An improved chaos genetic algorithm and its application in parameter optimization for robot control system

To avoid premature convergence and trapping into local minimum, a chaos genetic algorithm based on population high-efficiency mutation (CGAPM) is presented. According to achievements in society and biology, small world network, characterized in clustering and small-world effect, is introduced into GA to change the mutation from randomness to directionality. The chaotic variables are considered to produced the initial population with logistic mapping and chaos disturbance is performed after small-world mutation, thus the searching efficiency and accuracy are improved. The simulation results show that the proposed algorithm is stabile and effective. And the optimization results in the trajectory tracking of wheeled mobile robot verify that the developed method can obtain more satisfied parameters for control system.