The Study Of Decomposition-based Multiobjective Evolutionary Algorithm And Its Application In Small Hydropower Group Scheduling

In our country,small hydropower resources are abundant and widely distributed.How to make effective use of small hydropower resources is of great significance to the development of economy and ecology in rural areas.In recent years,the scheduling mode of small hydropower plants has changed from manual dispatching to intelligent optimization.The research and selection of excellent multi-objective optimization algorithm and the establishment of the actual scheduling model as tow main aspects play a vital role in scheduling.This paper focus on the study and improvement of decomposition-based multiobjective evolutionary algorithms,and an improved decomposition-based multi-objective evolutionary algorithm with an angle-baesd updating strategy has been proposed(MOEA/D-AU).Then,some improvement has been made on this algorithm and a decomposition-based multi-objective evolutionary algoritm with an adaptive angle-based updating strategy is proposed(MOEA/D-AAU).Finally,according to the actual situation,a reasonable optimal dispatching model of small hydropower stations is selected and established;MOEA/D-AAU is used in this model and working out results.The main work of this paper can be summarized as follows:1.As the evaluation criteria of the optimal solution of decomposition-based multi-objective evolutionary algorithm(MOEA/D)may easily lead to the poor distribution of the solution set,a methond of using angle to assist the selection of the optimal solution is added,which ensures the balance of convergence and distribution of the algorithm.The test results show that the new algorithm MOEA/D-AU has better distribution and convergence than the original algorithm MOEA/D.2.In order to make MOEA/D-AU performs better in the convergence process and has more space of adjustment.The idea of dynamic adjustment is introduced and a decomposition-based multi-objective evolutionary algorithm with an adaptive angle-based updating strategy has been proposed.The test results show that the new algorithm performs better in the convergence process and obtains a better result set under the condition of selecting the appropriate dynamic adjustment strategy.3.Based on reality,we have participated in the design and development of intelligent dispatching management system for small hydropower group.Considering the small hydropower group in the Jiu Gong watershed,the multi-objective scheduling model considering the maximum power generation,the minimum ecological water loss and the downstream irrigation water is selected.Using the proposed algorithm MOEA/D-AAU to solve the model,and provide a set of scientific scheduling schemes for the operation of small hydropower stations.4.Finally,summarizing the whole paper and puts forward further research and prospect.