Based On Simulated Annealing Algorithm For TSP

TSP is a typical NP problem, which mainly manifested in the time complexity on solving problem. With the increasing scale of problem, It is in accordance with the exponential growth. The traditional method for solving NP problems it is still unable to find an efficient algorithm based on polynomial time. With the development of computing performance, simulated annealing algorithm provides an efficient way to achieve ideal results on the TSP problem solving within a specific range. However, due to lack of global search capability permissions for large-scale TSP problem prone to a narrow range of global solution space into a local extreme defects.This paper put forward a variety of group-based simulated annealing algorithm to solve TSP problems. The algorithm on one hand with the introduction of a solution based on multi-operator collaboration generating function to enhanced by the operation of the randomness of the new solution set to achieve the global search capabilities. On the other hand the introduction of multi-population mechanism based on different parameters configured help to improve solution accuracy. The improved SA achieved verification and simulation on scale of14,31and34cities TSP. The results showed that the SA based on improved multi-population has more high quality and accuracy in comparing to traditional algorithm...