Motion control system in a hybrid architecture for middle-size soccer robot

Abstract

This paper proposes the motion control system in a hybrid architecture for middle-size soccer robot. A hybrid architecture combining hierarchical architecture and subsumption architecture is composed of four levels, such as the central control unit, the sensor information processing unit, the motion control system and the execution level. A local path pfanning algorithm fitting for autonomous soccer robot in dynamic environment is applied, and several actions and behaviors are properly designed for soccer robot motion control. The effectiveness of the proposed motion control system is validated by simulating and implementing it in a real soccer robot platform SmartROB2.

Active Sensor Planning for Multiview Vision Tasks

“…For navigation in an active way, a robot is usually equipped with a "controllable" vision head, e.g. The decision problem for one sensor was constructed with a standard tree for myopic Zhuang et al (2004) developed an adaptive panoramic stereovision approach for localizing 3D moving objects at the department of computer science at the University of Massachusetts at Amherst. 3.10).…”