Particle Swarm Optimization And Discrete Optimization Problems Of Applied Research,

With the rapid development of the computer science and technology, the demand for science and technology is in increase. Therefore high-performance optimization technology and intelligent computation is in urgent need. Particle swarm optimization algorithm (PSO) is a kind of rising intelligence optimizer, which stems from the simulation of birds flock's looking for food. PSO is simple in concept, few in parameters and easy in implementation, so it shows great potential in practice and has been widely applied in many areas. However, many problems which are solved by PSO are not discrete ones, in real life, a wide variety of problems can be represented by discrete optimization models. In this paper, the principle of PSO is analyzed and its applications in various discrete problems are studied.Firstly, based on the linear discrete-time system theory, the convergence behavior of PSO is presented, and the condition of convergence for PSO is deduced as well. On the basis of qualitative analysis of PSO parameters, the nonlinearly decreasing weight of PSO is proposed. Then the experimental analysis of inertia weight and acceleration constants is done, and the guideline of better choosing these parameters is summarized.Secondly, the application of PSO in 0-1 combination optimization problem is studied by studying its application in raw mix slurry optimal arrangement of alumina process. To improve the adaptability of PSO, convergence rate and evolutionary rate are introduced to adjust inertia weight decreasing nonlinearly, adaptively and dynamically. The improved strategy of inertia weight is introduced into discrete binary particle swarm optimization algorithm to solve raw mix slurry optimal arrangement problem, the simulation results show its superiority.Thirdly, the application of PSO in random combination optimization problem was studied by studying its application in pickup and delivery task assignment problem in transportation. A new discrete particle swarm optimization algorithm for solving this problem is proposed, an arctangent function is introduced to further adjust the position formula of standard PSO to ensure the equity and rationality of optimization. The experimental results prove that the proposed algorithm is effective and feasible.Finally, application of PSO in sorting problem is studied by studying its application in optimal operation for placing-in and taking-out of wagons at enterprise railway. This problem is solved by the PSO that presenting the concepts of swap operator and swap sequence. For this algorithm is easy to encounter local optimum, a hybrid algorithm PSO-SA is proposed, in which the global optimal position of PSO is adjusted by optimization of SA, thus the algorithm can skip the local optimum. Solving the optimal operation for placing-in and taking-out of wagons by PSO-SA, the results show that PSO-SA is effective to solve large-scale problem.