Hamid Khaloozadeh/ M. Reza Homaeinejad
The Journal of Mechanical Science and Technology, vol. 24, no. 6, pp.1337-1351, 2010

Abstract : The problem of controlling Space Free-flying Robots (SFFRs), which have many degrees of freedom caused by their mechanical manipulators, is challenging because of the strong nonlinearities and their heavy computational burden for the implementation of modelbased control algorithms. In this paper, a chattering avoidance sliding mode controller is developed for SFFR as highly nonlinear-coupled systems. To fulfill stability requirements, robustness properties, and chattering elimination, a regulating routine is proposed to determine the proper positive values for the coefficient of sliding condition. To solve the run-time problem, an explicit direct relationship between the SFFR’s output of actuators (force/torque) and the measurement of distances from the corresponding sliding surfaces is also assumed. To reach perfect performance, the parameters are estimated recursively using the Kalman filter as a parameter estimator. The explicit dynamics of a 14-DOF SFFR is derived using SPACEMAPLE, and the recursive prediction error method (RPEM) is used to parameterize the SFFR model. To alleviate the chattering trend, a multi-input sliding mode control law is proposed and applied to the given SFFR based on the online estimated dynamics to control its orientation and position to catch a moving target. To evaluate the new proposed algorithm in a more complicated condition, only on–off actuators are assumed for controlling the base of SFFR because it is the case in real systems. The obtained results show that the proposed regulated sliding mode controller can significantly reduce the chattering trend. Consequently, energy consumption will be substantially decreased, and running the control algorithm will be within a reasonable time duration.

Keyword : 14-DOF space free-flying robots; Variable structure systems; Parameter estimation; Kalman filtering