%0 Journal Article %A Nguyen, Thanh L. %A Allen, Sam %A Phee, Soo J. %D 2018 %T Direct torque control for cable conduit mechanisms for the robotic foot for footwear testing %U https://repository.lboro.ac.uk/articles/journal_contribution/Direct_torque_control_for_cable_conduit_mechanisms_for_the_robotic_foot_for_footwear_testing/9624779 %2 https://repository.lboro.ac.uk/ndownloader/files/17273387 %K Torque mode control %K Footwear testing %K Hysteresis model %K Cable conduit mechanism %K Robust adaptive control %K Mechanical Engineering %K Medical and Health Sciences not elsewhere classified %X © 2018 Elsevier Ltd As the shoe durability is affected directly by the dynamic force/pressure between the shoe and its working environments (i.e., the contact ground and the human foot), a footwear testing system should replicate correctly this interaction force profile during gait cycles. Thus, in developing a robotic foot for footwear testing, it is important to power multiple foot joints and to control their output torque to produce correct dynamic effects on footwear. The cable conduit mechanism (CCM) offers great advantages for designing this robotic foot. It not only eliminates the cumbersome actuators and significant inertial effects from the fast-moving robotic foot but also allows a large amount of energy/force to be transmitted/propagated to the compact robotic foot. However, CCMs cause nonlinearities and hysteresis effects to the system performance. Recent studies on CCMs and hysteresis systems mostly addressed the position control. This paper introduces a new approach for modelling the torque transmission and controlling the output torque of a pair of CCMs, which are used to actuate the robotic foot for footwear testing. The proximal torque is used as the input signal for the Bouc–Wen hysteresis model to portray the torque transmission profile while a new robust adaptive control scheme is developed to online estimate and compensate for the nonlinearities and hysteresis effects. Both theoretical proof of stability and experimental validation of the new torque controller have been carried out and reported in this paper. Control experiments of other closed-loop control algorithms have been also conducted to compare their performance with the new controller effectiveness. Qualitative and quantitative results show that the new control approach significantly enhances the torque tracking performance for the system preceded by CCMs. %I Loughborough University