FIELD: information technology.
SUBSTANCE: method of constructing fuzzy logic systems, where a sequence of fuzzy logic rules is created first. From each of these rules, a numerical characteristic is assigned - control quality index, where the fuzzy logic rules are enforced based on a trained neural network. Information signals or signals from the control object are transmitted to the inputs of the neural network. A sequence of output signals or a sequence of instructions and recommendations is formed at the output, where the trained neural network is a large trained artificial neural network. Each of the fuzzy logic rules is enforced by a separate fragment of the large trained artificial neural network (domain), where the number of domains corresponds to the number of fuzzy logic rules and also contains a certain excess number of backup domains. One of the domains functions of an arbitrator and switches outputs of the domains with outputs of the neural network based on the control quality index.
EFFECT: design of intelligent systems capable of solving complex control tasks, training and retraining a system during its operation, while changing functions of the system.
2 cl, 1 dwg
