Multi-Criteria Optimal Design of Cable Driven Ankle Rehabilitation Robot

An ankle rehabilitation robot has been conceptualized and designed to realize the range of motion, muscle strengthening and proprioception training exercises for ankle joint. The robotic device is intended to help patients and therapists in their cooperative efforts for the treatment of impaired ankle joint as a result of injury or stroke. After analyzing the ankle joint anatomy and its motions, a parallel mechanism is proposed for the robot. To mimic the human ankle joint and its muscles actuation, the proposed robot uses artificial air muscles configured in a fashion close to the actual muscle arrangement. The apparent advantages of the proposed robot over the existing ankle rehabilitation parallel mechanisms have been emphasized. As a matter of fact, the performance of parallel robots greatly depends on their dimensions and the configuration of their actuators. Thus to explore the potential of these robots, it is essential to obtain a set of kinematic parameters, leading to optimal robot performance. To achieve this, robot designs need to be optimized on the basis of performance indices such as, workspace, condition number and Euclidean 2-norm of actuator forces, under various operational constraints. The performance criteria and the constraints are discussed in detail to justify their influences on the robot design. The existing Multi-objective Optimization Approaches (MOA) e.g. weighted formula approach, population based approach and Pareto optimal approach have been discussed. The algorithm used in this chapter is based on genetic algorithms and attempts to draw advantages of the weighted formula and the Pareto optimal approaches simultaneously for the optimization of robot design. The results obtained from the optimization are discussed and important inferences for further work are drawn.