Cognitive intellignet control. Pt. 2: Quantum fuzzy inference algorithm in intelligent cognitive robotics

Abstract

In the first part of the article, a system for assessing operator emotions using deep machine learning based on soft computing and designing a cognitive control system was discussed. This work develops the approach of cognitive intelligent control, describing the strategy for designing intelligent cognitive control systems based on quantum and soft computing. The synergetic effect of quantum self-organization of the knowledge base, extracted from non-robust knowledge bases of an intelligent fuzzy controller, is demonstrated. The information-thermodynamic law of quantum self-organization of the optimal distribution of the basic qualities of control (stability, controllability and robustness) and the law of quantum information thermodynamics on the possibility of extracting additional useful work based on the extracted quantum information hidden in classical states are applied. Formed (without violating the second law of quantum thermodynamics) on the basis of the extracted amount of hidden quantum information, the «thermodynamic» control force allows the robot (as a control object) to perform quantitatively more useful work compared to the amount of work expended (on extracting quantum hidden information). The guaranteed achievement of the goal of controlling the robot is carried out on the basis of the designed intelligent cognitive con-trol system using the tools of the QCOptKBTM quantum knowledge base optimizer, the structure of which includes quantum fuzzy inference (QFI). The quantum algorithm for self-organization of non-robust knowledge bases of QFI structurally relies on synergistic effects from hidden quantum information to implement the implementation of the optimal distribution of control qualities. This technology makes it possible to increase the reliability of intelligent cognitive control systems in situations of control under conditions of danger, described using a cognitive neural interface and various types of interaction with robots. The examples have demonstrated the effectiveness of intro-ducing the QFI scheme as a ready-made programmable algorithmic solution for embedded intelligent control sys-tems. The possibility of using a neural interface based on a cognitive helmet with a quantum fuzzy controller to control a vehicle is shown.