Adaptive neural network control of flexible joint robots based on feedback linearization

This paper presents an interconnection and damping assignment passivity based control (IDA-PBC) for the position and impedance control of flexible joint robot. In considering of the joint damping and with some simplification of friction, a flexible joint robot system is descripted as an port-controlled Hamiltonian system with dissipation(PCHD). Then, giving a desired Hamiltonian function which both shapes potential and kinetic energies, the control laws of position control and impedance control is derived by the algebraic IDA-PBC method, which preserves the PCHD structure of the close-loop system. Comparing with the traditional passivity based control (PBC), the proposed method gives a new explanation of the control of flexible joint robot by introducing the self-damping and cross-damping terms, which make the gain selecting procedure more intuitive and comfortable. Simulation and experiments are done to verify the proposed controller, and the results shows a promising performance both in position and impedance control.

Adaptive Control of Flexible Joint Robots with Arbitrary Stiffness

In this paper, a stability control approach of flexible joint robot (FJR) based T-S fuzzy model using fuzzy Lyapunov function is proposed. First, the Takagi and Sugeno (T-S) fuzzy model is employed to approximate the flexible joint robot, then a fuzzy controller is developed based on parallel distributed compensation principle (PDC). The non-quadratic stability conditions for the flexible joint robot control system are proposed by using fuzzy Lyapunov function combined with linear matrix inequality (LMI). Finally, the simulation example is given to illustrate the effectiveness of the proposed method.

Improved Takagi–Sugeno fuzzy model-based control of flexible joint robot via Hybrid-Taguchi genetic algorithm

<para xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink"> This paper investigates the problem of global output feedback tracking control of flexible joint robots. Despite the fact that only link position and actuator position are available from measurements, the proposed controller ensures that the link position globally tracks the desired trajectory while keeping all the remaining signals bounded. The controller development uses a partial state-feedback linearization technique combined with the integrator backstepping control design method whereas a filter and an observer are utilized to remove the requirement of link and actuator velocity measurements. Partial state-feedback linearization of robot dynamics is performed by factoring the manipulator mass matrix into a quadratic form involving an integrable root matrix. The applicability of the proposed general design methodology is illustrated by an example of flexible joint planar robots. Numerical results for a two-link flexible joint planar robot are also provided. </para>

A switching control scheme is described for flexible joint robots which does not use link jerk or acceleration feedback. The control scheme does not require the numerical differentiation of the velocity signal or the inversion of the inertial matrices. Simulations are presented to verify the validity of the control scheme and its superiority over existing rigid robot controllers when applied to flexible joint robots. >

In paper ‘Robustness of Adaptive Control of Robots’ by Ghorbel and Spong, the adaptive control of flexible joint robots is investigated using a singular perturbation approach and composite Lyapunov theory. In this note, we correct a missing term from the boundary-layer system of that paper and then show that the stability analysis presented remains valid.

Real-time parallel adaptive neural network control for nonlinear servomechanisms—an approach using direct adaptive techniques

An adaptive neural network controller is presented for flexible link robots (FLRs) using the singular perturbation technique. The complex full model of FLRs can be decomposed into two separate time scale subsystems by modelling the elastic forces as the fast variables and the joint variables as the slow variables. A composite control strategy is adopted to control the two-time scale model with the slow subsystem being controlled by a neural network controller. The weights of the NNs are updated online based on direct adaptive techniques. A robust control term is also added for closed-loop stability. The fast subsystem is stabilized by a simple LQR control around the equilibrium trajectory defined by the slow subsystem.

Continuous sliding mode disturbance rejection control for flexible joint robots with input saturation.

On the control of flexible joint robots by dynamic output feedback

This paper presents a novel multivariable two-time scale control method for flexible joint robots. The fundamental limitations of the conventional two-time scale control are clearly explained based on a singular perturbation analysis with a linearized dynamic model. Unlike the conventional methods, the bandwidth of the slow control can be made high and the derivatives of the rigid control are not required. The closed-form solution of a slow model that is more accurate than the rigid model is presented. It is shown that the modeling error of the proposed two-time scale control can be made arbitrarily small by the control gain only. A real-world experiment with a six degrees-of-freedom commercial industrial robot was performed to validate the presented approach.

In this note, robust adaptive neural network (NN) control scheme is constructed for a class of unknown nonlinear systems with drift terms. The robust adaptive NN control laws are developed using backstepping technique which does not require the unknown parameters to be linear parametrizable and no regression matrices are needed. All the signals in the resulting closed-loop system are proved to be ultimately uniform bounded, and the system states are guaranteed to converge to zero.Keywordsadaptive controlbacksteppingneural networknonlinear system

The paper addresses the problem of controlling the joints of a flexible joint robot with a state feedback controller and proposes a gradual way of extending such a controller towards the complete decoupling of the robot dynamics. The global asymptotic stability for the state feedback controller with gravity compensation is proven, followed by some theoretical remarks on its passivity properties. By proper parameterization, the proposed controller structure can implement a position, a stiffness or a torque controller. Experimental results on the DLR lightweight robots validate the method.

Synchronization of networks composed by fully-actuated Lagrangian systems has received a lot of attention from the control theory community. Unfortunately, the case of networks composed by under-actuated systems of the same kind has not been deeply studied and the related literature is very reduced. The aim of this paper is to contribute towards the establishment of a control scheme for this class of networks by solving the particular case of a network composed by agents defined by Flexible-joint robots. The proposed schemes consider the presence of unknown delays in the communication channels. The usefulness of the controllers is validated through numerical simulations.