Research On Anti-sway Control Strategy Of Gantry And Overhead Crane Based On RBF Neural Network

Gantry and overhead cranes are widely used in production and logistics because of their strong load capacity,high site utilization rate,flexible operation,and strong versatility.Generally,the payload of a gantry and overhead crane is connected to the hoisting trolley through a wire rope.The start,braking,acceleration and deceleration of the trolley during the hoisting process will cause the payload to swing and the residual swing after reaching the target position,which affects the precise positioning of the payload and reduces work efficiency,seriously it will cause the payload to fall or the fatigue of crane metal structure,which poses a great safety hazard.Therefore,this paper presents an effective trajectory tracking control algorithm for the anti-sway control problem of gantry and overhead cranes with parameter uncertainties and external disturbances.The main content and results of this article are as follows:(1)The planar double-pendulum dynamic model of an underactuated gantry and overhead crane was established by the Lagrange energy equation of the second kind.Then the dynamics model was linearized near the equilibrium point of the system,and it was divided into the actuated part and the underactuated part,which laid a foundation for the subsequent controller design.(2)An ideal sliding mode tracking control law based on the precise dynamic model of the system was proposed to address the problem of the anti-sway trajectory tracking control problem of an underacuated gantry and overhead crane with random external disturbances.Then,an adaptive sliding mode controller based on the RBF neural network(ASMC)was proposed considering of the uncertainty of system parameters,in which the RBF neural network is used to estimate the unknown dynamics of the system,and the adaptive algorithm is used to estimate the quality of the trolley.Then the asymptotic stability of the error closedloop system under the action of ASMC was proved based on Lyapunov theory and La Salle’s invariance principle,and the control performance of ASMC was verified by simulation analysis.The simulation results show that ASMC can effectively track the ideal trajectory of trolley,thereby achieving the suppression of the residual swing angle of the payload.Simultaneously,the simulation results also show that the controller is strongly robust.(3)A directly robust adaptive controller based on the RBF neural network(DRAC)was proposed to solve the problem of high-frequency chattering in the control output of the ASMC controller,and to further improve the control performance.The controller is composed of one RBF neural network,thus the output of the neural network is directly used as the control input of the system.The control law does not include the switching control law part.Instead,the system relies on the design of the neural network weight adaptive law to keep it stable under disturbance conditions,and the controller structure is further simplified than ASMC.Then the asymptotic stability of the error closed-loop system under the action of DRAC was proved based on Lyapunov theory,and the control performance of DRAC was verified by simulation analysis.The simulation results show that the tracking performance and anti-sway effect of DRAC are greatly improved compared with ASMC,and it has stronger robustness to many system parameters.(4)A saturation control method based on RBF neural network(DRAC-OS)was proposed to solve the problem of controllers with limited control output.In order to ensure that the controller can still maintain the system stability even when the control output has a saturation limit,this method adds an anti-saturation module based on the DRAC controller,which is an RBF neural network.Then the asymptotic stability of the error closed-loop system under the action of DRAC-OS was proved based on Lyapunov theory,and the anti-saturation control performance of DRAC-OS was verified by simulation analysis.The simulation results show that DRAC-OS can maintain excellent tracking performance when the control output is limited.