模糊线性回归法在负荷预测中的应用

线性回归是电力系统中期负荷预测的常用方法。由于受众多不确定性因素的影响，历史数据和相关变量未来取值常常是不准确的，致使该方法的预测结果误差较大。为了提高电力负荷预测的精度，作者提出了一种改进的模糊线性回归预测方法，即加权模糊线性回归预测法，它将模糊线性回归法预测模型的求解归结为一个线性规划问题，并对该模型进行改进，按照回归变量的重要程度确定目标函数中各项的权重，并按照各历史数据的重要程度确定贴近度标准。文中提出的改进模型即加权模糊线性回归模型是可调的，能够灵活计及预测中的一些定性模糊因素。实际算例表明，文中的改进措施提高了模糊线性回归法的预测精度。

APPLICATION OF FUZZY LINEAR REGRESSION TO LOAD FORECASTING

Linear regression analysis is a most common method for mid term load forecasting. Because of the influence of many uncertain factors, the collected historical data and the future correlative variable data are often imprecise. It leads to the notable errors in forecasting results. In order to improve the accuracy of load forecasting, an improved fuzzy linear regression model, which is also named as weighted fuzzy linear regression, is proposed in this paper. Linear programming is used to solve the fuzzy linear regression problem. The fuzzy linear regression model is improved in two aspects. The weight of each term in the object function is determined according to the importance of each regression variable. The relevant value is determined according to the importance of each historical datum. The weighted fuzzy linear regression model is adjustable. It can take some qualitative and fuzzy factors into account easily. The practical calculation shows that the forecasting precision is improved with the method mentioned above.