The limitations of the primary genetic algorithm applied to practice are discussed
including poor local search ability
premature convergence etc. A hybrid annealing genetic algorithm is developed by combination of simulated annealing and powell searching method with genetic algorithm. Case studies on typical test functions show that the hybrid annealing genetic algorithm is efficient and reliable
and completes perfectly the Fuzzy optimization design of worm drive.