Research On Improvement And Application Of Artificial Ecosystem Optimization Algorithm

Artificial ecosystem optimization algorithm(AEO)is a meta-heuristic algorithm inspired by the energy flow of ecosystems on earth.At present,the algorithm has been widely used in scientific computing,engineering design,economic scheduling and other fields due to its characteristics of few parameters,high search efficiency,and strong exploiting ability.However,researchers also found that the algorithm has some shortcomings in the process of studying the AEO algorithm.For example,there is only one producer to guide the search of the population,the mechanism of the producer affects the balance between exploration and exploitation,and it is easy to fall into local optimum.In this paper,the deficiency of AEO algorithm is improved,and the improved algorithm is used to solve some applications.The main work of this thesis is as follows:(1)To achieve nonlinear data modeling,three strategies are introduced to enhance the performance of the AEO algorithm,and the enhanced artificial ecosystem-based optimization to self-organizing radial basis function Neural Network(EAEO-SORBF)is proposed.Three strategies are introduced to enhance the performance of AEO algorithm,and an encoding method which can optimize the network structure and parameters at the same time is designed.Finally,EAEO-SORBF and other 12 meta-heuristic algorithms were compared and analyzed on 12 UCI popular classification datasets and credit card fraud detection applications.The experimental results show that the EAEO-SORBF algorithm has high classification accuracy and good generalization ability when modeling strongly nonlinear data.(2)In order to better balance the exploration and exploitation of the algorithm,dynamic balance parameters and a parasitic strategy are introduced,and a simplified artificial ecosystem-based optimization(SAEO)is proposed.The proposed algorithm is applied to inventory control problem.The experimental results show that the SAEO algorithm has strong search performance on inventory control problems and can obtain high-quality solutions.(3)In order to solve the combinatorial optimization problem,a discrete artificial ecosystem-based optimization(DAEO)is proposed.Five neighborhood search strategies are introduced to enable AEO algorithm to search effectively in discrete space.Finally,the DAEO algorithm is applied to the spherical capacitated vehicle routing problem(SCVRP),and compared with other 5 discrete optimization algorithms on 32 test instances.The experimental results show that DEAO algorithm has strong competitiveness in solving SCVRP problems and can obtain a better solution than other meta-heuristic algorithms.