基于实数编码遗传算法的多层神经网络BP算法

基于实数编码遗传算法的多层神经网络BP算法 A MULTI LAYER NEURAL NETWORK BACKPROPAGATION ALGORITHM BASED ON REAL CODED GENETIC ALGORITHM 王保中 Wang Baozhong 1 first-author 康立山 Kang Lishan 1 何巍 He Wei 1 武汉大学软件工程国家重点实验室 武汉大学软件工程国家重点实验室 State Key Laboratory Software Engineering, Wuhan University State Key Laboratory Software Engineering, Wuhan University 提出用实数编码的遗传算法来优化多层神经网络的权值，并且将遗传算法与ＢＰ算法结合，能有效地避免ＢＰ算法陷入局部极小和遗传算法过早收敛，实验结果令人满意． In this paper genetic algorithm coding in real number is engaged to optimize the weights of multi layer neural network. Furthermore, by combining this real coded GA with BP algorithm, the entrapment in local optical optimum of BP and the premature of GA can be prevented efficiently. Some satisfactory results are obtained. 神经网络 遗传算法 BP算法 XOR问题 编码解码问题 neural network genetic algorithm backpropagation algorithm XOR problem encode/decode problem TP18;TP314 国家自然科学基金,教育部博士点基金 1998-03-01 2021-04-01 3