Adaptive navigation and motion planning for autonomous mobile robots

Abstract

Exploring autonomy in robotics is a meaningful task. The intuitive definition of autonomy is the capability of a robot to make a decision based on its own knowledge, acquired by its distributed sensors, without any human interference. Throughout this framework we discuss some algorithms and techniques underlying the subjects of adaptive navigation and motion planning for autonomous mobile robots. Mobile Robots will play an important role in many future applications, such as personal and service robots, handicapped aid, entertainment, space exploration, medical applications, nuclear industry, surveillance or autonomous transportation. It is the mobility that distinguishes a mobile platform from robot manipulators and gives the possibility to actively and adaptively interact with the environment and humans. In real world environments actual mobile robot platforms will need increased adaptability and autonomy with better techniques for navigation safety, map building, obstacle avoidance and path planning. This study will first focus on sensor integration and adapted interaction that is considered one of the major basic concepts for mobile platforms. The contributions of this study arise from a formulation of new methods and techniques for sensor integration, mapping, motion planning and adaptive navigation instead of previously dominant approaches. By manipulating the manner in which feature information of sensor data is incorporated into processing, it can be shown that significant improvements in the performance of the presented algorithms can be attained. Moreover, the simplicity and the efficiency of the applied techniques succeeded to reinforce the robustness of the overall system in static and dynamic environments. The key idea of that is to achieve the following tasks improvement of system autonomy, reinforcement of the overall stability, enhancement of precision, increment of flexibility, reduction of energy consumption and more adaptation.

In dieser Arbeit wurde die Pfadplanung und die adaptive Regelung mobiler Roboter diskutiert. Die Robustheit des Regelungssystems wurde durch Sensorintegration verstärkt, z.B. Sonarsensoren, Laserscanner, CCD-Kameras und einen geomagnetischen Kompass. Dabei wurde die Einsatzumgebung abgebildet, die Kollision vermieden und die Geschwindigkeit bei der Objektverfolgung geregelt.