In this paper, we expolore Multi-Agent Reinforcement Learning (MARL) methods for unmanned aerial vehicle (UAV) cluster. Considering that the current UAV cluster is still in the program control stage, the fully autonomous and intelligent cooperative combat has not been realised. In order to realise the autonomous planning of the UAV cluster according to the changing environment and cooperate with each other to complete the combat goal, we propose a new MARL framework. It adopts the policy of centralised training with decentralised execution, and uses Actor-Critic network to select the execution action and then to make the corresponding evaluation. The new algorithm makes three key improvements on the basis of Multi-Agent Deep Deterministic Policy Gradient (MADDPG) algorithm. The first is to improve learning framework; it makes the calculated Q value more accurate. The second is to add collision avoidance setting, which can increase the operational safety factor. And the third is to adjust reward mechanism; it can effectively improve the cluster’s cooperative ability. Then the improved MADDPG algorithm is tested by performing two conventional combat missions. The simulation results show that the learning efficiency is obviously improved, and the operational safety factor is further increased compared with the previous algorithm.
This article deals with a human–machine cooperative system for the control of a smart wheelchair for people with motor disabilities. The choice of a traded control mode is first argued. The paper then pursues two objectives. The first is to describe the design of the cooperative system by focusing on the dialogue and the interaction between the pilot and the robot. The second objective is to introduce a new cooperative mode. In this one, three features are proposed: two semi-autonomous features, a wall following and a doorway crossing, during which the user can intervene punctually to rectify a trajectory or a path, and an assisted mode where, conversely, the machine intervenes in a manual control to avoid obstacles. This mode of intervention of an entity, human or machine, supervising a movement controlled by the other is referred as “combined control.” Examples of scenarios exploiting the cooperative capabilities of the system are presented and discussed.
Around the cooperative path-following control for the underactuated surface vessel (USV) and the unmanned aerial vehicle (UAV), a logic virtual ship-logic virtual aircraft (LVS-LVA) guidance principle is developed to generate the reference heading signals for the USV-UAV system by using the “virtual ship” and the “virtual aircraft”, which is critical to establish an effective correlation between the USV and the UAV. Taking the steerable variables (the main engine speed and the rudder angle of the USV, and the rotor angular velocities of the UAV) as the control input, a robust adaptive neural cooperative control algorithm was designed by employing the dynamic surface control (DSC), radial basic function neural networks (RBF-NNs) and the event-triggered technique. In the proposed algorithm, the reference roll angle and pitch angle for the UAV can be calculated from the position control loop by virtue of the nonlinear decouple technique. In addition, the system uncertainties were approximated through the RBF-NNs and the transmission burden from the controller to the actuators was reduced for merits of the event-triggered technique. Thus, the derived control law is superior in terms of the concise form, low transmission burden and robustness. Furthermore, the tracking errors of the USV-UAV cooperative control system can converge to a small compact set through adjusting the designed control parameters appropriately, and it can be also guaranteed that all the signals are the semi-global uniformly ultimately bounded (SGUUB). Finally, the effectiveness of the proposed algorithm has been verified via numerical simulations in the presence of the time-varying disturbances.
This paper studies the precise position control of the hydraulic cylinder in the hydraulic support. The aim of this paper is to develop a method of hydraulic cylinder position control based on pressure and flow coupling, which takes the coupling feedback of load and flow into account, especially in the scene of cooperative control under the condition of multiple actuators and variable load. This method solves the problems of slow movement and sliding effect of hydraulic support in the traditional time-dependent hydraulic position control, as well as better realizes the intelligent and unmanned development of the fully mechanized mining face. First, based on the flow continuity equation and Newton Euler dynamic equation, the flow and stroke control model with the input and output pressure of hydraulic cylinder is established. Then, the effectiveness and correctness of the control model are verified by the comparison between the hydraulic system simulation software, AMESim, and the experiment. Finally, a test system is built. When the system pressure is large than 10 MPa, the error between the data determined by the fitting algorithm and the actual detection data is within 5%, which verifies the effectiveness of the theory and simulation model.
In order to improve the position tracking precision of dual permanent magnet synchronous motor (PMSM) systems, a unified nonlinear predictive control (UNPC) strategy based on the unified modeling of two PMSM systems is proposed in this paper. Firstly, establishing a unified nonlinear model of the dual-PMSM system, which contains uncertain disturbances caused by parameters mismatch and external load changes. Then, the position contour error and tracking errors are regarded as the performance index inserted into the cost function, and the single-loop controller is obtained by optimizing the cost function. Meanwhile, the nonlinear disturbance observer is designed to estimate the uncertain disturbances, which is used for feed-forward compensation control. Finally, the proposed strategy is experimentally validated on two 2.3 kW permanent magnet synchronous motors, and the experimental results show that effectiveness and feasibility of proposed strategy.
Technology-supported rehabilitation therapy for neurological patients has gained increasing interest since the last decades. The literature agrees that the goal of robots should be to induce motor plasticity in subjects undergoing rehabilitation treatment by providing the patients with repetitive, intensive, and task-oriented treatment. As a key element, robot controllers should adapt to patients’ status and recovery stage. Thus, the design of effective training modalities and their hardware implementation play a crucial role in robot-assisted rehabilitation and strongly influence the treatment outcome. The objective of this paper is to provide a multi-disciplinary vision of patient-cooperative control strategies for upper-limb rehabilitation exoskeletons to help researchers bridge the gap between human motor control aspects, desired rehabilitation training modalities, and their hardware implementations. To this aim, we propose a three-level classification based on 1) “high-level” training modalities, 2) “low-level” control strategies, and 3) “hardware-level” implementation. Then, we provide examples of literature upper-limb exoskeletons to show how the three levels of implementation have been combined to obtain a given high-level behavior, which is specifically designed to promote motor relearning during the rehabilitation treatment. Finally, we emphasize the need for the development of compliant control strategies, based on the collaboration between the exoskeleton and the wearer, we report the key findings to promote the desired physical human-robot interaction for neurorehabilitation, and we provide insights and suggestions for future works.