Abstract: The property of optimizing dynamic parameters of rigid multibody systems is analyzed in detail. An improved Niche genetic algorithm is introduced to resolve the problems of traditional optimization methods and simple genetic algorithms are used in this optimization. The Elitist model is utilized to ensure the stable convergence, and the Gaussian mutation operator is used to enhance the local search ability around every peak value. As an example, a five-degree-of-freedom automobile shows the method enhances the efficiency of con-vergence and global search greatly.

Key words: Genetic algorithm, Niche, Optimization, Rigid multibody system