Search

When a robot performs tasks such as assembly or human–robot interaction, it is inevitable for it to collide with the unknown environment, resulting in potential safety hazards. In order to improve the compliance of robots to cope with unknown environments and enhance their intelligence in contact force-sensitive tasks, this paper proposes an improved admittance force control method, which combines classical adaptive control and machine learning methods to make them use their respective advantages in different stages of training and, ultimately, achieve better performance. In addition, this paper proposes an improved Deep Deterministic Policy Gradient (DDPG)-based optimizer, which is combined with the Gaussian process (GP) model to optimize the admittance parameters. In order to verify the feasibility of the algorithm, simulations and experiments are carried out in MATLAB and on a UR10e robot, respectively. The experimental results show that the algorithm improves the convergence speed by 33% in comparison to the general model-free learning method, and has better control performance and robustness. Finally, the adjustment time required by the algorithm is 44% shorter than that of classical adaptive admittance control. [ABSTRACT FROM AUTHOR]

In robotics, active exploration and learning in uncertain environments must take into account safety, as the robot may otherwise damage itself or its surroundings. This paper presents a method for safe active search using Bayesian optimization and control barrier functions. As robot paths undertaken during sampling are continuous, we consider an informative continuous expected improvement acquisition function. To safely bound the contact forces between the robot and its surroundings, we leverage exponential control barrier functions, utilizing the derivative of the force in the contact model to increase robustness to uncertainty in the contact boundary. Our approach is demonstrated on a fully autonomous robot for ultrasound scanning of rheumatoid arthritis (RA). Here, active search is a critical component of ensuring high image quality. Furthermore, bounded contact forces between the ultrasound probe and the patient ensure patient safety and better scan quality. To the best of our knowledge, our results are both the first demonstration of safe active search on a fully autonomous robot for ultrasound scanning of rheumatoid arthritis and the first experimental evaluation of bounding contact forces in the context of medical robotics using control barrier functions. The results show that when search time is limited to less than 60 s, informative continuous expected improvement leads to a 92% success, a 13% improvement compared to expected improvement. Meanwhile, exponential control barrier functions can limit the force applied by the robot to under 5 N, even in cases where the contact boundary is specified incorrectly by -1 or +4 mm. [ABSTRACT FROM AUTHOR]

To address the problem that traditional force control methods have difficulty obtaining stable forces during a robot massage, a robot force control algorithm based on residual reinforcement learning is proposed. An initial strategy of the robot massage force is first constructed with impedance control, but the massage force often fluctuates when the skin environment is unknown to the robot. A reinforcement learning algorithm is then used to analyze the relationship between the robot contact state and offset displacement and compensate for the residuals of the impedance controller. To speed up the search for the compensation strategy, a neural network is constructed to fit a dynamic model of reinforcement learning with the data from the initial strategy, which can simulate the contact between the robot and skin. Offline training in a simulated environment can reduce the number of actual interactions in reinforcement learning and improve the practicability of the algorithm; to integrate the two algorithms, the output of the residual reinforcement learning strategy is smoothed. The experimental results show that the robot force control algorithm based on residual reinforcement learning converges after approximately 80 offline iterations. The force error in the online experiment is basically within ±0.2 N.

When a robot performs tasks such as assembly or human–robot interaction, it is inevitable for it to collide with the unknown environment, resulting in potential safety hazards. In order to improve the compliance of robots to cope with unknown environments and enhance their intelligence in contact force-sensitive tasks, this paper proposes an improved admittance force control method, which combines classical adaptive control and machine learning methods to make them use their respective advantages in different stages of training and, ultimately, achieve better performance. In addition, this paper proposes an improved Deep Deterministic Policy Gradient (DDPG)-based optimizer, which is combined with the Gaussian process (GP) model to optimize the admittance parameters. In order to verify the feasibility of the algorithm, simulations and experiments are carried out in MATLAB and on a UR10e robot, respectively. The experimental results show that the algorithm improves the convergence speed by 33% in comparison to the general model-free learning method, and has better control performance and robustness. Finally, the adjustment time required by the algorithm is 44% shorter than that of classical adaptive admittance control.

To address traditional impedance control methods' difficulty with obtaining stable forces during robot-skin contact, a force control based on the Gaussian mixture model/Gaussian mixture regression (GMM/GMR) algorithm fusing different compensation strategies is proposed. The contact relationship between a robot end effector and human skin is established through an impedance control model. To allow the robot to adapt to flexible skin environments, reinforcement learning algorithms and a strategy based on the skin mechanics model compensate for the impedance control strategy. Two different environment dynamics models for reinforcement learning that can be trained ofline are proposed to quickly obtain reinforcement learning strategies. Three different compensation strategies are fused based on the GMM/GMR algorithm, exploiting the online calculation of physical models and ofline strategies of reinforcement learning, which can improve the robustness and versatility of the algorithmwhen adapting to different skin environments. The experimental results show that the contact force obtained by the robot force control based on the GMM/GMR algorithm fusing different compensation strategies is relatively stable. It has better versatility than impedance control, and the force error is within ~±0.2 N. [ABSTRACT FROM AUTHOR]

