Mid-long term Load Forecasting Model Based on Error Compensation

Power grid planning is the foundation of power grid construction, and mid-long term load forecasting is a key technology in power grid planning, which is of great significance to power grid construction. Firstly, according to the characteristics of ‘S’ growth of power load on a long-term scale, logistic regression is used to construct a basic model of mid-long term load. Secondly, to reduce the error between the basic model and the actual load, the uncertainty of the error source is comprehensively considered. A probability model of error compensation was constructed using Gaussian process regression. In addition, because of the large amount of input of the error compensation model, there is a problem called “dimensionality disaster”. To solve this problem, the principal component analysis method is used to reduce the dimensionality of the input of the error compensation model. Finally, the effectiveness and accuracy of the proposed method are verified by numerical examples.