Research On Prediction Of Residual Life For Feed System Of CNC Machine Tools

The technology of predicting the service life of CNC machine is very important for its proper management and maintenance. To conduct life prediction for CNC machine, it is required to collect a large number of long-term operation data, but actually there big errors may occur in collected data due to complicated real environment and varied uncertainties.This paper is to study performance degradation prediction and life prediction method based on the relation between the fault and performance for feed system of CNC machine.The main content is as follows:(1)A kind of performance degrada tion prediction and life prediction method is proposed by analyzing the relation between the fault and performance of feed system of CNC machine from the perspective of fault-performance mapping based on the BP neutral network mapping model.(2) The growth mechanism of position error of mechanical transmission part of feed system is analyzed, and a kind of method predicting such error is proposed based on the fault-performance mapping. Position errors under different initial states are determined by establishing kinematical simulation models in ADAMS. A model mapping relations between position error and variables workbench- nut seat gap and ball screw gradient is established on the basis of BP neutral network. Finally a method of predicting the position error of mechanical transmission part of feed system is proposed by virtue of the BP neutral network mapping model.(3) A residual life prediction method is therefore proposed on the basis of performance threshold distribution. According to the the performance threshold distribution and performance degradation data distribution, the residual life is successfully predicted.(4) A test platform for feed system of CNC machine is set up.The proposed performance degradation prediction method is verified, and the residual life of test platform of feed system is predicted with the life prediction method.Finally, main contents of this paper are summarized and the orientation of further researches is determined.