What are the contents of intelligent diagnostic technology

Intelligent diagnostic techniques include: (1) expert system diagnosis. Expert system is a kind of artificial intelligence computer program that applies the knowledge and reasoning methods of a large number of human experts to solve complex practical problems. Generally includes knowledge base, database, reasoning machine, human-computer interface and knowledge base management system, interpretation system. Fault diagnosis expert system is an important branch of expert system applications. (2) Artificial neural network diagnosis. Artificial neural networks are widely valued in the field of fault detection and diagnosis for their massively parallel processing capability, adaptive learning capability, distributed information storage, robustness, fault tolerance and generalization capability. The application objects are mainly devices and subsystems. (3) Pattern recognition diagnosis. Pattern recognition diagnosis is to simulate and analyze the workflow of the system, together with human experience, to build a variety of failure modes, and according to the measurement information, to determine which mode the system belongs to, so as to detect and separate the fault. It includes both individual identification method and group identification method. (4) Fault tree analysis method. Fault tree analysis is a deductive analysis method that unfolds layer by layer from top to bottom. It is the least system or equipment failure for the top event, down layer by layer to find out all the causes of the event, with a special inverted tree logic causality diagram (i.e., fault tree), represents the logical relationship of events, and qualitative and quantitative safety and reliability analysis. This method is a more commonly used fault diagnosis method, mainly used for offline diagnosis of simple objects. (5) Fuzzy diagnosis. Fuzzy concepts are concepts whose connotation is certain but whose extension is uncertain, such as: "excessive voltage", "motor overheating" and so on. It is because of these fuzzy knowledge and empirical knowledge in fault diagnosis exists, so the fuzzy diagnosis technology has more occasions for use. (6) Gray system theory diagnosis. Gray system theory is China's scholars Deng Julong 1982, first to the international proposed, gray concept is the concept of certainty and connotation of uncertainty, such as "robot out of control". Gray system refers to part of the information is clear and part of the information is not clear system. Gray system theory is an extension of the cybernetic point of view and methodology, it is from the perspective of the system to study the relationship between the information, that is, the study of how to use the known information to reveal the unknown information, that is, the system's "whiteness problem". An operating device is actually a complex gray system. In this system, some information can be known, some information is inaccurate or impossible to know, fault diagnosis is the use of known information to recognize the characteristics of the information system containing unknowable, state and development trend, and to make predictions and decisions about the future, in fact, is a gray system of the whitening process. Among them, gray prediction model, gray correlation and gray clustering method and other theoretical methods have been initially applied in time series model ordering, prediction and fault diagnosis. (7) Wavelet analysis diagnosis. Wavelet transform is a new technology of signal analysis that has been rapidly developed and formed a research hotspot in recent years, and is considered to be a breakthrough progress of Fourier analysis method. (8) Genetic algorithm diagnosis. The main features of the genetic algorithm are the group search strategy and the transformation of information between the individuals in the group, which can climb multiple peaks in parallel, the search does not depend on the gradient information, and the variation rule of probability is used to guide its search direction. It is especially suitable for dealing with complex problems and nonlinear problems that are difficult to solve in traditional search methods, which not only avoids the defects of local optimization algorithms, but also can make use of the inherent knowledge to narrow the search space, and avoid the combinatorial explosion of other global optimization algorithms that produce searches. (9) Integrated diagnosis.