Trajectory Planning And Control Simulation Of Offshore Platform Knuckle Boom Crane

The knuckle boom crane is an important part of automatic drilling system.At present,the research on the crane is still to be carried on.The research on trajectory planning is less than the analysis of mechanism and parts.This paper uses the robot technology to do the following research on the crane:The kinematics and dynamic modeling of the crane can been obtained based on the relevant information.The crane has been simplified to a robot with two arms and the key points are connected by the 3-5-3 polynomial curve.The cost function is used as a weighted balance of traveling time and mechanical energy consumption.Using the genetic algorithm to calculate the objective function.An interpolation method based on five degrees B-spline for manipulators trajectory planning in joint space is proposed.The difference between the polynomial interpolation and the B-spline interpolation has been got.The reasonable number of interpolating points has been obtained based on the interpolated precision and time consumption.The multi-objective optimization problem based on the minimum time and energy consumption is established,and a novel multi-objective fireworks algorithm is proposed.The influence of two important parameters on the search efficiency has been studied,and the best parameter combination is obtained.The ideal trajectory was obtained after this optimization problem has been calculated.Secondly,the sliding mode control was compared with the PID control,and then the sliding mode control is applied to the control of the crane.RBF neural network approximates the nonlinear part of the system and uses the fuzzy sliding mode switching item instead of the fixed value of the switching item to eliminate the presence of chattering on the system.The simulation results show that the designed control law is reasonable and effective.Since the offshore platform is working in a complex condition on deep sea.The crane is influenced by the suspension of the wave.The system is simplified to a robot with a translational joint and three rotating joints and established a kinetic and dynamics model.Then the simulation results have been got.The influence of different amplitude and period of the wave on the tracking error is studied.