Forschungs- & Entwicklungsinformationsdienst der Gemeinschaft - CORDIS

Task-level optimal programming of industrial robots

The project concerns the development of industrially applicable procedures for optimal robot motion planning. The computational performance of modern robot control hardware allows the involvement of sophisticated on-line robot control algorithms with improved dynamic robot modelling. The idea to include the dynamic robot model into the robot control unit was developed in the project based on two different approaches: the continuous movement modelling and the discrete moving modelling. The results of the experimental investigation demonstrated that both approaches have their own spheres of applicability. The discrete approach developed has the following advantages: the simplicity of the model description; the simplicity of the boundary condition definition, which means, the description of the obstacle location and geometry; despite large dimensions to be solved, fast algorithms for the problem evaluation were constructed. In order to successfully implement the designed methods in real industrial products, a high modelling accuracy must be combined with a high numeric efficiency of the model evaluation. The continuous approach has the following advantages: high accuracy of the robot movement modelling; adequate involvement of the robot dynamic characteristics; easy interpretation of the problem solution. Though these advantages were demonstrated in the course of the project, the numerical efficiency of the methods must be improved. The proposed way of reaching this goal is the combination of the discrete and the continuous approaches, whereby the continuous methods should be applied to the trajectory parts with higher precision requirements. The investigation of the trajectory switching problem resulted in the development of a basic method for the solution stability analysis which is an efficient tool for estimating model applicability.

Reported by

Fraunhofer-Gesellschaft zur Förderung der Angewandten Forschung eV
Pascalstraße 8-9
10587 Berlin
See on map
Folgen Sie uns auf: RSS Facebook Twitter YouTube Verwaltet vom Amt für Veröffentlichungen der EU Nach oben