Neural network learning control of robot manipulators using gradually increasing task difficulty

Learning control for robotic manipulators has been developed over the past decade and to the best of the authors’ knowledge, it is still in its infant development stage; the authors believe that it will become one of the most promising directions in the control area in robotic manipulators. Learning control in robotic manipulators is mainly used to address the issue that the friction at the joints of robotic mechanisms and other uncertainties may exist in the dynamic models, which are very complex and may even be impossible to model mathematically. In this paper, the authors review and discuss the learning control in robotic manipulators and some issues in learning control for robotic manipulators are also illustrated. This review is able to give a general guideline for future research in learning control for robotic manipulators.

A discrete-time learning control for robotic manipulators is studied using its pulse transfer function. Firstly, the discrete-time learning stability condition which is applicable to single-input two-outputs systems is derived. Secondly, the stability of the learning algorithm with a position signal is studied. In this case, when the sampling period is small, the algorithm is not stable because of an unstable zero of the system. Thirdly, the stability of the algorithm with position and velocity signals is studied. In this case, we can stabilize the learning control system which is unstable in the case of learning with only the position signal. Finally, simulation results on the trajectory control of robotic manipulators using the discrete-time learning control are shown. The simulation results agree well with the analytical results.

Further results on adaptive iterative learning control of robot manipulators

Accurate estimation of inverse kinematics and dynamics is essential for precise robotic manipulator control. Traditional analytical models, while effective for nominal conditions, often struggle to capture complex nonlinearities and unmodeled dynamics, necessitating compensatory controllers. However, these combined approaches can become computationally intensive and prone to redundancy. Conventional machine learning methods have shown promise in error compensation but typically demand large datasets and incur high computational costs, particularly when addressing high-dimensional state spaces. To overcome these limitations, this paper proposes a novel data-efficient framework based on Liquid Neural Networks (LNNs) for learning inverse kinematics and dynamics directly from simulation platform measurements. The inherent capability of LNN to model temporal dependencies and adapt continuously to evolving system dynamics enables accurate approximation of nonlinear manipulator behavior without reliance on extensive offline training. The proposed method is validated on two distinct robotic platforms—a UR5e manipulator with 6 DoF and a WAM manipulator with 7 DoF using both simulation and real-world trajectory data acquired through a targeted sampling strategy. Both simulation and experimental evaluations with previously unseen trajectories demonstrate that the ILNN-based approach achieves high-fidelity approximation of both inverse kinematics and inverse dynamics, delivering superior generalization and robustness compared to conventional analytical and data-driven methods. These results highlight the potential of LNNs as an efficient and scalable solution for real-time robot control in dynamic and uncertain environments.

Adaptive learning control of robot manipulators via incremental hybrid neural network

A discrete-time learning control for robotic manipulators is studied from a viewpoint of zeros of a control system. The signal used for learning is a position signal of each joint detected by a pulse encoder. Firstly, we show that when sampling period is small, the learning is unstable because of an unstable zero of the system. Secondly, we show that we can stabilize the learning control system with small sampling period by using a low pass filter, and also show that a tracking error converges not to zero but to a certain value in that case. Finally, experimental results on the trajectory control of robotic manipulators using the discrete-time learning control are shown. These experimental results agree well with the analytical ones.

Although model-free reinforcement learning methods are illustrated to own abilities to learn a wide range of control and decision-making tasks, it is time-consuming and needs a huge number of samples to learn optimal policy. Therefore, model-based reinforcement learning algorithms are always treated as the most efficient ways to settle robotic control problems. In this work, a model-based reinforcement learning tracking control method with free energy is proposed. The control objective is to make sure that the proposed controller can let a robotic manipulator with parameters uncertainties and disturbance follow the desired trajectory and obtain good tracking performance. The main contribution of this work is to employ the free energy principle to optimize the problem of robotic control. The probability distribution of robotic dynamic model is learnt by a evolving neural network, a free energy based reinforcement learning control method uses the learnt model to find optimal control policy. The experiment results demonstrate that the proposed tracking control method is effective in solving robotic tracking control problems.

This paper presents a robust learning control algorithm which learns the entire span of robot trajectory within a finite time interval. The learning controller treats the uncertain parameters as well as unknown external disturbances with the aid of the linear parameterization property of the robot system and robust feedback control input. It is shown that the robot motion converges exponentially to the desired one as the iteration continues.

Adaptive iterative learning control for robot manipulators: Experimental results

Discrete iterative learning control of robotic manipulators

In this paper, we propose an error-tracking iterative learning control scheme to tackle the position tracking problem for robot manipulators with random initial errors and iteration-varying reference trajectories. Different from general usual ones, the control strategy in our work is to drive system errors perfectly track the desired error trajectories over the whole time interval as the iteration number increases, by which, the position trajectory and velocity trajectory can respectively track their reference trajectories during the predefined part operation interval. For fulfilling the control design, a new construction method of desired error trajectories is presented to remove the perfect initial resetting condition, which must be satisfied in most traditional iterative learning control algorithms. The uncertainties and disturbances in the robotic system dynamics are compensated by the robust approach and iterative learning approach, combinedly.

A real-time Machine Learning Control (MLC) of articulated robotic manipulators is presented in this work by exploiting the Lion Optimization Algorithm (LOA) which is combined with the Naïve Bayes (LNB).Here, the proposed model is used for the real-time MLC of robotic manipulators by exploiting the "Fractional Order Proportional-Integral-Derivative (FOPID)" control scheme.Using the proposed LNB model MLC control gain parameters are tuned.By considering the dynamics of the actuator, to convene significant timing constraints, a "Real-Time Operating System (RTOS)" on a microprocessor collaborates with the LNB-MLC.At last, the Mechatronic design and investigational setup of a 6-degree of freedom (DOF) expressed robotic manipulator are modeled.The performance analysis is presented to show the advantage of the adopted models.While compared with the existing control models, the adopted optimization model has practice and theoretical consequences regarding online parameter tuning, real-time ability, and convergent behavior.In both industry and academia, the adopted MLC approaches are appropriate to design the real-time modern controllers.

Learning control algorithms have been proposed for error compensation in repetitive robotic manipulator tasks. It is shown that the performance of such control algorithms can be seriously degraded when the feedback data they use is relatively sparse in time, such as might be provided by vision systems. It is also shown that learning control algorithms can be modified to compensate for the effects of sparse data and thereby yield performance which approaches that of systems without limitations on the sensory information available for control.

In this paper, a hybrid neural fuzzy control scheme is proposed for the control of flexible-joint robot manipulators with unknown dynamics. The control strategy is based on a feedforward artificial neural network to partially approximate the manipulator's inverse dynamics. A fuzzy sliding mode feedback controller is also used for the online adaptation of the neural network-based controller. Simulation results of various scenarios highlight the performance and stability of the proposed controller in compensating for the highly nonlinear unknown dynamics of the manipulator under different dynamical conditions and external disturbances

Perturbation-dependent selection of postural feedback gain and its scaling