A system of autonomous control of robotic devices based on an embedded virtual model on the example of a symmetrical walking robotic platform
Modern robotics is a dynamically developing direction in science and industry Robots themselves are able to solve a very wide range of tasks depending on the field of application. One of the most popular applications of robotic solutions is the search and rescue industry. Here, more than ever, the equipment able to move autonomously in conditions that are dangerous or deadly for humans are in demand. The most promising direction of development of robotics in this area is the creation of autonomous walking robots, which, unlike wheeled robots, are able to move under conditions of a priori uncertainty of the environment. Moving in such an environment requires a complex, intelligent control system. The goal of this work is to improve the efficiency of functioning of mobile autonomous robotic systems through application of new algorithmic and design solutions in their control systems. The authors propose changes to the structure of the embedded intellectual control system, expressed in the introduction of a predictive control module at the strategic level, interacting with a virtual model of the robotic complex. The robotics system will become capable of solving predictive control problems when, in parallel with the real-time control system, the operation of simulation software is possible, which will allow, based on the information about the environment and the current state of the robotic complex, to “predict” the future behaviour of the robot. In addition, the authors propose the concept and design of a symmetrical walking robotic platform based on a symmetrical three-dimensional propulsion system. The main difference from existing solutions lies in the special patented design of the propulsion system, which allows the robotic platform, made on its basis, to function both in the normal and in the overturned state without loss of functionality. A prototype was developed. Using these solutions will allow: 1.Exploring the potential of adaptive control algorithms.2.Reducing the development time of robotic complexes applying the portable virtual model.3.Significantly increasing the ability of robotic complexes to overcome irregularities and heterogeneity of the surfaces due to the new design solutions.
Control system; Controller; Model; Predictive algorithms; Walking robot