Based On The Information Entropy Weighting To The Research And Application Of Ant Colony Algorithm

Being inspired from the mechanism of biological evolution, Italy scholar M.Dorigo, V.Maniezzo et al. put forward a new type of evolutionary optimization algorithm according to the simulation of ant colony foraging behaviour, which called Ant colony algorithm. The algorithm has certain advantage in solving optimization problems in discrete areas such as TSP, assignment, network routing, workshop scheduling, and has achieved favorable results. At present, the ant colony algorithm has become an pop research focus and topic in the field of international intelligent calculation.This article focuses on the study of the improvement of ant colony algorithm and its application. Aiming at the shortcomings of ant colony algorithm in long-time of solving, lower-speed of convergence and non-ideal of local search, the thesis brings up the modified methords in three aspects separately:the path choice strategy, pheromone update strategy and calculation convergence judgment.Firstly, improving the path search strategy based on the MMAS algorithm. Raising the algorithmic randomness via the use of roulette method proposed by Xu Jingming. Then reducing the search space of dimension by using the candidate list, which leads the algorithm to be more focused on the search of interested portion. At the end, optimizing the quality of the solution by introducing the local search strategy. Secondly, in order to make better use of feedback information, increase the shorter path guide, the text introduced the weighted factor to improve the algorithm pheromone update strategy. Thirdly, by using the convergence judgment grounded on the information entropy algorithm, the text presents a weighted ant colony algorithm based on information entropy. The experimental results show that this algorithm has better ability of problem solving and execution efficiency.To verify the feasibility and validity of the algorithm, the entropy weighted ant colony algorithm based on the information entropy is applied to the field of intelligent transportation. This paper introduces construction process of practical traffic net-work, the correlation algorithm used to calculate the distance between any two points on earth and how to find the optimal path in the city traffic network by the weighted ant colony algorithm based on information entropy with combination of the AHP model. The experiment proves, this algorithm can accurately and quickly find the optimal path.