A New Neural Network with Genetic Algorithm in Searching Nonlinear Function Extremum

Abstract In order to more accurate for nonlinear function extreme, this paper used improved particle swarm optimization neural network combining with genetic algorithm method solve the problem. view of is easy appear “premature” faults, introducing adaptive threshold, initializing particles if they were under constraint conditions, making jump out optimal value position in previous search. Through experiment, contrasts and traditional BP network, faster convergence has smallest prediction error. Finally, algorithm, calculating extreme by using above three kinds trained forecast as an individual output fitness value. The proves most close theoretical calculation. It shows that effective.