Variable Correlation Analysis-based ACOR Algorithm An Its Applications In Economic Dispatch Of Power System

Ant colony algorithm is a meta heuristic optimization algorithm,inspired by the ant foraging behavior.In recent years,ant colony algorithm has been widely used in combinatorial optimization problems such as traveling salesman problem,graph coloring problem and network routing problem.With the further development of research,ant colony algorithm is also extended from discrete domain to continuous domain.Continuous domain ant colony algorithm uses Gaussian model to approximate the continuous probability distribution in real value space,and introduces the concept of tracking solution to accelerate the convergence speed of the algorithm.However,most of the continuous domain ant colony algorithms ignore the redundant correlation between the decision variables,which leads to low search efficiency and low convergence accuracy in solving some practical optimization problems.Based on this,this paper analyzes the correlation of decision variables in ant colony algorithm,proposes an improved algorithm based on variable correlation analysis,and carries out simulation verification,and further applies the proposed algorithm to the economic dispatch problem of power system.The main work of this paper is as follows:The first chapter introduces the relevant background knowledge of ant colony algorithm and power system and its research status at home and abroad,and introduces the main work of this paper.In the second chapter,we give some necessary preparatory knowledge,including ant colony algorithm and local preserving projection algorithm.In Chapter 3,a continuous ant colony algorithm based on local preserving projection and copula function is proposed.The local preserving projection algorithm is used to map the highdimensional data to the low dimensional space to decouple the redundant dependence of the data in each dimension,so as to reduce the redundant correlation and computational complexity between the decision variables,and improve the optimization efficiency of the algorithm.A new distribution estimation algorithm based on Copula function is proposed,and new solutions are sampled in the constructed low dimensional space.Finally,the effectiveness and convergence of the algorithm are verified by nine benchmark functions.In Chapter 4,a continuous ant colony algorithm based on mutual information and reverse learning is proposed.Specifically,in each iteration,each feasible solution interacts with the solution with the largest mutual information value,and at the same time,the position information between solution points is interacted with a certain probability.Furthermore,the improved reverse learning algorithm is used to increase the diversity of solutions and expand the search area.Finally,the effectiveness of the algorithm is verified by 12 benchmark functions.In the fifth chapter,the validity and practicability of the algorithm are verified in the load distribution problem of power system.Specifically,the proposed ACOR-LC algorithm and ACORMI algorithm are tested in the environment of 6 units,15 units and 40 units respectively.The test results show that the proposed algorithm has better,faster and more stable optimization performance than ACOR algorithm,ACOR-elite algorithm and Iacor-LS algorithm.The sixth chapter summarizes the main work of this paper and looks forward to some future research topics.