基于PSO与多变量支持向量机的通信基站能耗预测

Energy Consumption Prediction Model of Communication Base Station Based on PSO and Multivariate Support Vector Machine

为准确预测通信基站的能耗,提出了一种基于微粒群算法与多变量支持向量机(PSO-MSVM)的通信基站能耗预测模型。该模型基于温度和日期两个重要因素,运用MSVM建立通信基站能耗预测模型;采用PSO对模型核函数中的参数c和ε进行优化整定,以提高算法效率。湖南某地区通信基站的测试结果表明:与最小二乘法相比,PSO-MSVM的预测结果更接近基站实际能耗,预测精度达93.8%以上,表现出良好的工程价值。

In order to predict the energy consumption of communication base station accurately, a prediction model of energy consumption for communication base station based on particle swarm optimization and multi variable support vector machines（PSO-MSVM） is proposed. The model is based on two important factors of temperature and date. The prediction model of energy consumption for communication base station is established by MSVM. PSO is used to optimize the parameters c and ε in kernel function, so as to improve the efficiency of algorithm. The test results of the communication base station in Hunan area show that the PSO-MSVM is closer to the actual energy consumption of the base station compared with the least squares method, and the prediction accuracy is more than 93.8%, which shows good engineering value.