基于信息熵的梯级水库联合优化调度增益分配法

Benefit distribution method based on information entropy for optimal joint operation of cascade reservoirs

针对Shapley值法在求解多利益主体梯级水库联合优化调度增益分配问题中的局限性,引入基于信息熵的增益分配法,对乌江流域七座梯级水库联合优化调度的增益分配问题进行求解,并与指标分配法和Shapley值法的分配结果进行比较。结果表明,基于信息熵的增益分配法比Shapley值法需要的信息资料较少,计算效率高,可以得到合理且稳定的分配结果。该分配方法公平合理,避免了单指标分配法考虑问题不够全面的局限性,调动了水库参与梯级联合运行的积极性,有利于提高水电资源利用效率。

Difficulties in benefit distribution for optimal joint operation of cascade reservoirs are analyzed.Considering the limitation of the Shapley value method,a new benefit distribution method based on information entropy is developed and applied to seven cascade reservoirs in the Wu River basin.Compared to the Shapley value and single-index allocation methods,this method needs a smaller set of data and is more efficient in calculations;its results are more equitable and reasonable,motivating managers to participate joint reservoir operation and improving hydropower resources utilization efficiency.