Multiobjective Optimal Dispatching of Cascade Hydropower Stations Using Improved NSGA-II

何向阳,周建中,张勇传,覃晖
2011-01-01
Abstract:NSGA-Ⅱis a typical multiobjective evolutionary algorithm.Based on NSGA-Ⅱseveral improvements is presented: ① a fixed size external repository and a local search operator is introduced to obtain better spread of solutions and better convergence;② to facilitate easy implementation for the reservoir operation,a simple but effective decision-making approach was presented;③ to improve efficiency of NSGA-Ⅱ,a quick sort factor is adopted to sort the population according to each objective function value in ascending order of magnitude.Finally,the improved NSGA-Ⅱ is applied to a case study of multiobjective optimal dispatching of operating the Three Gorges cascade hydropower stations;and the results show that the imporved NSGA-Ⅱcan handle complex constraints and obtain better spread of non-dominated solutions,and make it easy for the reservoir operation to choose the final solution by calculating the weight of the respective objective of every non-dominated solutions.
What problem does this paper attempt to address?