Attempts have been made successfully to create a set of algorithms which use behavior-based artificial intelligence (AI) techniques to collectively yield desirable collisions avoidance behaviors in a dynamic operational environment. The algorithms consist of sensor-driven behavior or competence modules which are nessecary for dealing with various aspects of dynamic collision avoidance in urban traffic-like situations. These modules are implemented according to the principles of Subsumption Architecture, which is one of the key theories of behavior-based AI. A reduced-scale model operational environment was set up and the developed control system has been tested. The test environment is populated not only with stationary obstacles but also with moving objects such as humans and other test vehicles. Experiments were successfully repeated and results from these experiments indicate that it is feasible to construct control systems which allow fast and highly flexible manoeuvring of vehicles operating in tight proximities. Outlines of the algorithms developed, as well as the experiments conducted will be discussed.
Collision avoidance using behavioral-based AI techniques
Kollisionsverhütung mit verhaltensorientierter Technik der künstlichen Intelligenz
Intelligent Vehicles, Symposium, 1993 ; 141-145
1993
5 Seiten, 8 Bilder, 3 Quellen
Aufsatz (Konferenz)
Englisch
Collision Avoidance using Behavioral-Based AI Techniques
British Library Conference Proceedings | 1993
|COLLISION AVOIDANCE FOR CONNECTED VEHICLE BASED ON DIGITAL BEHAVIORAL TWIN
Europäisches Patentamt | 2020
|COLLISION AVOIDANCE SYSTEM, COLLISION AVOIDANCE METHOD, AND COLLISION AVOIDANCE PROGRAM
Europäisches Patentamt | 2023
|