Ant Colony Algorithm And Its Application In PID Control And Dynamic Matrix Control

Ant colony algorithm is a new stimulated evolutionary algorithm, which has been successfully applied to solve optimization problems, especially in discrete space optimization.Ant colony algorithm is implemented to solve TSP problem and nonlinear function optimization problem. Computer simulation results manifest its drawback of the pheromone update strategy. In the basic ant colony algorithm, it does not attach enough importance to the key nodes, which have more effect on object function. In this case, the nodes in the best path may be searched with a small probability. This may lead to worse searching guide and a large amount of inefficient search. According to the analysis above, a new pheromone update method is proposed. The key nodes are weighted by function in different searching period, in this way the pheromone can be used to guide the searching of ants correctly. It avoids large amount of invalid searching and promotes the searching efficiency greatly. First using mathematical method to prove the proposed to be convergent, later it is applied into optimizing PID controller and dynamic controller. By simulation, it can be concluded that the proposed algorithm has better convergent rate and obtains better control parameters.Based on adaptive control theory and ant colony algorithm, ant colony adaptive PID control is proposed, which is proved to be feasible in simulation experiments. Stimulated by genetic algorithm, the behavior of selection,crossover,mutation can be introduced to ant colony algorithm. Applying this algorithm to optimize PID parameters, simulation results indicate that the proposed can overcome the commonly seen disadvantage of low convergent rate and show good performance.