Research On Improved Estimation Of Distribution Algorithm And Its Application In Long-term Operation Of Cascaded Reservoirs

China is rich in hydropower resources,hydropower resources are in the stage of sustainable development.It is of great theoretical value and practical significance to study the optimal operation of reservoirs to make full use of the regulating and storage functions of hydropower stations and maximize the comprehensive utilization benefits of projects.In this paper,the improvement and application of the intelligent algorithm for the joint optimal operation of cascade reservoirs are studied from the aspects of theoretical method and practical application.First of all,on the basis of systematic research on related theories and models of long-term optimal operation of cascade reservoirs,this paper focuses on the differences of existing model solutions,and analyzes the significance of exploring new intelligent algorithms for optimal operation of cascade reservoirs.Secondly,Estimation of Distribution Algorithms(EDA)improvement mechanism was studied,in view of the weak local searching ability of the distribution of the estimation algorithm,iterative late is not easy to jump out of local optimal solution of the shortcomings,puts forward an Improved Estimation of Distribution Algorithms based on Simulated Annealing mechanism(S AEDA),at the beginning of the iteration kept EDA algorithm,the advantages of quick convergence and global optimization ability,and in the late iteration algorithm stagnation,using Simulated Annealing mechanism,using Metropolis accept criterion to the characteristics of a certain probability to accept an inferior solution,increase species diversity,the algorithm is optimal,jump out of the current and further search for the local optimal solution;Furthermore,the test results of six test functions by,applying SAEDA show that compared with EDA algorithm and PSO,SAEDA algorithm has higher convergence accuracy,stronger stability,faster convergence speed than EDA algorithm,and improved optimization performance.Finally,SAEDA algorithm was applied to calculating swarm optimization scheduling model of cascade reservoirs,calculation results show that SAEDA algorithm can obtain satisfactory results,improve the power generation benefits of cascade reservoirs,by comparing the proves the feasibility and practicability of the improved algorithm,to solve reservoir group of long-term optimal scheduling problem provides a new way.