Decentralized Optimal Fault Tolerant Control For Reconfigurable Manipulators System

Due to its high flexibility and easy maintenance,reconfigurable manipulator has strong applicability for complex environment survey and accurate operation tasks.In recent years,the research of reconfigurable manipulator has been widely developed,and it is often used in rehabilitation training,searching,rescuing,complex-accurate operating and other engineering applications.Reconfigurable manipulator is usually composed of power supply,control system and various related sensor modules.During the execution of tasks,reconfigurable manipulator can change its configuration to adapt to different task requirements.Each subsystem is composed of standardized mechanical modules and electrical components with the same interface,and these subsystems can be freely combined according to the needs of different tasks.According to the idea of reconfigurable design,the reconfigurable manipulator is not only convenient for installation and maintenance,but also greatly improves the stability of the system and make its structure more compact.Therefore,in-depth study of reconfigurable manipulator has far-reaching theoretical significance.Because of the high performance of the reconfigurable manipulator,people usually make it work in a variety of harsh environments.Due to external factors and long-time work,the actuator,sensor and other components of the reconfigurable manipulator occurs fault inevitably.At this time,the stability of the system will be destroyed and the control performance will decline.When a fault occurs in the system,the appropriate fault-tolerant control method is studied to compensate it,so that the system can maintain a certain performance and continue to run steadily,to greatly reduce the economic losses and safety hazards suffered by human beings.Therefore,the research of fault-tolerant control method is indispensable for reconfigurable manipulator system.At the same time,reconfigurable manipulator system should not only consider robustness and accuracy,but also consider the energy saving of the system.As an important part of modern control theory,the core problem of optimal control research is to select the appropriate control strategy for a given controlled system to achieve the optimal performance index of the system.For the reconfigurable manipulator system,it is necessary to obtain the optimal control strategy for solving the Hamilton-Jacobi-Bellman equation,which is a kind of nonlinear partial differential equation and difficult to obtain the optimal solution by analytical method.Adaptive dynamic programming method is a powerful tool to solve the optimal control problem of nonlinear systems.In adaptive dynamic programming system,neural network is designed to approximate the performance index function and estimate the solution of Hamilton-Jacobi-Bellman equation.In order to ensure the reconfigurable manipulator system with actuator fault to maintain stable operation and reduce the energy consumption cost of the controller,this paper combines the adaptive optimal control method with the fault-tolerant control method,and considers the output restriction of actuator and the combination of decentralized control and optimal tolerance-fault control,the main contents are as follows:1.Research on the significance and background of topic selection.From the research of reconfigurable manipulator at domestic and abroad,this paper analyzes the current situation and development of reconfigurable manipulator,and simply explains the existing problems and the advantages of decentralized control,fault tolerant control and adaptive dynamic programming,and determines the research objective of this paper.2.Dynamic modeling of reconfigurable manipulator.Combined with the concept of reconfiguration and the uncertainty of the system,the dynamic model of reconfigurable manipulator is obtained by constructing Newton-Euler equation.According to the dynamic model of the manipulator,the state space expression of the reconfigurable manipulator system is established,and the fault model of the reconfigurable manipulator is established by taking the actuator fault into account.At the same time,because the reconfigurable manipulator is composed of various modules,based on the local joint dynamic information,the dynamic model of the reconfigurable manipulator system based on joint torque feedback is constructed.The dynamic model of the reconfigurable manipulator system is decomposed into several dynamic subsystems,and the dynamic characteristics of the subsystems are deeply analyzed,which provides the necessary model basis for the subsequent control methods.3.Adaptive optimal fault tolerant control for reconfigurable manipulators with actuator output constraints.In view of the actuator output limited phenomenon in practical application,the actuator saturation is considered in the fault tolerant controller of reconfigurable manipulator system.In this paper,an optimal active fault tolerant control method based on adaptive dynamic programming is proposed for reconfigurable manipulator system with actuator output constraint.Firstly,using hyperbolic tangent function to deal with actuator saturation constraint of reconfigurable manipulator system.Secondly,the fault observer is designed to estimate the actuator fault,and the performance index function is constructed by using the fault observation function.Then the Hamilton Jacobi Bellman equation is solved by using the idea of adaptive dynamic programming.A critic neural network is established to approximate the optimal performance index function.Finally,based on Lyapunov theory,it proves that the closed-loop manipulator system is asymptotically stable under anti saturation optimal fault tolerant control,and the effectiveness of the proposed scheme is verified by simulation experiment.4.Research on decentralized active fault tolerant control method of reconfigurable manipulator based on adaptive dynamic programming.Comparing with centralized control,decentralized control allocates the input of the whole system to a group of local subsystem controllers,which can effectively reduce the huge computational burden of centralized control.In the actual application,the actuator of the system may be damaged to varying degrees.Therefore,considering the actuator fault in the manipulator system,an active decentralized fault tolerant control strategy based on adaptive dynamic programming is designed.Firstly,model-based compensation controller,fault-tolerant controller based on fault observer and optimal controller based on ADP are studied.Then,the critic neural network is used to approximate the optimal performance index function which designed in the decentralized optimal control,and the robust control method is used to compensate the subsystem friction term and cross-linking coupling term.Finally,the asymptotic stability of the closed-loop system is proved by Lyapunov theory and the effectiveness of the method is verified by experiments.