Eine Plattform für die Wissenschaft: Bauingenieurwesen, Architektur und Urbanistik
Eine Plattform für die Wissenschaft: Bauingenieurwesen, Architektur und Urbanistik
An object-oriented framework for spatial motion planning of multibody systems
This work presents an approach for handling optimal motion planning problems with direct methods which integrates motion interpolation tools, multibody dynamics methods, as well as nonlinear programming routines. Spatial path parametrization is hereby formulated in Euclidean space, providing a geometry-oriented user interface with intuitive visualization of the differential geometric properties of the motion, thus opening new ways for spatial path optimization. In particular, two main topics are explored in detail: the computation of time-optimal motions along given paths, including their extension to no-slip conditions and overall power consumption constraints, and the optimization of spatial path geometries with B-splines and nonlinear optimization routines. General optimal motion planning problems are discretized into nonlinear optimization problems by decomposing multibody motion in two main components, namely, the spatial path followed by the multibody system and the one-dimensional motion of the system along this path. The geometry of the spatial path is parametrized using curve fitting with B-splines, allowing for the path’s via-points and boundary conditions to be used as the design parameters of the optimization problem. The motion along the path is then computed by integrating the projected equations of motion or by solving the optimal problem along the path, and the simulated kinematic and dynamic time histories – measured at different poses on the multibody system – are passed to state-of-the-art gradient-based optimization routines in the form of cost and constraint functions. The user is enabled to choose the relevant design parameters, as well as to prescribe kinematic and dynamic cost and constraint functions with different levels of complexity. This novel approach for spatial path planning is of advantage in applications where the non-convexity and nonlinearity of the problem make automatic optimization difficult or even impossible. Its practical applicability is illustrated by theoretical ...
An object-oriented framework for spatial motion planning of multibody systems
This work presents an approach for handling optimal motion planning problems with direct methods which integrates motion interpolation tools, multibody dynamics methods, as well as nonlinear programming routines. Spatial path parametrization is hereby formulated in Euclidean space, providing a geometry-oriented user interface with intuitive visualization of the differential geometric properties of the motion, thus opening new ways for spatial path optimization. In particular, two main topics are explored in detail: the computation of time-optimal motions along given paths, including their extension to no-slip conditions and overall power consumption constraints, and the optimization of spatial path geometries with B-splines and nonlinear optimization routines. General optimal motion planning problems are discretized into nonlinear optimization problems by decomposing multibody motion in two main components, namely, the spatial path followed by the multibody system and the one-dimensional motion of the system along this path. The geometry of the spatial path is parametrized using curve fitting with B-splines, allowing for the path’s via-points and boundary conditions to be used as the design parameters of the optimization problem. The motion along the path is then computed by integrating the projected equations of motion or by solving the optimal problem along the path, and the simulated kinematic and dynamic time histories – measured at different poses on the multibody system – are passed to state-of-the-art gradient-based optimization routines in the form of cost and constraint functions. The user is enabled to choose the relevant design parameters, as well as to prescribe kinematic and dynamic cost and constraint functions with different levels of complexity. This novel approach for spatial path planning is of advantage in applications where the non-convexity and nonlinearity of the problem make automatic optimization difficult or even impossible. Its practical applicability is illustrated by theoretical ...
An object-oriented framework for spatial motion planning of multibody systems
Geu Flores, Francisco (Autor:in)
29.01.2014
Hochschulschrift
Elektronische Ressource
Englisch
Object-oriented analysis of spatial multibody systems based on graph theory
| Tema Archiv | 1997
Object-oriented multibody motorcycle modelling for control systems prototyping
| Tema Archiv | 2006
Using an extensible object-oriented query language in multibody system analysis
| Tema Archiv | 2001
Object oriented spatial analysis
| British Library Conference Proceedings | 1994
Collisions of multibody systems
| British Library Online Contents | 2001