抄録
Numerous combinations of the design parameters usually exist when a machine is designed by simulation. It is very difficult to make a search for the optimum combination when a designer combines the parameters by himself because the search is very much likely to be biased and manipulated.
We have developed a simulation program predicting the internal state of a vane pump. In this study, we have attempted to design the vane pump with small pressure fluctuation by using a simulation program that combines the genetic algorithm that is one of the optimum-value search algorithms with the program. The design parameters to be varied by the genetic algorithm were the shape of the cam ring in a vane pump.
In this paper, effectiveness of the genetic algorithm for a pump design is reported to have been verified by experiments using the cam ring based on the simulation.
