Shi, Jialei; (2024) Static Modelling and Control for Pneumatic-driven Soft Robots with Reinforced Chambers. Doctoral thesis (Ph.D), UCL (University College London).

Text Shi_10187480_thesis.pdf Access restricted to UCL open access staff until 1 March 2028. Download (25MB)

Abstract

Soft robots, made of compliant materials, exhibit distinct and unique characteristics such as flexibility, compliance, versatility, and inherent safety, setting them apart from traditional rigid-linked robots. On the other hand, modelling and control of soft robots are challenging due to their flexible and compliant nature, resulting from, e.g., inherent high dimensionality, large deformation and nonlinear behaviours of soft robots. To this end, this thesis presents a static modelling and model-based control framework for pneumatic-driven soft robots with fibre-reinforced chambers. The proposed framework presented in this thesis begins with a modelling framework aimed at understanding three key aspects of soft robots: forward kinematics, robot stiffness/compliance and contact force generation. The modelling is primarily rooted in Lie theory, especially screw theory. The forward kinematics model can be built on the piece-wise constant model or the Cosserat rod model. The static force generation model can predict tip forces across different actuation levels. Furthermore, the proposed modelling framework illustrates that the configuration-dependent compliance/stiffness of soft robots can be integrated based on the static configurations from kinematics models. In terms of addressing material nonlinearity, the modelling framework employs various approaches, including hyper-elastic models or the pressure-dependent dynamic modulus model. The proposed framework, grounded in these modelling approaches, culminates in a model-based control framework for achieving inverse kinematics control and on-demand force control, eliminating the need for feedback position/orientation and force sensing devices. Specifically, the inverse kinematics control is accomplished by solving nonlinear least-squares optimisation problems using the static Cosserat rod model. This enables various control modes; for example, the controller is capable of executing decoupled position control for the tip of each segment in a two-segment robot. Additionally, the on-demand force control, rooted in the compliance model, is realised through solving inverse kinematics problems. The modelling and model-based control framework is also extensively and thoroughly validated using both numerical simulations and experiments. The results demonstrate a high fidelity and effectiveness of the proposed framework.

Download activity - last month

Export as CSVJSONXML

Download activity - last 12 months

Export as CSVJSONXML

Downloads by country - last 12 months

Export as XMLJSONCSV

Archive Staff Only

View Item