New Super-twisting Sliding Mode Control of an ‎Upper Limb Rehabilitation Robot Based on the ‎TLBO Algorithm

Rehabilitation robots are very popular because they are beneficial tools in helping stroke patients and people with physical disabilities, so controlling them to get accurate performance is necessary. This paper presents a new super-twisting controller based on the determined gain with the TLBO algorithm (STA-TLBO) for an upper limb rehabilitation robot for the first time. One of the most important parts of designing the super twisting algorithm (STA) controller is determining the gains, which requires accurate calculations and obtaining disturbance. In this paper, the Teaching–Learning-Based Optimization (TLBO) algorithm is used to obtain the gains of the STA controller. To illustrate the validity of the proposed controller, the results are compared to PID, STA, and PID-TLBO controllers. The results indicate that the proposed controller ensures accurate tracking, finite-time convergence, and reduced chattering. The stability and the robustness of the PID-TLBO and STA-TLBO controllers are examined by three tests, parameter uncertainties, external disturbances, and step response. The results show that the STA-TLBO controller has a better performance than the others under different conditions; that means the proposed controller has a shorter convergence time, more accurate tracking, and fewer tracking error than the other three controllers.