Improved PSO Algorithm And Its Application On Combinatorial Optimization Problems

Based on detailed analysis on the standard PSO algorithm and its mathematical principles and flow chart, the main works of the paper are introduced. Firstly, we reconstruct of the PSO algorithm by introducing the dynamic social impact to discrete areas of the formula. The advanced algorithm has better performance in avoiding prematurely into the local optimum than the standards of PSO algorithm. Re-setting parameters and improving the PSO algorithm for solving TSP problem. Numerical experiment for 14 nodes of the standard TSP issue has been done. Secondly, the hybrid algorithm PGA has been given by combining the advantages of PSO and genetic algorithm. PGA has been applied to reconstruction in equipping electric network. Using the genetic algorithm in the feedback from random mutation operator and high frequency cross-operator and combining with PSO can avoid a partial optimal algorithm and guide the direction of the evolution of chromosome. PGA can converge to the global optimal solution fast than some other algorithms on experiments. Finally, writer's view about the PSO algorithm prospects and development trends has been given.