Research On The Control Of Lifting Load System Of Ship-mounted Cranes

Currently, with the global vigorous development of marine resources, the crane ship isone of the engineering ships and essential tool to transfer cargos, to build ports on the oceanof the offshore engineering. And it is also concerned by many countries. Crane ship willinevitably be affected by the wind and waves when it works on the ocean, which will leadship to swinging and cause operations to be abnormal. Seriously, it can cause the shipcapsizing. Consequences would be unbearable to contemplate. Therefore, it is necessary toresearch for crane ship hanging objects control system to make the hanging pendulum anglein a safe range. And improve the automation and intelligent of lifting load system, what hasgreat significance to ensure rapid and smooth operation of the crane ship.The background of this paper is precisely relying on the environment of marine.Therefore, the movement of ocean wave is analyzed firstly in this paper. The wave motionequation of sea waves is established in the first place. And then, in order to establish asystem of kinetic equations of hanging objects by using Lagrange method, the regular wave,as the interference of incentive system, is used to explore the problem. The following isusing the numerical analysis method simulated in MATLAB. Three factors which influencethe dynamic response of suspension system respectively is the rope length, the lifting speedand the excitation frequency.According to the results of the dynamic response of lifting load system, an overallframework of feed-forward feedback controller is set for lifting load control system. Thefeed-forward controller is built up by ELM. The feedback controller is designed bycombining the neural network control with the sliding mode variable structure control.Combine with the dynamic response results and the structure characteristics of the craneship, using two hydraulic cylinders as the actuator in the end of the crane’s manipulator.Their motion directions are perpendicular to each other. They can respectively for hullrolling and pitching Angle compensation. According to the method of the mechanical armmodeling, the lifting arm control system is established by D_H method.Next, the crane swing trajectory tracking is studied. It is used for feed-forward control.Due to the defects in the traditional neural network, a method of tracking the extremelearning machine based on the trajectory of ELM algorithm is proposed. The theory andsteps of ELM algorithm are discussed in this paper. At last, the results of this algorithmcompared with the traditional BP neural network algorithm through the simulation of tracking renderings.Finally, a neural sliding mode controller is designed. It is used for feedback control.Sliding mode variable structure control has strong robustness against external disturbances.Combined with neural network super self-learning ability, switching gain of sliding modevariable structure control can be adjusted by ELM. It can reduce the chattering of slidingmode control and improve the performance of the control. Then absorb controller to theswing control of lifting load system, analyze the simulation of lifting load control system.The results show that the control system can achieve better hanging objects swing anglecontrol.