基于灰色和马尔科夫系统理论的需求侧响应能力预测

近年来,随着分布式电源、电动汽车、智能用电器在用户侧接入比例越来越高,可以调动的需求侧资源将越来越丰富。探究多重变量对需求侧响应的影响,提出基于灰色关联度的多阶灰色预测模型和马尔科夫链模糊矩阵相结合的预测方法,对长期需求侧响应能力进行预测。采用的多阶灰色预测模型主要考虑到了负荷自变量的时间特性以及多重外在变量的相关性,而灰色关联度分析方法可以定量的研究事物之间的关联程度,当状态变量和时间变量是离散数据时,马尔科夫链对灰色理论中间累加过程中产生的误差有较好的调整效果,因此采用马尔科夫链模糊矩阵对预测误差进行修正,提高了长期需求侧响应能力的预测精度。结合我国上海市近10年的负荷数据,验证了所提预测方法的有效性。

Demand side response capacity prediction based on grey and Markov system theory

In recent years, with the increasing proportion of distributed power supply, electric vehicles and smart appliances accessing to the user side, the demand side resources that can be mobilized are more and more abundant. The impact of multiple variables on demand side response is explored, a multi-order grey prediction model based on grey correlation degree and a prediction method combining Markov chain fuzzy matrix are proposed to predict long-term demand side response ability. The multi-order grey forecasting model adopted mainly takes into account the time-space characteristics of load independent variables and the correlation of multiple external variables. When the state variables and time variables are discrete data, Markov chain has a good effect on adjusting the errors in the process of accumulating grey theory.Therefore, Markov chain fuzzy matrix is used to correct the prediction errors, and improves the prediction accuracy of long-term demand side response ability.Combining with the load data of Shanghai in recent 10 years, the validity of the forecasting method is verified.