摘要
利用萤火虫算法优化BP神经网络权值和阈值基础上,建立水电站厂房振动响应预测模型。针对萤火虫算法存在的收敛速度慢、易陷入局部最优等问题,引入动态随机局部搜索机制加快收敛速度,对当前最优解进行变异操作避免陷入局部最优,提出动态步长更新措施提高计算精度,改进最优解振荡问题。仿真实例表明,基于改进萤火虫算法优化的BP网络模型预测精度和收敛速度等性能得到明显改善,可用于水电站厂房结构振动响应预测。
A hydro-power house vibration prediction model was built based on the BP neural network optimized with the improved firefly algorithm( IFA-BPNN). Aiming at some disadvantages of FA including slow convergence,and easy to fall in local optimal values,a dynamic random local searching algorithm was introduced to speed up the convergent velocity,and do some mutation operations to avoid the optimization to fall into local optimal values. A dynamic step length updating measure was proposed to improve the accuracy of the optimization,and avoid the optimal solutions' oscillation problem. Simulation examples showed that the prediction accuracy and convergent speed of the IFA-BPNN method are obviously improved,it can be used to predict vibration responses of a hydro-power house.
出处
《振动与冲击》
EI
CSCD
北大核心
2017年第24期64-69,共6页
Journal of Vibration and Shock
基金
国家自然科学基金(51479165)
关键词
水电站厂房
振动
萤火虫算法
神经网络
hydro-power house
vibration
improved firefly algorithm(IFA)
back-propagation neural network(BPNN)
