变尺度混沌优化参数的LSSVM短期负荷预测

Short-term Load Forecasting Based on LSSVM with Mutative Scale Chaos Optimization Parameters

为解决短期电力负荷预测中LSSVM模型的参数难以确定的问题,利用变尺度混沌算法优化LSSVM模型的惩罚因子和核函数参数,构建了MSC-LSSVM模型,并将其应用于湖南省隆回县地区电网各小时点的数据分析和预测中。结果表明,MSC-LSSVM模型避免了人为选择参数的盲目性,预测精度较高。

In order to overcome drawbacks of parameters selection of LSSVM model for short-term load forecasting, the mutative scale chaos （MSC） algorithm is developed to optimize two parameters of LSSVM model such as penalty factor and kernel function. And then the MSC-LSSVM model is established to predict each hour load of Longhui City in Hu nan Province. The results show that the proposed model avoids the blindness of man-made choice of parameters and has high prediction precision.