Development of a Robot with an Intelligent Capability to Keep and Select a Path

The present paper is devoted to the development of a mobile robot, which is capable of following a pre-defined track and avoid collision with obstacles ahead of it. The avoidance of collision is implemented using a series of calculations and intelligent decision making support. The decisions are made according to the feedback given by the different sensors on the mobile robot, and a multilevel knowledge base programmed in the robot´s controller. The robot accommodates first order logic and meta-production rules for as knowledge representation models to describe the robot´s knowledge base, which defines the robot´s behavior. The use of meta-production rules allows the robot´s knowledge base to be fairly simple, but also flexible and extensible. Although the current intelligent capabilities of the robot are limited by avoidance of a single obstacle, the robot´s knowledge base can be extended by adding new semantic and meta-production rules, that would result extensions of the robot´s intelligence by new capabilities to avoid more complicated and especially multiple obstacles