Optimization Problem Solving Based On Adaptive Simulated Annealing Particle Swarm Optimization And Its Application

Optimization problem is often encountered in daily life,theoretical research and production practice.With the development of science and technology research,optimization problems in scientific theory and industry have presented characteristics of multi-modal,nonlinear and complex constraints.Traditional mathematical methods can not solve these problems efficiently in real time.This has led researchers to turn their eyes to evolutionary computing methods such as genetic algorithm(GA)and particle swarm optimization(PSO).Particle swarm optimization(PSO),as one of the typical algorithms,is widely used in the field of single-objective and multi-objective optimization due to its simple principle,fast convergence speed and easy implementation.However,p SO also has some defects,such as premature convergence and poor convergence accuracy,which are easy to fall into the local optimal region.Therefore,how to effectively maintain the diversity of the population,avoid particles falling into local optimum,and speed up the convergence of the population has been the focus of many scholars.In view of the shortcomings of particle swarm optimization algorithm in the process of population optimization,this paper incorporates the probability selection characteristics of simulated annealing algorithm to make relevant improvements.The main research contents are as follows:1.Several traditional numerical methods and new stochastic optimization methods for optimization problems are introduced.The basic principle of particle swarm optimization(PSO)is described,the advantages and disadvantages of the algorithm and the influence of each parameter on the algorithm performance are analyzed.Classical improved particle swarm optimization(PSO)with inertia weight and shrinkage factor is discussed.2.The advantages of particle swarm optimization in the field of multi-objective optimization are analyzed.Combined with the characteristics of particle swarm optimization and multi-objective optimization,analysis of the single objective of particle swarm optimization with multi-objective particle swarm optimization,at the same time introduces the multi-objective particle swarm optimization performance of three key problems: the determination of individual extremum,global extremum selection,maintenance of external species diversity,and introduces the corresponding improving ideas.3.An adaptive simulated annealing particle swarm optimization(ASAPSO)based on Metropolis criterion is proposed.ASAPSO uses a new adaptive extreme inertial weight method to balance the global search and local search process of particle population,and constructs the central particle of particle flight learning communication.The particle follows the central particle in a certain probability to approach the global optimal solution,which can effectively avoid the particle population falling into the local optimal region.Simulation experiments show that ASAPSO has significantly improved convergence rate and convergence accuracy compared with other test algorithms in various standard test functions,and can effectively improve the optimization efficiency of particle population.4.The image segmentation problem is regarded as a multi-objective optimization problem,the maximum inter-class variance and maximum entropy criterion are selected to construct the fitness function,and a multi-objective simulated annealing particle swarm optimization(MOPS-SA)is proposed and applied to the image segmentation problem.MOPS--SA selects one of the non-dominated solutions according to the weighted ratio as the optimal threshold combination to segment the image.Segmentation experiments on multiple Berkerly images show that.The proposed algorithm can improve the visual effect of image segmentation effectively.