Research And Simulation On Anti-swing Of Container Crane Using Fuzzy Intelligent Control

In order to promote the loading/unloading efficiency of container crane, it is very important to decrease or avoid the swing of container and spreader. Electric anti-swing is an active method. It does not rely on the experience of the operator, but take the control of both anti-swing and movement of the trolley into account, which makes the loading/unloading more efficient, and alleviates the operator's burden of work. It is a trend to make electric anti-swing into application in ports to realize the automation of loading/unloading.In conventional cybernetics, a precise mathematical model of some object is essential for automation. But, the process of loading/unloading with container crane is complicate and nonlinear. Also, there would be some affection of stochastic factors (e.g. wind) in ports. It is very difficult to build a precise mathematical model of the system. Or even some model is built, it would probably be impossible to solve because of the complication. That is, it is very difficult to control the anti-swing system by means of conventional cybernetics. Being different from conventional cybernetics, intelligent cybernetics needn't build the precise mathematical model. Especially, it suits the control of nonlinear and uncertain system. Therefore, it is a good scheme to apply intelligent cybernetics to the electric anti-swing of container crane.Two intelligent control methods, fuzzy control and fuzzy neural network control, are introduced in this thesis. Based on the kinetic characteristic of container crane, the design principle and means of fuzzy controller for the anti-swing of container crane are put forward. Generally, the input of controller adopts two variables as to the scheme considering the movement of trolley and the sway of load. This method does not take the change of position and swing angle into account, and then results in low precision of control. In this thesis, four variables are adopted for the input of controller, while the controller is still designed in term of two dimensions. Thereby, this method not only improves the control precision, but also simplifies the process of calculation. Furthermore, because of the subjectivity of theselection control rules and membership function in simple fuzzy control, this thesis put forward a design method of fuzzy neural network controller with clear space frame, good ability of self-study and nonlinear approach capacity using an improved fuzzy neural network model.With these two intelligent control methods, this thesis simulates the fuzzy intelligent control of container crane anti-swing on computer. According to the numeric solving of container crane dynamics model, a simulation software of fuzzy control and fuzzy neural network control is developed on the platform of Visual C++6.0, based on the anti-swing of container crane. In a certain extent, a fuzzy intelligent control simulation of the anti-swing of container crane is realized. Elementarily proved, a preferable anti-swing effect is obtained with the fuzzy intelligent control methods introduced in this thesis.