Research On Constrained Cooperative Differential Evolution Algorithm For Large-scale Economic Scheduling Problems

With the rapid development of social economy,the demand of the whole society and users for power supply service is increasing,the research on economic load dispatching of power system has not kept pace with the pace of social development,so the optimization of economic load dispatching of power system has been concerned by researchers for a long time.The purpose of economic load dispatching in power system is to rationally distribute the output power of each generator to minimize the consumption cost under the premise of ensuring the safe use of electricity and various generation constraints.With the continuous improvement of the model of power system and the increase of the scale of system units,the system model has become a non-convex,nonlinear,high-dimensional and difficult problem to deal with constraints,which leads to the traditional methods get local optimal solution too easily and converge slowly in dealing with such problems.In recent years,scholars have used intelligent optimization algorithm to deal with economic load dispatching problems,compared with traditional methods,intelligent optimization algorithm has the advantages of strong robustness and high accuracy,but it is difficult to deal with a large-scale combination economic load dispatching problems with many constraints.This paper focuses on the improved differential evolution algorithm combined constraint handling technology to deal with economic load dispatching problems in large-scale power systems.The following tasks were completed.Firstly,aiming at the shortcomings of slow convergence speed and easy to fall into local optimization of differential evolution algorithm,a hybrid harmony differential evolution algorithm is proposed.In this hybrid algorithm,a selection mutation method is used to improve the diversity of the population,and an adaptive mutation factor control mechanism is adopted to enhance the convergence ability of the algorithm.Experiments on benchmark function IEEE CEC 2005 show that the hybrid harmony differential evolution algorithm significantly improves the optimization ability of the differential evolution algorithm.Secondly,considering the shortcomings of some existing constraint handling technologies on balancing objective function and constraint violation,a hybrid constraint handling technology based on improved differential evolution algorithm is proposed.The hybrid constraint handling technique combines stochastic Ranking(SR)and De 'b rules,and uses a feedback regulation mechanism to adjust the processing times of SR and De 'b rules to balance the objective function and constraint violation.Compared with several improved differential evolution algorithms proposed in IEEE CEC2017,this method has strong constraint handling capability.Finally,for solving the problem of slowly convergence and difficulty to deal with constraints in dealing with large-scale combination economic load dispatching problems with many constraints,a constrained cooperative adaptive multi-population differential evolutionary algorithm is proposed.In this method,according to the population feasibility rate,sub-populations are generated according to the mechanism of "one-to-many" or "one-to-one",a cooperative constraint processing mechanism is applied to enhance the information exchange between feasible and infeasible solutions.In order to test the optimization performance of this method,it is tested on the power system of 6 units,13 units,38 units,40 units and 140 units.The experimental data show that this method is more accurate and efficient than other intelligent algorithms.On the basis of the algorithm,the multi-objective processing technology is combined to deal with the environmental economic dispatching problem,and the experiment on IEEE 30 Bus six-generator system proves that the algorithm has the potential to deal with multi-objective problems.