抽象的な

Genetic algorithms based on local variable weight synthesizing and its application to internal model control

Liu Jianchang, Chen Nan, Yu Xia

In this paper, a newobjective function of genetic algorithms based on local variable weight synthesizing is proposed to improve the imperfect selection of performance indicator and unclear weight distribution in objective function of controller parameters optimization. Using both error integral indicators and eigenvalues of the systemcalculated by local variableweight synthesizing as a parameters optimization objective function to achieve the purpose that eigenvalues of the system are all in a reasonable range and error integral values are smaller as well. Compared with traditional objective function, the modified objective function is more comprehensive, flexible and open.At last, applying it to the parameters optimization of internal model control and the simulation results have shown its effectiveness and superiority.