Model pattern recognition options of parametrec of technical condition for forestry machinery

Abstract

Pattern recognition methods are the most mathematical section of the theory of artificial intelligence, which solved the problem associated with the classification of objects arbitrary nature. Pattern recognition is one of those problems that are constantly in everyday life solves natural intelligence. The efforts of scientists is already over half a century, aimed at the development of methods and algorithms for automatically solving this problem. With respect to matters vibroacoustic diagnostics units and mechanisms for forestry machinery even received quite reliable data, the main issue is the question of classification. To include this data, which class (image). Determination of deterioration connection "piston – liner" when vibroacoustic diagnostics engine machines for forestry machinery is mitigated by a number of indirect (side) parameters: amplitude, power spectrum, phase, etc. Getting a clear diagnosis (classification) is a fairly urgent task, especially from the perspective of automation decision. System status describes the set of parameters for technical condition of vehicles Forestry works that define it. Recognition of the system is possible by classifying it as one of the all classes (diagnoses). The number of diagnoses (classes, typical conditions, standards) depends on the particular tasks and objectives of the study. Most simply adapted methods for multidimensional system of minimum risk and the greatest likelihood method. In cases where the method of statistical solutions need to define the boundary decision, estimated side task much more difficult. Therefore, to consider simplifying the process of recognition of the presence of one diagnostic parameter – phase occurrence vibration shot link.