This paper presents a versatile approach to the synthesis and design of a bond graph model and a Kalman filter observer for an industrial back-support exoskeleton. Actually, the main purpose of developing a bond graph model is to investigate and understand better the system dynamics. On the other hand, the design of the Kalman observer always should be based on a model providing an adequate description of the system dynamics; however, when back-support exoskeletons are considered, the synthesis of a state observer becomes very challenging, since only nonlinear models may be adopted to reproduce the system dynamic response with adequate accuracy. The dynamic modeling of the exoskeleton robotic platform, used in this work, comprises an electrical brushless DC motor, gearbox transmission, torque sensor and human trunk (biomechanical model). On this basis, a block diagram model of the dynamic system is presented and an experimental test has been carried out for identifying the system parameters accordingly. Both the block diagram and bond graph dynamic models are simulated via MATLAB and 20-sim software (bond graph simulation software) respectively. Furthermore, the possibility of employing the Kalman filter observer together with a suitable linear model is investigated. Subsequently, the performance of the proposed Kalman observer is evaluated in a lifting task scenario with the use of a linear quadratic regulator (LQR) controller with double integral action. Finally, the most important simulation results are presented and discussed.
This is an open access article distributed under the Creative Commons Attribution License
which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited