V.V.Kozhevnikov1, B.M.Kostishko1, M.Yu.Leontev1, E.R.Mingachev1, S.V.Pavlov2, V.V.Prikhodko1
1S.P. Kapitsa Technological Research Institute of Ulyanovsk State University, Ulyanovsk, Russia
2Sosny Research and Development Company, Dimitrovgrad, Ulyanovsk region, Russia
Received: 10 November 2018; accepted in revised form: 04 December 2018
Abstract. An intellectual control system (ICS) to control robots can be built (designed) on the basis of the mathematical model of cognitive digital automata (CDA). The intellectual control system in this case is a software and hardware complex, where the mathematical model of the CDA determines the control system as an intellectual one. The cognitive ability of the mathematical model is determined by the possibility of forming new knowledge based on the knowledge gained at the training stage. The creativity of a mathematical model is determined by the ability to construct sequences (logical chains) of generating new knowledge.
A specific feature of the mathematical model of the CDA consists in the fact that the description of the neural network (NN) structure serves as the initial structural scheme of automata, and the logical function “NOT-AND-OR” is used as the neuron model. The mathematical apparatus of Petri nets (PN) is proposed as a tool to construct the mathematical model of the CDA.
The structure, composition and algorithm of functioning of the robot’s intellectual control system based on the mathematical model of the CDA is discussed in the paper. In accordance with the algorithm, the ICS operates in three modes: training, manual and automatic control. Training a mathematical model of the CDA can be performed both in the manual mode and in the automatic control mode. The possibility of learning in the automatic control mode, in its turn, provides the possibility of regenerating knowledge and, accordingly, the possibility of cognitive control.
Keywords: intellectual control system, robots, cognitive automata, artificial intelligence, neural networks, machine learning, cognition, thinking, Petri nets, equation of states, mathematical modeling, synthesis, generation, analysis, logic.
c 2018 European Society of Computational Methods in Sciences and Engineering