摘要
将误差反向传播算法(BP算法)以一个算子的形式融入到遗传算法中,以提高遗传算法的优化性能。其基本思路是:在遗传算法收敛速度放慢时启用BP算子,把新一代群体作为BP算子的初始值再用BP算法训练网络,这样交替运行BP算法和遗传算法,直到达到问题要求的精度。通过对4例实验函数的优化,证明了混合遗传算法具有良好的收敛性和稳定性。实验对插入BP算子的遗传算法和传统遗传算法的优化结果进行了比较分析,结果表明BP算子的插入对遗传算法的优化性能、收敛速度和收敛精度方面都有了很大的改进。
In order to optimize the performance, mixed genetic algorithm, its main thought is to combine genetic algorithm with back propagation algorithm (BP), which realizes that BP algorithm is inserted into the genetic algorithm in the form of an operator. The result proves that mixed algorithm has excellence convergence property and robustness by optimizing four test functions. The experiment has compared the mixed algorithm with traditional genetic algorithm, and its result indicates the insertion of the BP operator has great influence on optimization performance, convergence speed and precision of the genetic algorithm.
出处
《计算机工程与设计》
CSCD
北大核心
2007年第3期651-652,共2页
Computer Engineering and Design
基金
山西省教育厅高校科技研究开发基金项目(200358)
关键词
遗传算法
杂交率
变异率
BP算法
人工神经网络
genetic algorithm
crossover probability
mutation probability
BP algorithm
artificial neural network