A Study On Improved Genetic Algorithm

Genetic algorithm is a probabilistic search algorithm which simulates the evolution method in nature. It can solve nonlinear problems which are difficult to solve by traditional search methods. But the classical genetic algorithm has some deficiency. For example, the algorithm often obtains a local optimal solution, converge slowly and so on. So a staged fitness function, a crossover operation that based on the mechanism of competition and imitation particle swarm operation are proposed in this paper. The new algorithm obviously improves the convergence speed and the global convergence probability. Finally, the improved genetic algorithm is applied to load allocation power generating units model, and then proves its practicability and validity.