Abstract
The flexible and adaptable nature of continuum soft robots makes them applicable to a wide range of operations not easily obtainable with conventional rigid-body robots. Thus, soft robots can be used in various operations such as manipulation tasks, minimally invasive surgery operations, robotic rehabilitation/wearable devices, inspection, and surveillance tasks. Unfortunately, the continuous nature of these robotic systems leads to significant modeling and control challenges. Presently, there are various modeling perspectives. However, a detailed review shows that current models are often characterized by problems such as high computational costs, quasi-static assumptions, imprecise inclusion of boundary conditions, spillover instability, etc. These problems limit the accuracy of the resulting model, requiring more effective modeling and control strategies. Therefore, this paper is aimed at improving the state of the art and science of current models by providing more effective strategies for the problems encountered. In this regard, the dynamic modeling of a two-link tendon-driven flexible manipulator based on hybrid parameter multibody system methodology will be presented to demonstrate these strategies. Using the model, path-planned dynamic controls based on pole placement, linear quadratic regulator, and sliding mode control methods will be implemented for a continuous time-varying path. Also, a comparison of the performance of the control methods, in addition to parametric studies for the optimal tendon connection points, will be presented. Results showed that the benefits of the modeling approach and strategies employed led to a highly accurate, real-time performance for the complex motions of the manipulator system.
