MODELS, METHODS, AND MEANS OF REPRODUCTION OF EXPERT KNOWLEDGE IN INTELLIGENT SUPPORT SYSTEM BUILDING-TECHNICAL EXPERTISE

The paper is devoted to solving such a scientific and practical problem as the creation of computerized infocommunication systems for support building-technical expertise to determine the causes of destruction and deformation of buildings and structures. The analysis of the current state of expert activity within the framework of building-technical expertise is carried out. Perspective directions of the introduction of intelligent infocommunication systems in the course of performance of building-technical expertise and expert researches are outlined. The architecture of Intelligent Support System Building-Technical Expertise and the communication scheme of experts with the system are shown. To mapping expert knowledge formalized in the form of fuzzy associative rules to the memory card of the Cascade ARTMAP category fuzzy artificial neural network, it is proposed to use a fuzzy Mamdani-type inference system. The main input data, on the basis of which a fuzzy conclusion is realized to establish the degree of influence of various environmental factors on the technical condition of buildings and structures, are systematized and presented in a form acceptable for processing by computerized systems. At the same time, the main focus is on the study of facilities that are built and operated on subsidence loess soils. The process of formalization of heuristics, which is based on the formation of associations related to information on the position of signs of deterioration of the technical condition of the objects of expertise and the position of the changed soil, is described. Examples of interpretation and fuzzification of input information on soil properties, characteristics of the soil base of the object of building-technical expertise, and the surrounding area are given. The described approach provides an opportunity to reduce the risks of making wrong decisions by using the system as an intelligent database. The use of an artificial fuzzy neural network of the Cascade ARTMAP category gives the system the ability to form an expert conclusion on the degree of influence of various environmental factors on the technical condition of objects in the fuzzy conditions of a partially observed environment.