Résumé
In this thesis, a fundamental framework for reconfigurable cable-driven parallel robots (CDPR) is developed. It consists of three main parts:1. Propositions of general architectures of large-dimension reconfigurable CDPRs which can be used to replace conventional means of handling large and heavy parts across a wide workspace, e.g., in a factory workshop or in an airplane maintenance workshop.2. Analysis tools that can aid the design and control of a general (reconfigurable) CDPR. Various issues are addressed including kinematic, dynamic and elastostatic modeling, collision detections and cable modeling.3. Methods and tools to solve the CDPR reconfiguration problem efficiently. This problem is considered as a general nonlinear optimization problem. The thesis focuses on methods to formulate this optimization problem into standard forms which can be solved by using available (suitable) optimization software in order to obtain optimal CDPR reconfigurations.