Ant Colony Algorithm And Its Application In The Distributed And Flexible Job Shop Scheduling Problem

As the rapid development of economic globalization and the increasing competition in the market, China’s manufacturing industry is in unprecedented trouble. In the process of transformation, more and more enterprises have realized the importance of the distributed and flexible production mode, which needs the cooperation of different workshop plants. Because of the difference of technological level, material resource and equipment performance in different plants, the existing production scheduling strategies are difficult to produce the expected results. Therefore, the paper proposed an improved ant colony algorithm based on parameter adaptive control to improve the performance of the scheduling of the distributed and flexible production mode.As a classical intelligent optimization algorithm, the performance of ant colony algorithm depends on the value of the parameters strongly. However, most traditional ant colony algorithm used fixed value of the parameters during the process of looking for the best solution, this made the algorithm trap into local optima easily and converge slowly, thus affecting the performance of the algorithm. This paper firstly introduced the theoretical knowledge and model of ant colony algorithm, and analyzed three parameters, including the heuristic factor α, the expectation heuristic factor β and the information persistent factor ρ, which had a critical influence on the performance of the algorithm. On the basis of above, the paper proposed a two-stage improved ant colony algorithm, which was based on adaptive adjustment of the parameters. Then the improved algorithm was applied to solve the distributed and flexible job-shop scheduling problem after verified its effectiveness in TSP problem. The main research results of this paper includes the following aspects:(1) In this paper, the two-stage improved ant colony algorithm was used to solve the distributed and flexible job-shop scheduling problem, and two pheromone matrix was designed according to the features of the problem.(2) According to the previous theoretical results, the value of the parameters should be different in each state of the algorithm. So, in the first stage, a clustering technology based on the shuffled frog leaping algorithm and K-means algorithm was used to judge the state of the ant colony. Then the parameters were adjusted adaptively in the different states.(3) Due to the complexity of the algorithm was too high when we used the clustering technology to adjust the parameters only, the improved algorithm applied chaos theory to adjust the parameters in the second stage, and the parameters were then tuned based on the ergodicity of chaos to jump out of local optima.The results of this paper not only solved the distributed and flexible job shop scheduling problem effectively, but also provided a new way of thinking for the improvement of swarm intelligence algorithm.This has a positive significance for the promotion and application of the swarm intelligence algorithm.