Dynamic modelling and parameter identification for cable-driven manipulator

Abstract

Cable-driven manipulators (CDMs) have several advantages, but unknown dynamic characters restrict their control performance. The motors of CDMs are placed at the base and power is transmitted by cables between the motors and driving pulleys. These cables interact with the links of the manipulator, which poses a challenge for obtaining the dynamic model. In this study, the interaction between cables and manipulators is analysed, and a dynamic model is derived based on Newton-Euler method. To eliminate excessive variance in recursive equations, the impact of pretension force is considered and some equivalent assumptions are proposed. To improve the accuracy of parameter identification, limited terms of Fourier series are adopted for the identification trajectory. Considering various limitations of CDMs, such as maximum joint angle, speed and acceleration, artificial bee colony algorithm is used to optimize the coefficients of the identification trajectory. Simulations verify that the dynamic model can precisely calculate the driving torque of CDMs. Moreover, parameter identification experiment affirms the efficiency of the proposed parameter identification method.