考虑多时间尺度的梯级水电站组合交易风险自适应感知方法

Adaptive Perception Method for Combined Transaction Risk of Cascaded Hydropower Stations Considering Multi-time Scales

为了控制交易风险发生概率,规避交易风险,助力收益最大化,提出考虑多时间尺度的梯级水电站组合交易风险自适应感知方法。综合考虑多时间尺度,将梯级水电站交易方式划分为上下两层,各层利用多时间尺度的马尔科夫决策描述梯级水电站的交易过程,搭建梯级水电站组合交易序贯模型,并采用人工蜂群算法求解获取组合交易最佳策略,搭建基于条件在险价值的组合交易风险度量模型,给定置信水平,求解组合交易最佳策略的风险值,实现梯级水电站组合交易风险自适应感知。实验结果表明,所提方法能够选取最佳组合交易决策,决策后的水电出力可满足实际负荷需求,拥有理想的决策性能;能够感知各时间段内组合交易的风险值,且应用该方法后可有效控制交易风险发生概率,应用效果极佳。

In order to control the probability of transaction risk,avoid transaction risk and maximize income,an adaptive perception method of combined transaction risk of cascade hydropower stations considering multi-time scales is proposed.Considering multi-time scales comprehensively,the transaction mode of cascade hydropower stations is divided into upper and lower layers.Each layer uses Markov decision of multi-time scales to describe the transaction process of cascade hydropower stations,and a sequential model of cascade hydropower station portfolio transaction is built.The artificial bee colony algorithm is used to solve the sequential model and obtain the best strategy of portfolio transaction.Then a portfolio transaction risk measurement model based on Conditional Value at Risk(CVaR)is built to solve the risk value of the best strategy of portfolio transaction under given confidence level.Finally,the adaptive perception of the risk of portfolio transaction of cascade hydropower stations is realized.The case study results show that,(a)the proposed method can select the best combination transaction decision,and the hydropower output after decision can meet the actual load demand and has ideal decision performance;and(b)it can perceive the risk value of portfolio transactions in each time period,and the application of this method can effectively control the probability of transaction risk and has excellent application effect.