Benefits and challenges of design process that employs a multi-objective optimization approach for industrial applications

In this paper, total performance of two 20 kVA UPS realizations based on different design processes are compared in terms of losses and volumes. One of the design process employs a multi-objective optimization approach and Pareto analysis for selecting optimal design parameters. The other design process employs traditional industrial design approach. Employing the multi-objective optimization approach and Pareto analysis in the design process enables comparison and selection of suitable combinations of power semiconductor device for achieving minimum semiconductor losses. Thus, not only lower semiconductor losses but also lower heat sink volumes are achieved. In addition, it also enables selection of suitable magnetic material and design parameters, switching frequencies and relative current ripple, for reducing losses and volumes of inductors. As a result, the total volume of realized system based on the optimized design is 47% less than the other realization. On the other hand, costs are considered as one of the most important performance index in the industrial application. Challenges for including systematic costs analysis into the design process are also discussed.