Control of vehicles and robots: creation of planning systems in the state space (MIPRA)

Abstract

To control machines and robots, in addition to the well-known reflex level control systems, it is proposed to use logical level decision-making systems that are created on the basis of mivar expert systems. It is shown that the use of mivar expert systems makes it possible to solve problems of automatic planning of actions of robots in the state space. These tasks relate to STRIPS planning. The results of the study of the Mivar-based Intelligent Planner of Robot Actions (MIPRA) project have dramatically reduced the computing infrastructure requirements from servers to conventional computers and the time it takes to plan tasks for robot activities from weeks to seconds. Similar results after creating special knowledge bases can be obtained for groups of robots, multi-level heterogeneous robotic systems and cyber-physical systems of various bases and purposes, which is an important section in the control of machines and robots. Solving automatic planning tasks makes it possible to quickly develop driver assistance systems and move on to creating a fully autonomous smart car.

Creating a “Logical intelligent plant care system” in digital agriculture based on Mivar approach

“…The project presented in this work can be used as a hybrid expert system . Advantages of this system include high speed solution development with linear computational complexity, availability of requirements for computing equipment, the ability to enter information into the knowledge base directly to the expert agronomist [16], and the ability to solve and visualize complex logical problems on ultra-large data sets [17]. Thanks to the algorithms based on active trainable evolutionary mivar network, it is possible to implement a plant care system that considers individual characteristics of crop growth, make decisions in conditions of heterogeneous sensor data (data collection), and timely adjust the process of plants growing.…”