摘要： 针对传统预测模型精度不高的问题，提出基于小波核支持向量机的复合预测模型。采用小波分析提取燃气负荷相关的特征值，通过粒子群优化算法确定小波核支持向量机的参数，利用支持向量机(SVM)解决非线性回归和时间序列问题。实验结果证明，该预测模型的预测精度比BP神经网络和传统高斯核SVM高。

关键词: 支持向量机, 核函数, 粒子群优化, 燃气负荷, 小波, 预测模型

Abstract: Facing an uncertain, nonlinear, dynamic and complicated system, gas load forecasting generally can not get a sufficient accuracy result when using traditional forecast model. This paper proposes a wavelet v-Support Vector Machine(SVM) compound model, wavelet analysis extracting the feature of gas load, and PSO determining the parameter of Wv-SVM model, solving nonlinear regression and time series problems. Experimental results show that the proposed model outperforms the back propagation neural network and traditional Gauss SVM model.

Key words: Support Vector Machine(SVM), kernel function, Particle Swarm Optimization(PSO), gas load, wavelet, forecast model

中图分类号: 

• TP18