As the life span increases and the availability of physicians becomes more and more scarce, robotic rehabilitation for post-stroke patients becomes more and more demanding, especially due to the repeatability character of the rehabilitation exercises. Both lower and upper limb rehabilitation using robotic systems have proved to be very successful in different stages of the rehabilitation process, but only a few address the immediate (critical) post-stroke phase, especially when the patient is hemiplegic and is unable to stand. The paper presents the kinematic modelling, singularity analysis and gait simulation for a new 4-DOF parallel robot named RECOVER used for lower limb rehabilitation for bedridden patients. The robotic system has been designed for the mobilization of the lower limb, namely the following motions: the hip and knee flexion and the plantar adduction/abduction and flexion/dorsiflexion. The kinematics has been studied and the singularity configurations have been determined to achieve a failsafe rehabilitation robot. Numerical simulations prove that the system can be used for gait training exercises in safe conditions.
Contrary to basic kinematic modelling, during tight turns, wheelchairs can slide sideways by non-negligible magnitudes: Collision risks may be increased
