Abstract:A novel heuristic method GA-PSO (genetic algorithm-particle swarm optimization) is proposed. Its genetic principles and application technologies are discussed on emphasis. The algorithm combines the advantages of GA's global search and diverse population and PSOs strong evolutionary direction and convergence. On the basis of PSO, it evolves by updating its velocity and location of the particles. Its new population is produced by a mother particle which is the elite particle of the past generation. The genetic operations of GA are implemented to the new population, such as elite reservation and roulette wheel selection, arithmetic crossover, gaussian mutation and micro rotation. The particles are coded by real numbers and the inertia weight is adjusted by the self adaptation mechanism. The examples show that the GA-PSO algorithm is correct and effective.

Optimization algorithm GA-PSO and its application technologies in engineering