阶梯电价背景下基于贝叶斯分析的用电需求模型  

Electricity Demand Research Under Multistep Electricity Price Based on the Bayesian Estimation

作　　者：琚艳芳 刘喜梅[1] 

机构地区：[1]华北电力大学经济与管理学院

出　　处：《华北电力大学学报（社会科学版）》2014年第3期37-43,共7页Journal of North China Electric Power University(Social Sciences)

摘　　要：阶梯电价不仅引导居民合理、节约用电，而且减少了用户之间的电价交叉补贴，但在实际应用中存在多种不确定因素，如居民用电需求变动以及各档电量的确定范围等。针对居民用电需求变动，本文提出了基于贝叶斯估计方法的阶梯电价用电需求模型。首先提出基于阶梯电价的需求函数；其次对阶梯电价用电需求函数进行贝叶斯分析，分别从统计模型、似然函数、后验分布以及加速收敛四个方面分析；最后，对1250个用户的用电数据进行估计，将影响因素带入模型得出各用户的用电需求，确定了贝叶斯估计对用电需求模型构建的适用性。Implement of multistep electricity price not only reasonably guide residents to save electricity, but also electricity reduce the electricity price cross complementary among users. It is necessary to study the electricity demand to determine quantity on each scale. The demand function based on multistep electricity price is first put forward; Bayesian analysis was carried out on the scale electricity price demand function, respectively from the statistical model, likelihood function, posterior distribution and convergence acceleration. Finally, 1250 electricity data are estimated.

关 键 词：贝叶斯估计 居民用电需求 GIBBS采样 马尔可夫链蒙特卡罗方法 

分 类 号：F407.61[经济管理—产业经济]