综合最优灰色支持向量机模型在季节型电力负荷预测中的应用

Integrated Optimum Gray Support Vector Machine Model for the Seasonal Power Load Forecasting

季节型电力负荷同时具有增长性和波动性的二重趋势,使得负荷的变化呈现出复杂的非线性组合特征。对此,提出了一种综合最优灰色支持向量机预测模型,研究了同时考虑2种非线性趋势的复杂季节型负荷预测问题,说明了此优化模型分别优于2种单一负荷预测模型。在此基础上,对一般粒子群算法引入粒子速度自适应可调机制,并利用改进粒子群算法优化组合预测模型中的权值。对电力负荷预测应用实例的计算结果表明,该模型较大提高了季节型负荷预测的精度,具有较好的性能。

Seasonal power load forecasting possesses the dual properties of growth and fluctuation simultaneously,characterizing the load variations with complex non-linear combination.Therefore,this paper puts forward the integrated optimum gray support vector machine model to study the problem of complex seasonal load forecasting with double nonlinear trends.The optimum model turns out superior to the two single load forecasting models.Furthermore,improved particle swarm optimization algorithm,proposed by introducing the adjustable mechanism of the adaptive particle velocity,can optimize the weight of combination forecasting model with effective dynamic adaptability.The case calculation results show that the proposed method can enhance the accuracy of the seasonal load forecasting greatly,exhibiting superior performance.