中长期电力消费需求与市场发展密切相关，目前的电力需求预测方法对市场的关注度远远不够。文中从市场着眼，建立基于宏观经济、产业结构和终端能源结构等市场因素的电力消费需求预测模型。将经济增长构成要素分为投资、消费、出口3个方面，运用向量自回归（VAR）模型表征电力消费的时间序列与经济指标时间序列的关系。采用灰色理论常见模型GM（1，1）的精确形式DGM（1，1）模型对分产业用电量进行预测。根据马尔可夫预测无后效性的特点，建立最优化模型求取预测年份的终端能源概率。采用残差均方根来反映预测模型对历史数据的拟合程度，确定各预测方法权重，进行组合预测。文中以某省电力消费数据为实例，预测该省中长期电力需求，并通过校验证明了方法的准确性。%The medium and long term electricity consump-tion has a close relationship with the market development. The existing forecasting methods for electricity requirement,howe-ver, pay little attention to the market. With focus on the market, this paper establishes forecasting models based on macro economy,industry structure and energy end-use struc-ture. Economy development is characterized by investment, consumption and export. The Vector Autoregression(VAR) model is employed to certify whether there is a long term co-integration between electricity requirement and economy inde-xes. This paper applies Grey theory to predict each industry separately. The DGM (1,1)model,which is the precise form, is used instead of the common model GM (1,1). The Markov Prediction is suitable for the energy end-use structure. The paper sets up the optimization model and obtains energy probability. This article applies residual root mean square method to confirm the weight of each model. According to the electricity require-ment of one province,codes are written to predict electricity requirement during the period of 2012 to 2020. And the real data are compared to certify the accuracy of the whole model.
展开▼
