Genetic algorithms are a technique for modeling the evolution of organisms and repeating intersections and sudden mutations to select suitable solution sets. The method is characterized by multi-point searches. Toshiba has applied genetic algorithms to multiobjective optimization problems with respect to industrial motors and other inverter power units. These power units have five objective variables, such as heat resistance and cooling fan current speed. As a result, Toshiba was able to ascertain trade-off relationships between objective variables and Pareto solution sets, making possible the calculation of design proposals using weighting by means of paired comparison of multiple objective variables. This method is expected to result in increased design quality and efficiency in comparison with conventional series design, a method in which objective variables are decided sequentially.
To Top