Research On Improved Particle Swarm Optimization And Its Application In Flexible Job Shop Scheduling

Swarm intelligence optimization algorithm is a stochastic optimization algorithm which simulates the biological behavior in nature.It does not require complex constraints,nor does it need to create the corresponding mathematical model for specific problems.Therefore,swarm intelligence algorithm has developed into one of the classic ways to solve optimization problems.Particle swarm optimization(PSO)is proposed by simulating the flight behavior and foraging activities of bird swarms,and it has become a hot spot of scientific theory research because of its advantages such as few parameters and easy to realize.By observing the flight and foraging activities of birds,a particle swarm optimization algorithm with complete theoretical background has been created.Since the particle swarm algorithm was proposed,it has become a hot topic in scientific theoretical research because of its advantages such as fewer parameters and easy implementation.At the same time,it is widely used to solve practical problems in various fields.However,when solving complex or large-scale optimization problems,particle swarm optimization algorithms often cannot maintain the diversity of population,premature convergence and excessive dependence on parameter values.Therefore,it is necessary to carry out in-depth research on the particle swarm algorithm and explore a larger application space.The main points of thesis are as follows:1.Aiming at the problem that particle swarm optimization algorithm is easy to fall into local optimization and dependent on the parameter value when solving complex optimization problems,a particle swarm optimization algorithm with independent adaptive parameter adjustment is proposed.Firstly,the concepts of population fitness variance,particle evolution ability and evolution rate are defined,and then the independent adjustment strategies of inertia weight and learning factors are given.Experiments show that the algorithm not only effectively regulates the self-cognition and social cognition capabilities of particles,but also effectively balance the ability of local search and global search.2.In order to solve the problem that the population diversity of PSO is difficult to maintain,the convergence speed and accuracy are poor,a two-strategy cooperative particle swarm optimization algorithm is proposed.The population is divided dynamically based on the fitness value and evolution ability of the particle,and the particle reconstruction strategy and the differential variation strategy are adopted according to the characteristics of the sub-population.Through comparative experiments,it is found that the algorithm can not only effectively maintain population diversity,avoid falling into local extreme value,but also improve the convergence performance of the algorithm,especially in solving high-dimensional complex optimization problems.3.In order to verify the effectiveness and practicability of the algorithm,the improved algorithm is applied to solve the flexible job shop scheduling problem,which has a wide range of research value in combination optimization and actual production.In addition,this thesis also selects four excellent algorithms for solving FJSP problem as the comparison algorithm.Through comparison results,it is found that this algorithm can find a more ideal scheduling scheme,which effectively proves the efficiency and practical significance of the algorithm.