Continuous Space Ant Colony Algorithm And In The Industrial Process Control Applications

Optimization is an important branch of mathematics,and a wide range of disciplines of subject. For the purpose of the actual problem, the best program can be chosen from many programs. Recently, accompanied by the rapid development of computer technology and optimized methods, all kinds of theoretical study have developed and practical applications are increasingly widespread. Ant colony algorithm has been put forward only in recent years, which is a new type of simulated evolutionary algorithm. The biggest probability will eventually approach the optimal solution of the problem after several iterations.Ant Colony Optimization Algorithm (ACO) is a new stimulated evolutionary algorithm attacking hard combinatorial optimization problems. It has showed a great deal of salient character and performed great value in its application. Choosing the analysis of character of Ant System (the basic algorithms of ant colony) as the research background, this paper focuses on the principle, the model, the behavior and its characteristics. In this paper, a kind of ACO algorithm have been proposed for solving the continuous space optimization and PID controller parameters optimization in industrial process control. Finally the results obtained via computer simulation show its validity. The main works and innovations as follows:1. The biological mechanism, the development and character of the ant colony algorithm are outline. The current situations of research and application of the ant colony algorithm is introduced. The superiority in solving combinatorial optimization problems and the shortages in solving the continuous space optimization problems are presented; in this paper, we also introduce the principle, the model, the characteristics and the management about the basic algorithms of ant colony (Ant System). And a list of improvements of continuous space optimization has been referenced and presents;2. On the foundation of summarizing and analyzing the existing research results, many details of original Ant Algorithm model have been implemented, including the number of division space, the number of searching ants and other parameters.3. Facing the limitations of the random search in the solution space, we use the certainty search method, Variable Metric Algorithm, the improved algorithm VACA has been gotten. The improved algorithm has been applied to one-dimensional and multi-dimensional continuous function optimization. Compared with the simulation results of the original ACO, the effectiveness of the approach can have been proved.4. In this paper, PID controller parameters optimization can be realized using the improved algorithm, to verify the effectiveness of its practical application. And experiments show that optimization efficiency and quality has been improved.The ant colony algorithm is a kind of stochastic explorative algorithm which has showed many excellent characteristics. It is demonstrated that the ant colony algorithm is more adaptive to the genetic algorithms and the simulated annealing algorithms which were fashionable for a time. Although some successful applications have been presented, it also has many problems for solving and making further investigation. Ant colony algorithm, which is different from other heuristic algorithm, has not formed the theory system. Parameter selection relies on more experiments and experience. It takes more time to calculate and it is prone to stagnation which shows that the algorithm in theory and practice has many problems to solve and research.