基于IKH-LSSVM的架空导线脱冰跳跃高度估算  

Estimation for Ice-Shedding Jump Height of Overhead Conductor Based on IKH-LSSVM

作　　者：马晓宇 郑权 陈海旭 MA Xiaoyu;ZHENG Quan;CHEN Haixu(Fuzhou Power Supply Company State Grid Fujian Electric Power Co.,Ltd.,Fuzhou 350004 China;Huangshi Power Supply Company State Grid Hubei Electric Power Co.,Ltd.,Huangshi 435100 China)

机构地区：[1]国网福建省电力有限公司福州供电公司,福建福州350004 [2]国网湖北省电力有限公司黄石供电公司,湖北黄石435100

出　　处：《红水河》2025年第1期111-115,共5页

摘　　要：为了准确获取架空导线脱冰跳跃的最大高度,笔者以导线分裂数量、覆冰厚度、档距、脱冰率等4个参数作为输入量,以导线脱冰跳跃最大高度作为输出量,构建基于改进磷虾群算法(improved krill herd,IKH)优化最小二乘支持向量机(least squares support vector machine,LSSVM)的架空导线脱冰跳跃高度估算模型。通过利用混沌映射初始化和柯西变异策略对磷虾群算法进行改进,采用IKH算法获得LSSVM参数的最优值,从而搭建IKH-LSSVM模型。算例分析结果表明,采用IKH-LSSVM模型估算的结果具有较高的决定系数(0.971)、较低的均方根误差(0.928 m)和平均相对误差(4.623%),各项误差指标均为最优,验证IKH-LSSVM模型在架空导线脱冰跳跃高度估算方面的有效性。In order to accurately obtain the maximum height of ice-shedding jump of overhead conductors this paper takes the four parameters of conductor splitting number ice thickness span and ice-shedding rate as input variables and the maximum height of ice-shedding jump of overhead conductors as output variables and constructs an estimation model for ice-shedding jump height of overhead conductors based on least squares support vector machine LSSVM optimized by improved krill herd IKH algorithm.The krill herd algorithm is improved by using chaotic mapping initialization and Cauchy mutation strategy and the optimal value of LSSVM parameters is obtained by IKH algorithm and the IKH-LSSVM model is built.The results show that the IKH-LSSVM model has a high determination coefficient 0.971 a low root mean square error 0.928 m and an average relative error 4.623%and all the error indexes are optimal.The effectiveness of the IKH-LSSVM model in estimating the ice-shedding jump height of overhead conductors is verified.

关 键 词：脱冰跳跃高度 架空导线 改进磷虾群算法 最小二乘支持向量机 

分 类 号：TM726.3[电气工程—电力系统及自动化]