Neural-Network-Based Sliding-Mode Control of an Uncertain Robot Using Dynamic Model Approximated Switching Gain

Disturbance Rejection Event-Triggered Robust Model Predictive Control for Tracking of Constrained Uncertain Robotic Manipulators

A novel hierarchical control framework combining computed-torque-like control (CTLC) with disturbance-observer-based event-triggered robust model predictive control (DO-ET-RMPC) is proposed for the trajectory tracking control of robotic manipulators with bounded disturbances and state and control input constraints. The CTLC approach is first used to cancel the exact nonlinear dynamics of the original tracking error system to obtain a set of decoupling linear tracking error subsystems, thus reducing the optimization complexity of model predictive control (MPC). The composite DO-ET-RMPC scheme is then developed based on the so-called dual-mode MPC approach to robustly stabilize the tracking error subsystems, which could improve the robustness of MPC and save its computational resources simultaneously. The continuous-time theoretical properties of the DO-ET-RMPC scheme, considering disturbances and state and control input constraints simultaneously, are provided for the first time, including the avoidance of Zeno behavior, robust constraint satisfaction, recursive feasibility, and stability. In the end, the superiorities of the proposed control scheme are verified by the comparative simulations.

Multi-stage trajectory tracking of robot manipulators under stochastic environments

For robot manipulators composed of Lagrange subsystems driven by direct current (DC) motors under stochastic environments, multi-stage trajectory tracking is investigated in this paper. The main challenge is how to achieve the end-effector drive of manipulators from a given initial state to a final state. First, the inverse kinematics method and the partition of the task space are adopted to tackle multi-stage trajectory planning. Second, the adaptive backstepping technique is used to design tracking controller for stochastic Lagrangian subsystems. Then, based on the state-dependent switching signal, a multi-stage switched controller is designed for trajectory tracking of robot manipulators. All signals in the close-loop error switched system are bounded in probability, and the tracking error in mean square can be made arbitrarily small enough by parameters-tuning The effectiveness of the proposed control method is illustrated by simulation results.

Bidirectional recurrent learning of inverse dynamic models for robots with elastic joints: a real-time real-world implementation

Collaborative robots, or cobots, are designed to work alongside humans and to alleviate their physical burdens, such as lifting heavy objects or performing tedious tasks. Ensuring the safety of human-robot interaction (HRI) is paramount for effective collaboration. To achieve this, it is essential to have a reliable dynamic model of the cobot that enables the implementation of torque control strategies. These strategies aim to achieve accurate motion while minimizing the amount of torque exerted by the robot. However, modeling the complex non-linear dynamics of cobots with elastic actuators poses a challenge for traditional analytical modeling techniques. Instead, cobot dynamic modeling needs to be learned through data-driven approaches, rather than analytical equation-driven modeling. In this study, we propose and evaluate three machine learning (ML) approaches based on bidirectional recurrent neural networks (BRNNs) for learning the inverse dynamic model of a cobot equipped with elastic actuators. We also provide our ML approaches with a representative training dataset of the cobot's joint positions, velocities, and corresponding torque values. The first ML approach uses a non-parametric configuration, while the other two implement semi-parametric configurations. All three ML approaches outperform the rigid-bodied dynamic model provided by the cobot's manufacturer in terms of torque precision while maintaining their generalization capabilities and real-time operation due to the optimized sample dataset size and network dimensions. Despite the similarity in torque estimation of these three configurations, the non-parametric configuration was specifically designed for worst-case scenarios where the robot dynamics are completely unknown. Finally, we validate the applicability of our ML approaches by integrating the worst-case non-parametric configuration as a controller within a feedforward loop. We verify the accuracy of the learned inverse dynamic model by comparing it to the actual cobot performance. Our non-parametric architecture outperforms the robot's default factory position controller in terms of accuracy.

