Electronic Thesis and Dissertation Repository


Doctor of Philosophy


Electrical and Computer Engineering


Dr. Mehrdad R. Kermani


In recent years, research on physical human-robot interaction has received considerable attention. Research on this subject has led to the study of new control and actuation mechanisms for robots in order to achieve intrinsic safety. Naturally, intrinsic safety is only achievable in kinematic structures that exhibit low output impedance. Existing solutions for reducing impedance are commonly obtained at the expense of reduced performance, or significant increase in mechanical complexity. Achieving high performance while guaranteeing safety seems to be a challenging goal that necessitates new actuation technologies in future generations of human-safe robots.

In this study, a novel two degrees-of-freedom safe manipulator is presented. The manipulator uses magneto-rheological fluid-based actuators. Magneto-rheological actuators offer low inertia-to-torque and mass-to-torque ratios which support their applications in human-friendly actuation. As a key element in the design of the manipulator, bi-directional actuation is attained by antagonistically coupling MR actuators at the joints. Antagonistically coupled MR actuators at the joints allow using a single motor to drive multiple joints. The motor is located at the base of the manipulator in order to further reduce the overall weight of the robot. Due to the unique characteristic of MR actuators, intrinsically safe actuation is achieved without compromising high quality actuation. Despite these advantages, modeling and control of MR actuators present some challenges. The antagonistic configuration of MR actuators may result in limit cycles in some cases when the actuator operates in the position control loop. To study the possibility of limit cycles, describing function method is employed to obtain the conditions under which limit cycles may occur in the operation of the system. Moreover, a connection between the amplitude and the frequency of the potential limit cycles and the system parameters is established to provide an insight into the design of the actuator as well as the controller. MR actuators require magnetic fields to control their output torques. The application of magnetic field however introduces hysteresis in the behaviors of MR actuators. To this effect, an adaptive model is developed to estimate the hysteretic behavior of the actuator. The effectiveness of the model is evaluated by comparing its results with those obtained using the Preisach model. These results are then extended to an adaptive control scheme in order to compensate for the effect of hysteresis. In both modeling and control, stability of proposed schemes are evaluated using Lyapunov method, and the effectiveness of the proposed methods are validated with experimental results.