| 限幅的变步长BP算法 |
| |
| 引用本文： | 杨彦利,汤武初,赵亮,伉大俪.限幅的变步长BP算法[J].机械工程师,2005(11):105-106. |
| |
| 作者姓名： | 杨彦利 汤武初 赵亮 伉大俪 |
| |
| 作者单位： | 1. 大连交通大学,辽宁,大连,116028
2. 大连圣力来监测技术有限公司,辽宁,大连,116033 |
| |
| 摘 要： | 针对BP算法学习速度慢和易陷入局部最优等缺陷,提出了一种新的变步长算法--限幅的变步长算法CVLBP.这种算法根据误差来调整下次的学习步长,同时把步长限制在一定范围内,这样在计算量增加很少的情况下,使学习时间大为缩短的同时,也避免了过多的振荡,防止了系统的发散.仿真结果表明,该算法具有跳出局部最优的能力,同时对初始权值和闻值具有一定的鲁棒性.
|
| 关 键 词： | BP神经网络 自适应 变步长学习 VLBP算法 LMS算法 |
| 文章编号： | 1002-2333(2005)11-0105-02 |
| 收稿时间： | 2005-06-19 |
| 修稿时间： | 2005年6月19日 |
Control Variable Learning Rate BP Algorithm |
| |
| YANG Yan-li,TANG Wu-chu,ZHAO Liang,KANG Da-li.Control Variable Learning Rate BP Algorithm[J].Mechanical Engineer,2005(11):105-106. |
| |
| Authors: | YANG Yan-li TANG Wu-chu ZHAO Liang KANG Da-li |
| |
| Affiliation: | 1.Dalian Jiaotong University,Dalian116028,China;2.Dalian Shenglilai Monitoring Technology Limited Corporation,Calian 116033,China |
| |
| Abstract: | A novel variable step size(control variable learning rate backpropagation,CVLBP)algorithm is proposed in view of the faultiness of BP algorithm.The step size which was limited in a certain value was determined by the learning error.Thus the learning time is cut down greatly and avoid some oscillation in the learning process,even avoid divergence,but the calculation is added little.The simulation result shows that this algorithm can step out the local optimal and has certain powerful intelligence to initial weight and threshold value. |
| |
| Keywords: | BP neural network adaptive variable step size learn VLBP(variable learning rate backpropagation) LMS(least mean square) |
| 本文献已被 CNKI 维普 万方数据 等数据库收录！ |
|
