Control system development for intelligent rehabilitation robotic system
A novel approach of control system development for an Intelligent Rehabilitation Robotic System (IRRS) is presented in this thesis. The major benefit of the MRS would be to physically rehabilitate stroke patient's limb motor and neuromuscular functions through artificial intelligence, real-time control, dynamical control and computer science. The proposed hybrid control system of the MRS consists of a traditional PID controller, combined with a fuzzy logic controller and neural network controllers to deal with system static, dynamics and external disturbances. The hybrid control system is designed, modeled and simulated through diverse methods by using different software including Coloured Petri-Nets for real-time analysis, Simulink for PID, Fuzzy Logic and Neural Network controller design and test, and LabVIEW for visual simulation and rapid-prototype establishment. Simulation results demonstrate satisfactory results. Preliminary experimental results are achieved from PID feedback and feed-forward control. System movements are tested accurate through several performance tests.