Research On Improved Ant Colony Optimization

An ant colony optimization (ACO) is a stochastic optimization algorithm for complex optimization problems, which first proposed by M. Dorigo in 1991 and solved the traveling salesman problem well. In the past two decades, ACO algorithms are successful in applications of many NP-hard problems, such as traveling salesman problem, quadratic assignment problem, job-shop problem, vehicle-routing problem and graph coloring problem and so on. Because of its excellent performance, ACO is also obtained increasing attention of researchers.The improved ACO algorithms and applications are studied in this thesis. The main contents are as follows:(1) In order to overcome the drawbacks of premature convergence or low convergence speed, a novel hybrid ant colony optimization algorithm (NHACO) is proposed. The algorithm adopts the crossover and mutation operators of genetic algorithm to avoid trapping into a local optimum. Furthermore, a novel initial distribution of pheromone is used to improve the search ability. The results of test functions demonstrate the efficiency of the proposed algorithm.(2) The improved continuous ant colony optimization algorithm (ICACO) is presented for the continuous space optimization problems. Satisfactory solutions of 9 benchmark functions are obtained by using the ICACO algorithm.(3) The ICACO algorithm is applied to optimize the RBF neural network for the soft-sensor modeling of a continuous stirred tank reactor (CSTR). The objective function is designed by considering the conflicting between the modeling precision and the network complexity. The simulation results show the effectiveness of the proposed algorithm.