Research On Improved Particle Swarm Optimization For Mine Production Planning

In recent years,particle swarm algorithm is widely used in engineering problems because of simple structure and coding.Many experts apply particle swarm algorithm to mine production.The formulation of mine production plans can also be turned into the optimization problem.However,the shortcomings of the particle swarm algorithm are obvious,and the effect in formulating mine production plans needs to be improved.Therefore,this article conducts research on the particle swarm algorithm and applies it to the design of the prototype of the mine production plan management system.This paper designs two algorithms,CWBPSO(Change the Worst and the Best Particle Swarm Optimization)and LPSO(Levy Particle Swarm Optimization).CWBPSO has adopted two improvement measures.First,in order to enhance the search capabilities,combining the selection and crossover ideas in the Genetic Algorithm,in each iteration,two particles in the population are randomly selected as parents,and arithmetic crossover is performed to generate child.For child,if new child is better than the worst,the worst will be replaced,otherwise the original is retained.Second,in order to prevent falling into the local optimal solution,an interference factor is used to interfere with the best particle.If the best particle affected by interference is better than the original best particle,it will replace the original particles.When the particles are dispersed,the interference factor is small;when the particles are concentrated,the interference factor is large,making it difficult to fall into the local optimal solution.In order to prevent falling into a local optimal solution,LPSO is designed in conjunction with Levy flight.When the particles have not become better after several iterations,a Levy Flight step is added to the particle.Finally,we conducted experiments on 7 test functions such as Sphere function and3 dimensions.Results show that the first improvement measure can improve the convergence speed,and the second measure can improve the accuracy.The third measure can improve the convergence speed and accuracy.The improvement measures proposed in this paper can effectively improve the performance of particle swarm optimization and can be applied to solve the problem of mine production planning.