Нейро-нечеткая сеть для оценки остаточного ресурса тракторных двигателей

Abstract

The issue of improving the technical operation of transport-technological machines, in particular, the diagnostics of the technical condition of tractor engines and determining its residual life is considered. The purpose of the research was to create a neuro-fuzzy network to determine the residual life of the tractors engines BELARUS-80.1/82.1 and KhTZ-17221 by the amount of the engine shafts wear. An analysis was made by using applications of intelligent systems for diagnosing transport and technological machines and their possible application in other areas; the wear of the tractors crankshafts was experimentally determined; experimental data were statistically processed and the resource of tractors was calculated. The resource indicators were determined for the most critical unit, the crankshaft, according to the main parameters - the amount of its wear at the points of attachment of the connecting rod and main journals. The wear parameters in the work were determined in the course of experimental studies with subsequent statistical processing. The distribution functions of the wear value in four critical places of the crankshaft are obtained. The experimental data served as the initial data for the neuro-fuzzy network, which is implemented in the MATLAB environment. Experimental data on the wear of tractor engine units and the values of the residual engine life obtained on their basis are presented.