Research On 3D Packing Based On Hybrid Heuristic Algorithm

As bin packing problem is solved and applied,logistics,warehousing and industries can be provided with efficient and convenient services.However,due to the complexity of bin packing problem and non-university of bin packing algorithm,the researches on bin packing problem and bin packing algorithm become a practical problem to be solved.With easy concepts and intelligent background,particle swarm algorithm which is easy to realize attracts widespread attention in solving bin packing problem.The researches on bin packing problem,a typical problem among combinatorial optimization problems,are of great significance in practical application and theoretical research and have a broad research prospect.This paper mainly improves the external parameter and internal topological structure of pturticle swarm algorithm and obtains hybrid heuristic algorithm to solve bin packing problem by combining particle swarm algorithm with heuristic algorithm.The main work and conclusions summarized in this paper are shown below:1.This paper adjusted the external parameter based on discrete problem,improved algorithm performance by adopting the stochastical decreasing of inertia weight and learning the linear increasing weight of factor c1 and linear decreasing weight of factor c2 and obtained the optimal parameter setting of solving bin packing problem,which increased the convergence rale by 86%and 36%compared with tradition particle swarm algorithms and WPSO algorithm respectively.2.This paper perfected the internal topological structure of particle swarm algorithm,constructed a dynamic topological struture and improved the dynamic particle swarm algorithm by fitness judgment before and after particle adjustment in a way of double judgement.In the experiment on 500 iterations of the improved particle swarm algorithm,the optimal solution could be obtained every 30 iterations,while the optimal solution could be obtained by other algorithms every 90,120 and 150 iterations respectively.Therefore,the convergence rate of the algorithm is improved.3.This paper obtained the new particle swarm algorithm based on the improvement of the internal parameter and internal topological structure of particle swarm algorithm,got hybrid heuristic algorithm by combining new particle swarm algorithm with three space division algorithm and solved three-dimensional bin packing problem.In different cases,this algorithm can obtain the optimal packing sequence every 30,50 and 80 iterations.However,other algorithms reach its optimal filling rate every 80 to 200 iterations,which is relatively slow.Therefore,it was found that the algorithm realized in this paper performs excellently in the aspect of convergence rate.