In robotics, active exploration and learning in uncertain environments must take into account safety, as the robot may otherwise damage itself or its surroundings. This paper presents a method for safe active search using Bayesian optimization and control barrier functions. As robot paths undertaken during sampling are continuous, we consider an informative continuous expected improvement acquisition function. To safely bound the contact forces between the robot and its surroundings, we leverage exponential control barrier functions, utilizing the derivative of the force in the contact model to increase robustness to uncertainty in the contact boundary. Our approach is demonstrated on a fully autonomous robot for ultrasound scanning of rheumatoid arthritis (RA). Here, active search is a critical component of ensuring high image quality. Furthermore, bounded contact forces between the ultrasound probe and the patient ensure patient safety and better scan quality. To the best of our knowledge, our results are both the first demonstration of safe active search on a fully autonomous robot for ultrasound scanning of rheumatoid arthritis and the first experimental evaluation of bounding contact forces in the context of medical robotics using control barrier functions. The results show that when search time is limited to less than 60 s, informative continuous expected improvement leads to a 92% success, a 13% improvement compared to expected improvement. Meanwhile, exponential control barrier functions can limit the force applied by the robot to under 5 N, even in cases where the contact boundary is specified incorrectly by −1 or +4 mm.

Aiming at the transient overshoot, slow tracking response, and poor steady-state tracking accuracy during the dynamic contact between the robot and the physical environment, a robot impedance control strategy based on fractional order was proposed. First of all, the current method of integer-order impedance control contact process was analyzed, which only relies on changing the update rate and in-decreasing elastic items to adapt to different tasks. A more flexible fractional-order impedance control was designed. According to the inherent properties of fractional impedance, the gain coefficient was designed to map fractional impedance to integer order impedance to facilitate calculation and practical application, and some characteristics of improving integer order impedance control were introduced into fractional impedance to further improve the performance. The dynamically adjustable boundary was analyzed. The simulation results show that the fractional impedance proposed in this paper can directly reduce the contact force overshoot peak generated during the contact process, and has a certain inhibitory effect on the unstable oscillation behaviour of the system. It shows that the fractional impedance is suitable for the application of the contact interaction between robot and dynamic unknown environment.

In robotic polishing applications, the rotating polisher will generate strong vibration disturbance, which makes it more difficult to detect and control the contact force. This paper proposes a novel and practical method to detect and control the contact force using the built-in sensors (motor encoders and joint torque sensors) for actual robotic polishing in harsh conditions of strong vibration disturbance. An extended state observer based on the robot dynamic model is developed to estimate the contact force in real time, and a new and efficient adaptive filter combining insights of the notch filter with the tracking differentiator is designed to relieve the strong vibration disturbance of torque signals from the eccentrically rotating polisher. On the basis of efficient force detection, a hybrid position/force control method based on the inner joint torque controller is proposed to realize accurate force control in actual robotic polishing of curved surfaces. With the proposed contact force detection and control methodology, the robot is potential to achieve satisfied polishing applications in which some usual accessories, such as the six-axis force/torque sensor or Remote Center Compliance device are absent. Experimental results confirm the effectiveness and accuracy of the proposed contact force detection method in conditions of strong vibration disturbance from the rotating polisher. In addition, the curved surface polishing experiments indicate that the new hybrid position/force control framework performs well on rejecting the strong disturbances while maintaining high force control accuracy of polishing, and the quality of polished surface is greatly improved. [ABSTRACT FROM AUTHOR]

To address traditional impedance control methods' difficulty with obtaining stable forces during robot-skin contact, a force control based on the Gaussian mixture model/Gaussian mixture regression (GMM/GMR) algorithm fusing different compensation strategies is proposed. The contact relationship between a robot end effector and human skin is established through an impedance control model. To allow the robot to adapt to flexible skin environments, reinforcement learning algorithms and a strategy based on the skin mechanics model compensate for the impedance control strategy. Two different environment dynamics models for reinforcement learning that can be trained offline are proposed to quickly obtain reinforcement learning strategies. Three different compensation strategies are fused based on the GMM/GMR algorithm, exploiting the online calculation of physical models and offline strategies of reinforcement learning, which can improve the robustness and versatility of the algorithm when adapting to different skin environments. The experimental results show that the contact force obtained by the robot force control based on the GMM/GMR algorithm fusing different compensation strategies is relatively stable. It has better versatility than impedance control, and the force error is within ~±0.2 N., Competing Interests: The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest., (Copyright © 2024 Xiao, Zhang, Zhang, Chen, Zou and Wu.)

