Title:Optimal Trajectory Planning and Model Predictive Control for Elastic and Redundant Multibody SystemsAuthor(s):Klemens SpringerAbstract:As a basis and for experimental validation of the proposed methods, an elastic articulated robot, an elastic gantry robot and a rigid articulated robot are used. Therefore, the setup, modeling and existing control strategies are discussed. Thereafter, the theory of parameter optimization and optimal control problems as well as essential solution algorithms are explained. In combination with the basic theory of curve description and trajectory planning, they constitute the fundament for the scientific contributions within this thesis. In the case of time-optimal trajectory planning for elastic robots, oscillation excitation is discussed as a consequence of the structural elasticities of the elastic articulated robot. Hence, an obligatory consideration of additional constraints within the optimization problem is proposed in order to avoid plastic deformations. For time-optimal paths for redundant robots, an explicit approach, which is based on a separation of the generalized coordinates into a redundant and a nonredundant part as well as a subsequent optimization of the trajectory is introduced. Since time-optimal point-to-point motions are not real-time computable for most robots and are still a challenging task, the last part of trajectory planning is dedicated to this topic. Within the proposed method, a time-optimal reference trajectory is learned with dynamic movement primitives and appropriately adapted to a new target position in real-time. Model predictive control constitutes the second part of scientific contributions within this thesis. An introduction to basic theory is given as a basis for proposing three different formulations for model predictive control of nonlinear systems and for discussing their characteristics. The focus is on the problem of computational expense and implementability at a real elastic system with vibration excitation as well as on the consideration of high-order state derivative constraints.Page Reference:121 page(s)Publishing:4/2015