摘要
针对传统BP神经网络搭建的电梯群控算法中出现的易于陷入局部极值、收敛速度慢、预测值与实际值偏差较大等问题,本文通过分析研究,在使用BP神经网络,拟合某台电梯对某一楼层呼梯信号响应满意度函数的基础上,应用Adam算法优化神经网络的权值和阈值,使用Dropout缓解过拟合现象,减小误差,提高网络预测精度。仿真结果表明,与传统的优化算法相比,此算法收敛速度更快,模型预测准确率更高,减少了候梯时间,提高了电梯运载效率。
For the elevator group control algorithm built by the traditional BP neural network,it is easy to fall into the local minimum,the convergence speed is slow,and the deviation between the predicted value and the actual value is large.Through analysis and research,on the basis of using the BP neural network to fit the satisfaction function expression of a certain elevator’s response to a certain floor call signal,the Adam algorithm is used to optimize the neural network weights and thresholds,and Dropout is used to alleviate The phenomenon of fitting reduces errors.The simulation results show that compared with the traditional optimization algorithm,the convergence speed is faster,the model prediction accuracy is higher,the waiting time and the long waiting time are reduced,and the elevator carrying efficiency is improved.The results show that the average relative error is less than 2%.
作者
雷剑
LEI Jian(School of Electrical Engineering,University of South China,Huanan Hengyang 421000,China)
出处
《智能计算机与应用》
2020年第11期101-105,共5页
Intelligent Computer and Applications
关键词
Adam算法
BP神经网络
电梯群控算法
多目标优化
Adam algorithm
BP neural network
elevator group control algorithm
Multi-objective optimization
