Localization and navigation of mobile robots in an environment with variable properties

Abstract

A method of localization and navigation of a mobile robot in an environment with variable properties in conditions of limited possibilities was proposed for remote control which provides for a possibility of switching the mode of robot control to a state of autonomous navigation. The method is based on combined application of a fuzzy model and an RL-algorithm that makes it possible to improve the set of fuzzy rules using the signal of reinforcement. Improvement of the method of localization of mobile objects using iBeacon and NFC technologies in a space with known maps of premises was proposed which enables reduction of the number of transmitters necessary for localization. The method of identification of mobile object movement routes was modified with the use of the modified Jump Point Search algorithm. Essence of the modification consists in the use of the algorithm of Manhattan distance between coordinates of the route points. This makes it possible to reduce impact of individual surges on the results of calculations compared with the basic algorithm. The obtained results can be used in mobile robot control systems in an environment with variable properties at limited possibilities for remote control. The results of testing the proposed methods and corresponding computational procedures confirm their performance and prospects of practical application. Application of the above approach makes it possible to take into account obstacle configurations and adjust the navigation strategy to improve the system quality (in 95 % of the test experiments, the mobile robot reached the target in an environment with various types of obstacles).