Studies on Use Reliability of Chinese Tractor (Part 1)

Examination of Reliability Model by Neural Network

Abstract

This paper describes the use reliability of Chinese tractors, as assessed by measuring working hours until failure occurred in an agricultural field. A failure intensity function and reliability function for working hours before the first failure were evaluated using a nonhomogeneous Poisson process. Parameters in the Weibull process and Weibull distribution estimated using a neural network are almost the same as the results obtained with the probability paper and other statistical methods. As the shape parameter (β) in the failure intensity function is β<1, the behavior of the mean time between failures is a monotonous increase. Working hours before the first failure follow a Weibull distribution, and the shape parameter β is 1.24>1.