A fault prediction device for subway electric sliding doors

Abstract

A failure prediction device for an electric sliding plug door of a subway comprises a current sensor module, a data acquisition card and a computer system. The computer system is provided with a data storage software module, a failure prediction software module, a display module and an alarm module. According to the execution steps of the device, the data storage software module receives and stores sliding plug door motor working current data acquired by the current sensor module; the failure prediction software module carries out signal preprocessing, feature extraction, state recognition, transition probability calculation and expected remaining service life prediction on the obtained current data, so that poor lubrication failures of the sliding plug door are predicted; the display module is used for displaying predicted expected remaining service life and opening and closing frequency; the alarm module gives out an alarm when an expected remaining service life prediction value is lower than a set threshold value. The failure prediction device is used for monitoring motor currents of the electric sliding plug door in the opening and closing process, predicting the poor lubrication failures of the sliding plug door and giving an alarm when the poor lubrication failures of the sliding plug door occur.

Description

technical field [0001] The invention relates to a fault prediction device for an electric sliding door of a subway, in particular to a device for predicting a failure of poor lubrication of a sliding door of a subway. Background technique [0002] The subway has become an important equipment of the urban rail transit system. The electric sliding door is a key component to ensure the safety of passengers and the operation of the system on the subway. Its working status directly affects the operational efficiency of the subway. The electric sliding door system is driven by the door control unit to drive the brushless DC motor to drive the screw nut transmission mechanism to open and close the door. During use, due to the frequent switching movement of the door load for a long time and impurities such as solid particles mixed into the mechanical transmission parts, as the number of opening and closing increases, the grease of the transmission system is gradually scraped off, re...

AI Technical Summary

Problems solved by technology

[0003] (1) When forecasting based on time series artificial neural network, its forecasting range is limited to short-term forecasting, and the forecasting results are prone to fall into local optimum, and large samples are required, and the results are not stable during training

[0004] (2) Most of the fault prediction algorithms based on degraded states reported in the literature ignore the residency between states and the k-means clustering tends to fall into a local optimal solution, although HSMM takes into account the residency information of states and uses Fault prediction has achieved certain research results, but it only considers the retention situation within each state, and does not add information between states, and its modeling is complicated

Images