A logical-linguistic routing method for unmanned vehicles with the minimum probability of accidents

Forming optimal motion control laws for unmanned vehicles (UVs) by analyzing sensory data about the choice environment is an integral part of designing their situational control systems. The weakly predictable variability of the UV operating environment and the imperfection of measuring means reduce the possibility of obtaining comprehensive information about the environment state. Therefore, routing to minimize travel time and the probability of an accident is performed under uncertainty. An effective way to solve this problem is using logical-probabilistic and logical-linguistic models and algorithms. This paper is intended to develop new optimal routing methods for UVs with estimating the probability of an accident based on the logical-linguistic classification of route segments. For this purpose, the rows of parameters and characteristics of reference route segments are created and compared with the logical-probabilistic and logical-linguistic parameters and characteristics of classified route segments considering their significance for routing. After processing sensory and statistical data, the proposed logical-probabilistic and logical-linguistic methods are used to estimate the probabilities of accidents and minimize a performance criterion. As a consequence, the accuracy and speed of optimal routing for UVs are both increased. The results of this research can be used in the central nervous system of intelligent robots to classify route segments obtained by analyzing sensory and statistical data, which will improve the quality of motion control in an uncertain environment.