自适应粒子群支持向量机风速组合预测模型

COMBINATION MODEL FOR FORECASTING WIND SPEED BASED ON ADAPTIVE PSO-SVM

提出一种优化的支持向量机风速组合预测模型,首先通过模糊层次分析法对参与组合的单项预测模型进行遴选,在当前风速样本集下自适应决策预测效果较优的单项预测模型的输出值作为支持向量机的输入,将实际风电场风速值作为支持向量机的输出,并采用粒子群算法优化支持向量机组合模型的参数。基于实际运营的风电场数据进行仿真分析,自适应遴选出BP神经网络、RBF神经网络、小波神经网络和遗传算法优化BP神经网络这4种单项预测模型参与支持向量机组合,结果表明所提方法的预测精度不仅高于单项模型,且高于线性组合预测模型和神经网络组合预测模型。

A combined model for forecasting wind speed based on optimized support vector machine （SVM） was proposed by using data from wind farms. The single forecasting models were chosen by fuzzy hierarchy process （FAHP） at first, the output single forecasting model which has the better adaptively deciding performance in present wind speed samples was selected as the input of SVM, and the actual wind speed of wind farm was taken as the output of SVM. The particle swarm optimization （PSO） was used to optimize the parameters of combination model at the same time. The data of actual wind farm were simulated and analyzed. The four single forecasting model, such as back propagation （BP） neural network, radial basis function （RBF） neural network, genetic algorithm neural network and wavelet neural network were chosen adaptively to participate in SVM combination. The results showed that the accuracy of the proposed method is not only higher than that of any single model, but also than the traditional linear combined forecasting model and the neural network combined forecastin~ model.