Accurate position-force control is a core and challenging problem in robotics, especially for manipulators with redundant DOFs. For example, trajectory tracking based control usually fails for grinding robots due to intolerable impact forces imposed onto the end-effectors. The main difficulties lie in the coupling of motion and contact force, redundancy resolution and physical constraints, etc. In this paper, we propose a novel motionforce control strategy in the framework of projection recurrent neural networks. Tracking error and contact force are described in orthogonal spaces respectively, and by selecting minimizing joint torque as secondary task, the control problem is formulated as a quadratic-programming (QP) problem under multiple constraints. In order to obtain real-time optimization of joint toque which is non-convex relative to joint angles, the original QP is reconstructed in velocity level, where the original objective function is replaced by its time derivative. Then a dynamic neural network which is convergence provable is established to solve the modified QP problem online. This work generalizes projection recurrent neural network based position control of manipulators to that of position-force control, which opens a new avenue to shift position-force control of manipulators from pure control perspective to cross design with both convergence and optimality consideration. Numerical and experimental results show that the proposed scheme achieves accurate position-force control, and is capable of handling inequality constraints such as joint angular, velocity and torque limitations, simultaneously, consumption of joint torque can be decreased effectively.

Motion planning is a core issuein the field of robotic control, which directly affects the programming efficiency of robots. In this article, we study the motion planning problem of manipulators for simultaneous obstacle avoidance and target tracking, and propose a novel real-time planning method in a complex workspace. One important feature of the proposed method is that the robot can avoid colliding with obstacles by easily defining “virtual fences,” which are described by a group of level set functions. Thus, the feasible space can be abstracted as inequality constraints. Taking the predefined task, physical constraints, and feasible space constraints into consideration, the motion planning problem is formulated into a quadratic programming (QP) one, in which the redundant degrees of freedom are used to optimize the velocities of the robot. Then, the control command is obtained by an established recurrent neural network, which is capable of solving the QP problem in an online manner. Theoretical conduction and verification in several typical workspaces demonstrate the efficacy of the established method, such as the abilityto remove physical fences, quick rearrangement, and performance optimization.

Collision avoidance plays a major part in the control of the wheeled mobile robot (WMR). Most existing collision-avoidance methods mainly focus on a single WMR and environmental obstacles. There are few products that cast light on the collision-avoidance between multiple WMRs (MWMRs). In this article, the problem of simultaneous collision-avoidance and target tracking is investigated for MWMRs working in the shared environment from the perspective of optimization. The collision-avoidance strategy is formulated as an inequality constraint, which has proven to be collision free between the MWMRs. The designed MWMRs control scheme integrates path following, collision-avoidance, and WMR velocity compliance, in which the path following task is chosen as the secondary task, and collision-avoidance is the primary task so that safety can be guaranteed in advance. A Lagrangian-based dynamic controller is constructed for the dominating behavior of the MWMRs. Combining theoretical analyses and experiments, the feasibility of the designed control scheme for the MWMRs is substantiated. Experimental results show that if obstacles do not threaten the safety of the WMR, the top priority in the control task is the target track task. All robots move along the desired trajectory. Once the collision criterion is satisfied, the collision-avoidance mechanism is activated and prominent in the controller. Under the proposed scheme, all robots achieve the target tracking on the premise of being collision free.

Robot Learning from humans is a promising paradigm for directly transferring human skills to robots. This learning allows robots to encapsulate task constraints and motion patterns from human demonstrations as well as acquire skills that can be adapted to unseen scenarios. Even though many state-of-the-art skill-learning successes have been achieved, simultaneously addressing variability from a complex and long-horizon manipulation task and generalizing it to external uncertainty remains a challenge. This efficient skill learning has to allow for handling large-scale, high-dimensional demonstrations, adapting to environmental changes (starting, via and end points and obstacles), generalizing to task constraints (trajectory precision, stiffness), and measuring uncertainty in the reproduction. To this effect, we present a novel robot skill-learning framework called SVGP-CoGP that will implement all the aforementioned properties by encoding task variability from multiple demonstrations using Sparse Variational Gaussian Processes (SVGP) and adapting to additional constraints via a coregionalized multi-output GP (CoGP) based on SVGP. The proposed method can significantly reduce the computational complexity of model fitting by making use of the variational inference of GP models, which makes it possible for robots to learn skills from complex and long-horizon tasks. We evaluated and compared the effectiveness and strengths of our framework with existing probabilistic methods on a Kinova robot that performed emergency button-pressing tasks. The results indicated that our framework allowed the robot to learn skills from complex and long-horizon manipulation tasks that outperformed baselines both in quantitative evaluation and in an online test.