Mill Load Control System Based on the Improved Ant Colony Neural Network

Abstract

Taking into account the complexity, nonlinearity and uncertainty of mill load control system, an improved ant colony neural network control strategy was proposed, based on the combination between dynamic local pheromone update and global pheromone update with an adaptive control pheromone. This improved algorithm was used to tune PID parameters and to control mill load. The results from simulation test indicated that many assumptions about the controlled object characteristics in mill load system were effectively avoided via the improved ant colony neural network PID approach. Meanwhile, the complicated nonlinear equations were also simplified. Consequently, the mill load system can be controlled quickly and accurately with a satisfactory accuracy and dynamic performance.