基于遗传算法和BP算法的混合算法

结合神经网络的优化问题,提出一种新的混合算法.该算法是在遗传操作中嵌入BP算子,有效地结合了遗传算法全局寻优与BP算法快速收敛的特点,同时采用二进制编码和实数编码将神经网络的结构与权值混合编码到串中,实现了结构与权值的同步优化.仿真结果表明,新算法既能够快速地收敛到全局最优解,又能够在简化网络结构复杂度的同时得到网络权值的最佳逼近.

A Hybrid Algorithm Based on Genetic Algorithm and BP Algorithm

By embedding BP operator into genetic algorithm,a hybrid algorithm is put forward.This algorithm assimilates advantages both of the globe convergence of genetic algorithm and of the fast convergence of BP algorithm,and it encodes the construction and the weights hybridly with real code and binary code.The test results show that the hybrid algorithm can not only converge to the global optimal solution quickly but also can optimize the construction and the weights of neural network synchronously.