| 基于实数编码的改进遗传神经网络研究及应用 |
| |
| 引用本文： | 徐永春.基于实数编码的改进遗传神经网络研究及应用[J].广东广播电视大学学报,2008,17(5):101-103. |
| |
| 作者姓名： | 徐永春 |
| |
| 作者单位： | 广东广播电视大学,广东广州,510091 |
| |
| 摘 要： | 本文介绍了BP网络收敛慢与全局收敛的不足等特点，提出以实数的编码机制及在迭代过程中改进的自适应交叉、变异算子更新及种群更新等改进遗传策略，以此提高BP网络学习效率和BP模型预测的准确率，并以甘蔗产量的实例数据进行模拟仿真分析验证。
|
| 关 键 词： | 遗传算法 改进BP网络 甘蔗 预测 |
Application Research on the Neural Networks with Improved Genetic Algorithms Based On Real Coding |
| |
| XU Yong-chun.Application Research on the Neural Networks with Improved Genetic Algorithms Based On Real Coding[J].Journal of Guangdong Radio & Television University,2008,17(5):101-103. |
| |
| Authors: | XU Yong-chun |
| |
| Institution: | XU Yong - chun ( Guangdong Radio ＆ TV University, Guangzhou, Guangdong, China, 510091 ) |
| |
| Abstract: | This paper,firstly,presents defects on slow convergence of the network and the total global convergence of back -propasation(BP)neural network.Then we propose a method that improves the simple genetic algorithms(SGA)with the in- dividual choices update based on improved cross-adaptive,mutation operator based on real coding.The method is used to opti- mize BP net and raise the accuracy of the BP forecast.Finally,a mathematical study on the case of sugar cane forecast is veri- fied. |
| |
| Keywords: | genetic algorithm Bp network sugar cane forecast |
| 本文献已被 CNKI 维普 万方数据 等数据库收录！ |
|
