摘要
针对当前输电线路温度预测算法精度低的问题,提出一种基于改进的天牛须搜索(Improved Beetle Antennae Search,IBAS)优化反向传播(Back Propagation,BP)神经网络的输电线路温度预测算法。首先,为平衡天牛须搜索(Beetle Antennae Search,BAS)的局部搜索能力与全局搜索能力,在BAS算法的步长寻优过程中引入自适应收敛因子;其次,为使BAS算法避免陷入局部最优解,引入t分布变异算子来增加天牛种群;最后,采用IBAS算法优化BP神经网络的阈值与权值,建立IBAS-BP输电线路温度预测模型。仿真结果表明,IBAS-BP算法的R2为0.98,接近1,预测精度高。
Aiming at the problem of low accuracy of thecurrent transmission line temperature prediction algorithm,a transmission line temperature prediction algorithm based on Improved Beetle Antennae Search(IBAS)optimized Back Propagation(BP)neural network is proposed.Firstly,in order to balance the local search ability and global search ability of Beetle Antennae Search(BAS),an adaptive convergence factor is introduced in the step size optimization process of BAS algorithm.Secondly,in order to prevent BAS algorithm from falling into local optimal solution,t-distribution mutation operator is introduced to increase the population of longicorn beetles.Finally,IBAS algorithm is used to optimize the threshold and weight of BP neural network,and establish IBAS-BP transmission line temperature prediction model.The simulation results show that the R2 of IBAS-BP algorithm is 0.98,close to 1,and the prediction accuracy is high.
作者
谭晓宇
TAN Xiaoyu(School of Electrical and Information Engineering,Anhui University of Technology,Huainan Anhui 232001,China)
出处
《信息与电脑》
2022年第13期37-39,43,共4页
Information & Computer
