Tracking Control of a Rotary Flexible Joint Robot using Adaptive fuzzy IMC

Recently, rotary flexible joint robots have become extremely common and essential as many applications require flexibility, such as performing refueling works, inspections, and maintenance work. However, the rotary flexible joint is a nonlinear, coupled, and under-actuated system. This makes the control procedure a challenging task. The control objective of this work is to track the desired trajectory and minimize the vibrations in the joint. Therefore, auto-tuning and adaptive control using fuzzy logic was combined with the internal model control (IMC) concept to develop an adaptive-based fuzzy IMC (AFIMC). By doing so, the proposed controller benefits from the high accuracy and enhanced performance of the intelligent techniques (i.e. fuzzy logic) as well as the simple structure, easy tuning, and robustness of the conventional IMC. In this paper, a conventional IMC was designed first for the rotary flexible joint system. In addition, triangular and trapezoidal membership functions were utilized to design two mamdani-type fuzzy logic controllers that are used to auto-tune the closed time constants instead of being fixed as in the conventional IMC. By doing so, the tracking performance and robustness of the conventional IMC are improved. This is proved after comparing the performance of the adaptive fuzzy IMC, and the conventional IMC in the absence and presence of disturbance through simulations. Simulation results show that the adaptive fuzzy IMC controller performs better than the conventional IMC since it is able to achieve the control objective and reject the disturbance more efficiently.