This paper presents a unified control framework for both set-point and time-varying force control of robot manipulator by introducing an improved position-based impedance control (IPBIC). In order to essentially achieve accurate force control, especially time-varying force tracking, a new target impedance function compensated by a force controller is presented. The essence of the improved method in realizing time-varying force tracking, as well as the coupled stability of the manipulator-environment system is investigated. To further improve the force control performance, the Newton-type iterative learning control (ILC) is introduced upon the closed-loop system. A case study on a two-link robot model demonstrates the effectiveness of this method. [ABSTRACT FROM AUTHOR]

In the paper attention is drawn to adverse dynamic properties of filters implemented in commercial measurement systems, force/torque sensors, which are increasingly used in industrial robotics. To remedy the problem, it has been proposed to employ a time-variant filter with appropriately modulated parameters, owing to which it is possible to suppress the amplitude of the transient response and, at the same time, to increase the pulsation of damped oscillations; this results in the improvement of dynamic properties in terms of reducing the duration of transients. This property plays a key role in force control and in the fundamental problem of the robot establishing contact with rigid environment. The parametric filters have been verified experimentally and compared with filters available for force/torque sensors manufactured by JR3. The obtained results clearly indicate the advantages of the proposed solution, which may be an interesting alternative to the classic methods of filtration. [ABSTRACT FROM AUTHOR]

This paper considers the problem of time-varying force control for robot manipulators in the presence of uncertainties from both the robotic model and working environment. The position-based impedance control (PBIC) method is employed and in order to achieve accurate time-varying force tracking, an improved PBIC is proposed. A neural-network-based robust controller is proposed to compensate for the system uncertainties, and an adaptive law is developed to identify the uncertain environmental parameters. Simulation results on a two-link robot manipulator confirm the effectiveness of the method in achieving time-varying force tracking. [ABSTRACT FROM AUTHOR]

This paper describes an approach to teaching concepts of automatic control, aimed principally at mechanical engineering students on courses that do not include a dedicated control module. The application area employed is robot force control, which has been the focus of much research during the past three or more decades. In this work, a robot is modelled in MATLAB and Simulink as a simple one degree-of-freedom ‘arm’, which is based on a velocity-controlled DC motor. The paper describes the development of this model and how it can be approximated as a standard second-order system from step response data. Three example solutions for designing a suitable force controller are then discussed, for a situation where the stiffness of contact at the robot/task interface is known and constant. The approach can be used in conjunction with a more traditional course in automatic control, or act as standalone tutorial for those with a more specific interest in force control. [ABSTRACT FROM AUTHOR]

In many applications involving a robot in contact with a surface it is important to control the interaction between the manipulator and its environment, usually by employing force sensors. However, sometimes it is desirable to remove them due to a variety of reasons, e.g. high costs, noisy measurements and a narrow bandwidth. To overcome these drawbacks, in this work it is proposed as a velocity/force observer based on the Generalized Proportional Integral (GPI) technique. Joint velocities and contact forces are estimated with only position measurements and then used in a force/position control scheme. Ultimate boundedness of the observation errors is formally proven and an arbitrarily small ultimate bound is then achieved. Simulation results are used to validate the proposed approach. [ABSTRACT FROM AUTHOR]

To precisely implement the force control of robot manipulators in an unknown environment, a control strategy based on fuzzy prediction of the reference trajectory in the impedance model is developed. The force tracking experiments are executed in an open-architecture control system with different tracking velocities, different desired forces, different contact stiffnesses and different surface figurations. The corresponding force control results are compared and analyzed. The influences of unknown parameters of the environment on the contact force are analyzed based on experimental data, and the tunings of predictive scale factors are illustrated. The experimental results show that the desired trajectory in the impedance model is predicted exactly and rapidly in the cases that the contact surface is unknown, the contact stiffness changes, and the fuzzy force control algorithm has high adaptability to the unknown environment. [ABSTRACT FROM AUTHOR]

The perception of touch opens new perspectives for so-called 'intelligent robotics'. Force/torque sensors are currently a key component of autonomous assembly processes or of the dynamically developing sector of collaborative robots. Response time is a critical parameter of force control, which has a direct effect on impact forces when the robot initiates contact with the environment. This paper indicates parameters of one of commercial force/torque sensors by JR3, in particular, its pre-defined low-pass filters. Their stationary nature introduces i.e. significant delay in the time domain, resulting in a negative impact on the overall dynamics of force control. To remedy the problem, our proposed approach is to employ a novel discrete time-variant filter with appropriately modulated parameters, owing to which it is possible to suppress the amplitude of the transient response and, at the same time, to increase the pulsation of damped oscillations; this results in the improvement of the dynamic properties in terms of reducing the duration of transients. Differences between a commercial, stationary filter and the recommended discrete time-variant filter have been shown experimentally, using a dedicated test environment. [ABSTRACT FROM AUTHOR]