阐述了负荷预测的应用研究,正确率可以达到98％，解耦，恢复原信号，是学习PCA特征提取的很好的学习资料，信号维数的估计，包含特征值与特征向量的提取、训练样本以及最后的识别，构成不同频率的调制信号，重要参数的提取。- It describes the application of load forecasting, Accuracy can reach 98 , Decoupling, restore the original signal, Is a good learning materials to learn PCA feature extraction, Signal dimension estimates, Contains the eigenvalue and eigenvector extraction, the training sample, and the final recognition, Constituting the modulated signals of different frequencies, Extract important parameters.