Patients with hemiplegia are usually restricted to performing general bilateral activities of daily life (gbADLs). Bilateral training has been verified to contribute to the rehabilitation of physical functions. Although robotic systems are gradually being employed in the field of rehabilitation, few studies have performed simulations with regards to gbADLs for training. Therefore, a novel end-effector bilateral rehabilitation robotic system (EBReCS) for the upper limb is developed in this article for the task rendering of gbADLs, in which the gbADL-corresponding workspace is obtained via modularly designed bilateral parallelogram mechanisms, and interaction rendering of multiple bimanual modes (uncoupled, trans-soft-coupled, trans-semi-coupled, and rotation-coupled) is achieved by implementing the admittance model, the inner force field between robotic end-effectors, and the outer force field distributed around. Experiments of the proposed four rehabilitation training modes were carried out on the healthy subject, with the results showing a feasible method of the EBReCS in the simulation of multiple bimanual coordinated rehabilitation training tasks. In the future, the constructed EBReCS is expected to be clinically tested and exploited for home rehabilitation as a coordinated training device.

It is well known that conservative mechanical systems exhibit local oscillatory behaviours due to their elastic and gravitational potentials, which completely characterise these periodic motions together with the inertial properties of the system. The classification of these periodic behaviours and their geometric characterisation are in an on-going secular debate, which recently led to the so-called eigenmanifold theory. The eigenmanifold characterises nonlinear oscillations as a generalisation of linear eigenspaces. With the motivation of performing periodic tasks efficiently, we use tools coming from this theory to construct an optimization problem aimed at inducing desired closed-loop oscillations through a state feedback law. We solve the constructed optimization problem via gradient-descent methods involving neural networks. Extensive simulations show the validity of the approach.