Neural network-based adaptive second-order sliding mode control for uncertain manipulator systems with input saturation

In order to solve the trajectory tracking problem for robotic manipulators with dynamic uncertainty, external disturbance and input saturation, a novel second-order sliding mode control scheme based on neural network is proposed in this paper. First of all, a model-based second-order non-singular fast terminal sliding mode controller (SONFTSMC) is designed to overcome the chattering problem under the consideration of uncertain parameters. Then attention is focused on the scenario that all those nonlinear uncertainties are unknown, and a new fuzzy wavelet neural network (FWNN) is designed to estimate those unknown uncertainties via lumping them into one compounded uncertainty. In addition, all parameters in FWNN are adjusted autonomously by using an adaptive method. The proposed second-order non-singular fast terminal sliding mode (SONFTSM) control method not only improves the convergence speed and tracking accuracy of the robotic manipulator, but also enhances its robustness. Finally, the advantages of SONFTSM control strategy over existing sliding mode control methods are verified with comparative simulations.

Enhanced compensation control of pneumatic muscle actuator with high-order modified dynamic model

Dynamic behaviour of the pneumatic muscle actuator (PMA) is conventionally modelled as a pressure-based first-order equation under discrete loads, which cannot exactly describe its dynamic features. Considering PMA's nonlinear, time-varying and hysteresis characteristics, we propose a novel high-order modified dynamic model of PMA based on its physical properties and working principle, with coefficients being identified under external dynamic loads. To tackle PMA's nonlinear hysteresis problem in high-frequency movements, a global fast terminal sliding mode controller with the modified model-based radial basis function (RBF) neural network disturbance compensator (RBF-GFTSMC) is designed. Comparison experimental studies are carried on a designed PMA platform that can provide continuously changing loads. Results show that the RBF-GFTSMC has superior trajectory tracking performance and disturbance compensation capability under wide-ranged frequencies and external loads, which can be potentially used to achieve precise control of PMA-actuated robots.

A Bus-Scheduling Method Based on Multi-Sensor Data Fusion and Payment Authenticity Verification

It is of great significance to ensure public transportation management capabilities by improving urban public transport services. One method is to solve the problems related to the quality of data submitted for public funding as well as the accuracy and transparency of the supervision and review processes; moreover, improving public-transportation-service systems is a viable method to solve such problems. With technological advancements and the application of new technologies such as automatic driving and multiple payment, it has gradually become difficult for user-data verification systems, based on the original single bus payment method, to cater to these new technologies. Diversified payment and complex management methods have highlighted the need for new verification methods. Firstly, in this paper, we constructed the Origin–Destination (OD) model of bus-passenger flows by using real-time transmission of passenger-multiple-payment data, on-board-video passenger flow detection data and vehicle real-time positioning data. On this basis, the bus waybill data of other intelligent bus systems and the wait data of bus stations were integrated, so as to establish the travel chain theory by matching passenger flow and the temporal and spatial distribution model. Then, an OD analysis of public-transport passenger flows could be carried out, with a detailed analysis of vehicle, station and line-passenger flow, and the travel characteristics of public transport passenger flow could be excavated. Then, according to the means-end chain theory, the spatiotemporal distribution of the passenger flow data was obtained to carry out an OD analysis of the passenger flow, so as to perform a refinement analysis of the vehicle, station, and passenger flow. Thereby, the characteristics of the passenger flow were explored. Subsequently, payment-authenticity-verification models were established for the data-validity assessment, video-data analysis, passenger-flow estimation, and early warnings in order to determine the authenticity of the payment data. Lastly, based on the multi-sensor passenger flow data fusion and the data authenticity verification models, combined with the application of new technologies such as the use of autonomous buses, the test was promoted. That is, by taking intelligent bus scheduling as the scenario, the research method was tested and verified with real-time passenger flow data according to historical data. The results showed that the method accurately predicted the passenger flow, and had a positive role in improving the efficiency of payment-data-authenticity verification. The application of the method can enhance the management and service quality of public transportation.

