Research Of Intelligent Control Method For Special Crane Based On Ant Colony Optimization Algorithm

The object of study of this thesis is one kind of special crane, which is a combination ofindustrial robot and crane system. The special crane is a new intelligent ship-borne craneequipment, it possesses both robotic and crane systemic structural characteristics controlmode, it consists of a set of jibs, slewing sustain device, slewing mechanism and manysensors which hinged together in order to Complete pitching, rotation and other movements.The special crane is driven by each joint to make the jibs move to deliver the load which isend of the jib to the designated location. In view of this special crane’s characteristics ofmovement is similar to the multiple joints system, the thesis establishs the kinematic equationand dynamic equation of the special crane by refer to the inference rules of three jointsmanipulator model, which is given the mathematical model of special crane.Second, because of the special crane work on the vessel, so it can’t avoid the adverseimpact in the work by the movement of waves and many other uncertain factors, and thenonlinearity of special crane is very strong, the dynamic behavior of special crane is complex.So it is a kind of complex, nonlinear mimo system, which has the dynamic characteristics ofchanging with time and strong coupling and so on, it is very difficult to control. Withintelligent control methods such as fuzzy control and neural network develop rapidly, thecontrol problem of this kind of complex mechanical system won more solutions. Because ofthe simple fuzzy control and neural network control has its advantages and disadvantages, ifthe two methods together constitute a fuzzy neural network (FNN) controller, it not only cansum up and reasoning information, but also to process information parallel. At the same timethe fuzzy neural network also has the ability to learn itself and has the excellent ability ofgeneralization, which provides a good control method to solve the control problem of thecomplex mechanical system. The thesis makes a research of the simulation of trajectorytracking control for special crane dynamic model by using the controller of FNN; thesimulation results verify the validity of the control system.Finally, the thesis studies the basic ant colony algorithm and the adaptive ant colonyalgorithm and makes them apply to the optimization of parameters of fuzzy neural networkcontrol system, and makes a simulation of trajectory tracking control for special crane systemby using MATLAB simulation software. Simulation results show that compare with thetraditional PD control, fuzzy control and the fuzzy neural network system which based on theBP algorithm training, the fuzzy neural network control system based on the ant colony algorithm training has the more stable and accurate trajectory tracking effect and the trainingspeed is faster than BP algorithm. Compare with the basic ant colony algorithm, the adaptiveant colony algorithm is faster in training speed and Better trajectory tracking error effect.Through the simulation the thesis analyzes the feasibility and effectiveness of the system.