Mobile manipulation has a broad range of applications in robotics. However, it is usually more challenging than fixed-base manipulation due to the complex coordination of a mobile base and a manipulator. Although recent works have demonstrated that deep reinforcement learning is a powerful technique for fixed-base manipulation tasks, most of them are not applicable to mobile manipulation. This paper investigates how to leverage deep reinforcement learning to tackle whole-body mobile manipulation tasks in unstructured environments using only on-board sensors. A novel mobile manipulation system which integrates the state-of-the-art deep reinforcement learning algorithms with visual perception is proposed. It has an efficient framework decoupling visual perception from the deep reinforcement learning control, which enables its generalization from simulation training to real-world testing. Extensive simulation and experiment results show that the proposed mobile manipulation system is able to grasp different types of objects autonomously in various simulation and real-world scenarios, verifying the effectiveness of the proposed mobile manipulation system.

This paper deals with position tracking control of a single-rod electro-hydraulic actuator subject to external disturbances and parameter uncertainties. In previous disturbance observer design methodologies for electro-hydraulic actuators, parameter uncertainties have been commonly regarded as disturbances and lumped together with external perturbations. However, in practical electro-hydraulic systems, system parameters are unknown and varying. If considerable parameter uncertainties exist in the system or if the disturbance dynamics induced by parameter uncertainties exceed the bandwidth of the disturbance observer, estimation accuracy will degrade, which will significantly affect system performance. To solve this problem, an extended disturbance observer is proposed in this paper to estimate disturbances while dealing with parameter uncertainties. In addition, a nonlinear position tracking controller is designed for position tracking based on the proposed disturbance observer using a backstepping technique. The proof of the stability of the overall closed-loop system is based on Lyapunov theory. The performance of the proposed controller is verified through simulations and experiments using a shock absorber as a load force generator. A detailed nonlinear physical model of the load force is developed and implemented in the simulation. The results show that the proposed nonlinear position tracking controller, together with the extended disturbance observer, provide excellent tracking performance in the presence of parameter uncertainties and external disturbances.

For many electro-hydraulic pump-controlled systems, supply pressure control of the hydraulic pump is of great importance. However, the control performance is significantly affected by unknown time-varying load flow requirements. To improve the supply pressure tracking performance in the presence of unknown time-varying load flow disturbances, a nonlinear controller is designed based on the control-oriented mathematical model presented in this article. First, a disturbance observer is used to estimate the unknown time-varying load flow; a nonlinear feedforward controller is then derived using the differential flatness property of the system based on the estimated load flow; in addition, a feedback controller is implemented to stabilize the system. Sliding mode control is also used to compensate for load flow estimation error. The stability of the whole system is proved using Lyapunov theory. Experimental results demonstrate that the proposed control strategy has good supply pressure tracking performance.

No abstract