Research On Trajectory Planning And Vibration Control Of Rotating Beam Based On Structured Light Vision Detection

In aerospace and industrial applications,lightweight and efficient rotating flexible beam structure is widely used.Due to the structural characteristics of low stiffness and small damping,the low frequency elastic tip vibration will be excited and attenuated slowly when it performs the rotation operation task,which has a great impact on the service life and operation efficiency of the flexible beam.In addition,the flexible beam system has many nonlinear characteristics such as rigid flexible coupling,friction and reducer backlash.Therefore,it is necessary to design a reasonable and effective control strategy to improve the rotation smoothness and quickly suppress the residual vibration.The main work of this thesis includes:(1)The mathematical models of single flexible beam system and dual flexible beam system are established.Based on the assumed mode method to describe the elastic vibration,the dynamic model is derived and the kinematic equation is further extracted to provide the theoretical basis for the subsequent vibration control strategy;The nonlinear characteristic of backlash and coupling of flexible beam system is analyzed,and the composite control scheme of dual beam system has been determined.(2)The structured light vision sensor which composed of a point laser and industrial camera is adapted to measure the vibration of rotating flexible beam.Based on Zhang's calibration method,a calibration method of camera,laser line equation and motor shaft equation is proposed.The theoretical derivation of 3D reconstruction and vibration information extraction is completed.The error caused by the laser not perpendicular to the beam surface is compensated,and the overall measurement steps are summarized.(3)From the perspective of feed forward control to improve the rotation stability of single and dual flexible beam system,the optimal trajectory is designed for point-to-point motion planning.For the single beam system,polynomial and Fourier series are used as the basic trajectory,and particle swarm optimization is used to optimize the vibration objective function and the basic trajectory parameters based on the identified autoregressive moving average model.For the dual beam system,the coupled kinematics equation of the double beam is identified based on the state space form.The piecewise quintic polynomial spline function is considered as basic trajectory,and its parameter is optimized by adaptive genetic algorithm to achieve the optimal vibration suppression trajectory of dual beam system.(4)From the perspective of feedback control,the vibration controller is designed to suppress the elastic vibration caused by rotation motion of the flexible beam,so as to improve the operation efficiency.Due to the nonlinear uncertainties of flexible beam system,the effect of conventional linear controller is limited.Considering the nonlinear mapping ability and memory ability of the diagonal recurrent neural network,the corresponding identification network and control network are designed to realize the online identification and vibration suppression of the single flexible beam system.The fuzzy predictive control algorithm is designed for the dual flexible beam system,which breaks through the control dead zone of the backlash by using the segmented characteristics.Each motion section corresponds to the change of dominant fuzzy model.Combined with the predictive control,the vibration can be quickly suppressed.(5)Based on the above-mentioned measurement and control strategy,a single beam and dual beam experiment platform is built.For the sake of generality,the dual beam platform includes symmetrical beam and asymmetrical beam.The experiment result show that the vibration response of the flexible beam is smaller when the proposed two optimal trajectories are planned.Diagonal recurrent neural network control and fuzzy predictive control are used for the single beam system and dual beam system respectively.Compared with PD control,the experimental results show that the vibration suppression effect of the proposed control algorithm is better,especially for the small residual vibration.