Zoran Kunica

Behavior based mobile robot in automatic assembly

This work deals with the development of intelligent autonomous robot using behavior based control approach. The hypothesis is that intelligence, as something what assumes the understanding, creativity and self-improving, should rely on the learning ability. Planning of intelligent robot behavior addresses three main issues: finding task solutions in unknown situations, learning from experience and recognizing the similarity of problem paradigms. The presented behavior based model integrates perception, recognition, problem solving and learning capabilities. The reinforcement learning method is used here to evaluate robot behavior and to induce new, or improve the existing, knowledge. The acquired action (task) plan is stored as experience which can be used in solving similar future problems. To provide the recognition of problem similarities, the adaptive fuzzy shadowed (AFS) neural network is applied. This behavior based approach to the robot intelligence is simulated on the mobile robot model and verified on the Pioneer 2DX, the real mobile robot, using primarily the sonar perception of working environment to manage its performance in unknown surroundings for given task

Development of a design procedure for automatic assembly system

The paper is aimed at development of concurrent engineering (CE) assembly planning methodology and tools, trying to establish a general and integral approach to assembly system planning and modeling, urging the need for connection between assembly planning field and concepts that usually lay outside engineering domain, and involvement of still uncovered and computer non-implemented aspects of the planning. As the basis for mentioned, seven variants of the assembly system for a simple product are presented. The variants are compared regarding several criteria and the most appropriate is chosen.

Knowledge Based System for Geometry Features Reasoning from Three-Dimensional CAD Data

Integrated planning of flexible assembly combines the designing of product, process and system into a single computer aided system. Such an approach can significantly enhance planning results. But, computer reasoning on product geometry appears as the primary problem for computer implementation in assembly planning. In the present paper the knowledge based system for geometry features extraction from 3D CAD data provides formalization of a product geometry in order to enable the processing of its geometry properties. Implementations in assembly sequence planning, mating features deducing and automatic robot programming are presented.