电力负荷预测的最优分段校正模型研究及应用

中长期电力负荷同时具有增长性和波动转折性的二重趋势，这使得电力负荷的变化呈现出复杂的非线性特征。传统的中长期负荷预测模型仅考虑一种变化趋势，从而预测效果较差，不能符合实际需要。为了提高预测精度，更精确地反映负荷变化规律，通过改变GM(1，1)模型中微分方程的初始条件，且对历史负荷进行最优分段，在此基础上建立了电力负荷预测的最优分段校正模型，使模型具有二重趋势性特点。通过实例验证，较大提高了预测精度。

Study and application of optimum and segmental correct model for power load forecasting

Medium and long-term power load has dual trend, that is increasing and fluctuating turning. This makes the change of power load take on the characters of complexity and non-linear. A certain kind of variation tendency is only considered in traditional medium and long-term load-predicting model, which makes the predicting results relatively bad, and can't meet reality's needs. In order to improve forecasting precision and reflect the law of load changing accurately, the initial condition of differential equation in GM (1,1) was changed and the historical load data was segmented optimally. Based on these, the optimum and segmental correct model for power load forecasting that has dual trend characters was set up. Verified by the instance, the model can improve prediction precision prominently.