The Application Research Of Ant Colony Optimization Algorithm For Intelligent Control On Special Robots

With the development of science and technology, robot technology integrates the developing achievements of multi-disciplinary widely. It being used more and more widely in the field of human life, and continue to raise humans' attention. The special robot is a kind of machine equipment which can auto work. It can both accept the command control and work according to the principles of artificial intelligence control techniques. The working condition of the special robot is that it uses the joints to drive connecting rods so that the load which hanging at the end of the robot could be transported and loaded into the designated location. The special robot can be considered as a multi-joints mechanical arm system. Therefore, the thesis deduced kinematic and dynamic equations of the special robot by means of establishing mathematical model of mechanical arm, and established the mathematical model of the special robot.Various kinds of disturbance and changing of parameters are impact the special robot when the robot is working. So, the robot is a complex multi-input and multi-output nonlinear coupling system, it has dynamic characteristics of time-varying, strong coupling and nonlinear. Therefor, it's difficult to control the robot. The intelligent control gives a good method to deal with the controlling of such complex robot system. The fuzzy neural network (FNN) has combine both fuzzy logic control and neural network, it has both inductive reasoning ability of fuzzy control and parallel processing, self-learning and associate ability of neural network. It's able to control the robot system better. The thesis used fuzzy neural network intelligent control system to control the special robot, and maked simulation research. The result of simulation showed good control effects of the method.Finally, the thesis combined ant colony algorithm and fuzzy neural network intelligent control for intelligent optimization training of the network parameters. The dynamic simulation and control research for the special robot system were built in Simulink of MATLAB. The result showed that fuzzy neural network control system based on ant colony algorithm has a better result for trajectory tracking than fuzzy neural network intelligent controller based on BP algorithm. It improves the generalization ability of the control system, and makes the convergence of the system training faster than BP algorithm. And it avoids the shortcoming of easy to fall into local extreme value when training in neural network. The simulation analysis showed the feasibility and effectiveness of the system.