论文部分内容阅读
分别介绍了采用BP神经网络模型和线性回归模型进行电价预测的方法和结果。方法的突出特点是在预测模型中引入了一个衡量市场力的新指标———发电容量必须运行率 (MRR) ,从而充分考虑了市场力对电价的影响 ,提高了电价预测的精度 ,特别是增强了短期预测模型对最高限价的预测能力。文中对MRR指标进行了简单的介绍 ,并针对电价预测的不同特点 ,对预测模型和预测变量的选择进行了探讨 ,提出了自己的观点。基于浙江电力市场实际运营数据的初步预测结果表明 ,所建预测模型是适用的 ,选择的预测输入变量是恰当的 ,电价预测精度能够满足电力市场实际运营的需要。
The methods and results of price forecasting using BP neural network model and linear regression model are respectively introduced. The salient feature of the method is the introduction of a new measure of market power called MRR in the forecasting model, which fully considers the effect of market power on electricity price and improves the accuracy of electricity price forecasting, especially Enhance the ability of short-term forecasting models to predict the price ceiling. In this paper, the MRR index is introduced briefly. In view of the different characteristics of the electricity price forecasting, the paper discusses the choice of the forecasting model and the forecasting variables and puts forward my own views. The preliminary forecast results based on the actual operation data of Zhejiang Electricity Market show that the forecasting model is suitable and the forecast input variables selected are appropriate, and the price forecasting precision can meet the actual demand of electricity market.