Intelligent Bus Scheduling Control Based on On-Board Bus Controller and Simulated Annealing Genetic Algorithm

The stable and fast service of a bus network is one of the important indicators of the service quality and management level of urban public transport. With the continuous expansion of cities, the bus network complexity has been increasing accordingly. The application of new technologies such as self-driving buses has made the bus network more complex and its vulnerability more obvious. Therefore, how to collect information on passenger flow, traffic flow, and transport distribution using intelligent means, and how to establish an effective intelligent bus scheduling control method have been important questions surrounding the improvement of the level of urban bus operation. To address this challenge, this paper proposes the design method of a bus controller based on data collection and the edge computing requirements of autonomous driving buses; and installs them widely on buses. In addition, an intelligent bus control scheduling method based on the simulated annealing genetic algorithm was developed according to the current scheduling requirements. The proposed method combines the strong local search ability of the simulated annealing algorithm, which prevents the search process from falling into a local optimum, and the strong search ability of the genetic algorithm in the overall search process, leading an intelligent bus control scheduling method based on the simulated annealing genetic algorithm. The proposed method was verified by experiments on the optimal scheduling of multi-destination public transport as an example, we verified the research method, and finally, simulated it using historical data. There is good model prediction of the experimental results. Therefore, the intelligent traffic control can be realized through efficient bus scheduling, thus improving the robustness of the bus network operation.

IMU Motion Capture Method with Adaptive Tremor Attenuation in Teleoperation Robot System

Teleoperation robot systems can help humans perform tasks in unstructured environments. However, non-intuitive control interfaces using only a keyboard or joystick and physiological tremor reduce the performance of teleoperation. This paper presents an intuitive control interface based on the wearable device gForcePro+ armband. Two gForcePro+ armbands are worn at the centroid of the upper arm and forearm, respectively. Firstly, the kinematics model of the human arm is established, and the inertial measurement units (IMUs) are used to capture the position and orientation information of the end of the arm. Then, a regression model of angular transformation is developed for the phenomenon that the rotation axis of the torsion joint is not perfectly aligned with the limb segment during motion, which can be applied to different individuals. Finally, to attenuate the physiological tremor, a variable gain extended Kalman filter (EKF) fusing sEMG signals is developed. The described control interface shows good attitude estimation accuracy compared to the VICON optical capture system, with an average angular RMSE of 4.837° ± 1.433°. The performance of the described filtering method is tested using the xMate3 Pro robot, and the results show it can improve the tracking performance of the robot and reduce the tremor.

Multi-Agent Reinforcement Learning with Optimal Equivalent Action of Neighborhood

In a multi-agent system, the complex interaction among agents is one of the difficulties in making the optimal decision. This paper proposes a new action value function and a learning mechanism based on the optimal equivalent action of the neighborhood (OEAN) of a multi-agent system, in order to obtain the optimal decision from the agents. In the new Q-value function, the OEAN is used to depict the equivalent interaction between the current agent and the others. To deal with the non-stationary environment when agents act, the OEAN of the current agent is inferred simultaneously by the maximum a posteriori based on the hidden Markov random field model. The convergence property of the proposed methodology proved that the Q-value function can approach the global Nash equilibrium value using the iteration mechanism. The effectiveness of the method is verified by the case study of the top-coal caving. The experiment results show that the OEAN can reduce the complexity of the agents’ interaction description, meanwhile, the top-coal caving performance can be improved significantly.

Chattering-Suppressed Sliding Mode Control for Flexible-Joint Robot Manipulators

In this paper, sliding mode tracking control and its chattering suppression method are investigated for flexible-joint robot manipulators with only state measurements of joint actuators. First, within the framework of singular perturbation theory, the control objective of the system is decoupled into two typical tracking aims of a slow subsystem and a fast subsystem. Then, considering lumped uncertainties (including dynamics uncertainties and external disturbances), a composite chattering-suppressed sliding mode controller is proposed, where a smooth-saturation-function-contained reaching law with adjustable saturation factor is designed to alleviate the inherent chattering phenomenon, and a radial basis function neural network (RBFNN)-based soft computing strategy is applied to avoid the high switching gain that leads to chattering amplification. Simultaneously, an efficient extended Kalman filter (EKF) with respect to a new state variable is presented to enable the closed-loop tracking control with neither position nor velocity measurements of links. In addition, an overall analysis on the asymptotic stability of the whole control system is given. Finally, numerical examples verify the superiority of the dynamic performance of the proposed control approach, which is well qualified to suppress the chattering and can effectively eliminate the undesirable effects of the lumped uncertainties with a smaller switching gain reduced by 80% in comparison to that in the controller without RBFNN. The computational efficiency of the proposed EKF increased by about 26%.