Transmission Line Ice Cover Thickness Prediction Model Based on IDBO-LSSVM
Aiming at the problem of low accuracy of ice cover thickness prediction due to the influence of multiple meteorological factors on transmission line ice cover,a transmission line ice cover thickness prediction model based on improved dung beetle optimizer(IDBO)-least square support vector machine(LSSVM)was proposed.Firstly,the Pearson correlation coefficient(PCC)was used to calculate the correlation between the ice thickness of transmission lines and different meteorological factors.High correlation meteorological factors were selected to determine the input variables.Secondly,the dung beetle optimizer(DBO)was improved by introducing Halton sequences,Levy flight strategies,and T-distribution perturbations.Finally,IDBO was used to optimize the parameters of LSSVM,including the adjustment factor and kernel function width,further improving the prediction accuracy of the model.When the prediction results of IDBO-LSSVM were compared with other seven prediction models using the historical monitoring data of a transmission line in a certain region as a sample,the average absolute errors were reduced by about 27%,36%,25%,23%,24%,44%and 39%,respectively.The results indicated that the transmission line ice cover thickness prediction model based on IDBO-LSSVM could effectively improve the prediction accuracy.
国家科技期刊平台
NSTL
万方数据
