Multi-Object Optimal Design of Analog Filter Based on Improved Genetic Algorithm

A multi-object optimization method for analog filter design based on genetic algorithm (GA) is proposed. The complete objective function is a weighted sum of deviations between the properties of designed filter and these of desired filter, including the magnitude, phase responses and step response, etc. The optimization is achieved by GA to minimize the complete objective. For overcoming the disadvantages of simple GA, an improvement is given. A four order filter designed by proposed method testified the validity of the method.