The Vibration Control Method Of Manipulators Handling A Flexible Payload

With modern science and technology changing and the rapid development of social economy, robot technology progresses continually. Manipulators have been widely used in the fields such as machinefacture. aeronautics, automobile industry, and electronic industry because of their own advantages. In aerospace, printed circuit boards, automobile industry and other fields, due to the applications of large-scale or thin metal plates, it is inevitable to cause vibration during operation. The accuracy and safety requirements are high in these areas. The vibration must be controlled. In the aerospace field, with the development in space, more and more flyers have been sent into space orbit. After a long time, the position or posture of them will change, which will bring damage to their operation, so they must be maintained. It is unable to complete these tasks totally depending on human and sometimes must rely on space robots. Because these panels are mostly made of flexible materials and bulky, vibration will be caused in their operations, not only makes the robot difficult to control, but also brings uncertainty into the operation of the flying objects. Therefore, the robot control must take into account the coupling characteristics of vibration.Through the current research at domestic and abroad, the study of manipulators operating flexible payload system is still at the initial stage. The coordinated control and vibration suppression have not formed a complete system in the academics, especially that the study of vibration suppression is rare. Most scholars only study a certain specific problem during operating flexible object. There are still many key problems to be solved. At present, a complex manipulator operating flexible payload system, has highly nonlinear characteristics, strong coupling features and infinite-dimensional degrees of freedom. Complete descriptions of system model, dynamic analysis and modeling have not formed a relatively complete theoretical system. The vibration of flexible payload occurs in the process of manipulator-coordinated operation, but an effective self-contained control strategy has not appeared. It should be processed when the load reaches a particular posture. When the flexible load is forced by external force, system kinematics and dynamics characteristics need to be further studied, as well as vibration control and compliant motion control methods. Therefore, the study of manipulators handling flexible payload vibration is significant. In this paper, the vibration control problems of manipulators operating flexible payload system are studied.First, the paper considers the flexible beam structure of the system, the finite element method is adopted to establish the dynamic model of the flexible payload, and the flexible multibody system is discretized into N units which connect only in the nodes. Each unit is an elastomer, and every node includes two degrees of freedom which are angular displacement and line displacement. Then, according to Lagrange equations the dynamics model of the system is derived. Combine each unit and then obtain the dynamic equations of flexible payload. The relationship of coordinates is expressed. The relationship between the joint angle coordinate and the end effectors coordinate is given, through which the dynamics equation of the system vibration is established.Second, in the paper, vibration control of flexible payload is based on the input forming technology, combining with neural network approximating modeling uncertainties. Because of the intrinsic characteristics of flexible payload, in the actual operation process, it will inevitably produce vibration. According to the actual requirements. the input generator with certain robustness is designed. As the rational robustness constraints are established and the relationship among residual vibration, pulse amplitude and time is given, the SI input shaper can be directly designed under arbitrary mode. And the use of the cyclical nature of the input shaper shortens the molding time. But the uncertainties produced during establishing the input shaper and model will affect the overall control performance of the system, so the RBF neural network algorithm is used to approximate the system model, combining with adaptive control law and then the closed-loop control based on error is designed. The stability of the system is proven based on Lyapunov equations and the simulation is done.In the actual industrial production, extraction of flexible load has unpredictability, extraction of quality, density and elastic modulus may change. According to the unpredictability of the flexible payload, the controller incompletely depending on model is designed with auto-disturbance-rejection control technique. Because installing sensors in flexible load is not realistic, the ends of manipulator can be installed sensors so the vibration information of the contact points can be measured and derive the vibration state. Taking it as the output, for the manipulators and flexible payloads. design two extended state observer in parallel. By dynamic estimating every part of the unknown model dynamic and external disturbance. according to the relationship between the estimate value and the actual value, the feedback control law is designed to improve the system performance, containing vibration suppressing and trajectory tracking. Finally, simulations validate the effectiveness of the design.Finally, according to the research experience in manipulator systems, flexible multibody systems and thorough study on mechanics and control theory, the conclusion and the perspective of future research are given at the end of the paper.