knowledge, no quantitative design for capacitor banks is available by considering all the design phases. This paper proposes a design method for capacitor banks with multiple design constraints. Considering the multi-physics aspects, the performance factors (e.g., lifetime, efficiency, cost and volume) of the capacitors are modeled. By optimizing the capacitor bank configurations (e.g. hybrid capacitance ratio of types, and the capacitance and numbers for individual capacitor), the solutions which fulfill the requirements of applications are identified. To verify the accuracy of the proposed method, a capacitor bank for the DC link of a 5.5 kW PV inverter is optimized and demonstrated.