Tapped delay line filter based on real genetic algorithm and its application in adaptive inverse control

A design method of tapped delay line filter (transversal filter) based on real genetic algorithm is proposed. By improving continuous space genetic algorithms, it is applied in the adaptive inverse control of dynamic system. Plant and inverse controller are expressed by a tapped delay line filter respectively. In order to make the inverse adaptive process of plant model change with the plant model timely and without lag, the model-reference inverse model is trained offline. Finally, the effectiveness of this method is demonstrated by simulation result of a stabilized non minimum phase plant model and model-reference inverse.