Synergetic knowledge bases

Modern knowledge bases must largely correspond to human thinking and the reality of the world. Synergetic knowledge bases create the possibility of joint use of both "hard" computing, which require the accuracy and uniqueness of the solution, and "soft" computing, allowing a given error and uncertainty for a specific problem. A methodology for creating synergetic systems for the representation of knowledge using artificial intelligence technologies is proposed. The methodology is based on knowledge base methods and can be used to develop design and management systems in industries. A model for representing linguistic variables is proposed. The method of creating fuzzy knowledge bases and the stages of the inference mechanism are considered. The fuzzy inference is described using the example of the Mamdani mechanism. A functional diagram of the creation of fuzzy inference systems based on a structured clear knowledge module is proposed. A method for creating knowledge bases for the implementation of neural network models is considered. An example of a knowledge base for training neural networks is given.