Online fault detection device installed in a railway vehicle and used for a high-speed rail operating component

摘要

An online fault detection device, fully installed within a railway vehicle and used for a high-speed rail operating component, is characterized by the fact that it includes a GPS module, a first and a second noise sensor, a 3G module, a hard disk, a microprocessor,a status indicator and a warning indicator;where the input of the microprocessor is connected to an output of the GPS module and an output of the noise sensor respectively;the output of the microprocessor is linked to an input of the status indicator and an input of the warning tonal;the microprocessor is interactively connected to the 3G module and hard disk respectively;the GPS module is trained to collect position information and train speed information;the first noise sensor is installed on the ground of the railway vehicle and is trained to collect noise signals transmitted by a railway vehicle body, and the second noise sensor is trained to collect noise signals in the railway vehicle;which are transmitted through the rail vehicle body and through the air;the hard disk is designed to pre-set, for the train, a normal operating spectral database consisting of permanent state components of order noise levels, which are generated by different sounds collected during the normal operation of the train, which do not show any fault, and a fault spectral database,which consists of permanent state components of noise levels,which are generated by different sounds, which are collected when the faulty train is operated or a faulty operating component is operated and for storing the noise signals collected by the noise sensor;the microprocessor is trained to produce a waterfall block spectrum by periodically collecting the noise signals through the noise sensor, the waterfall color spectrum being a 3-D map, the x-axis representing the frequency of the spectrum,the y-axis represents time and the chronologically measured spectra are arranged in order as a waterfall, z-axis represents the energy of the spectrograms and their size is distinguished by color, which is a binary function of frequency and time,and to perform a short-term Fourier transformation on a time domain data stream of the noise signals,and then to compare the waterfall block spectrum with an ambient noise spectrum, which is a waterfall block spectrum, which is produced using a predetermined ambient background noise.in order to obtain a characteristic type of noise measured, thereby obtaining a high order of a Section Frequency parameter of an IIR low pass filter;the microprocessor is further trained to calculate the spindle rotation speed of the rail vehicle according to the position information and the train speed information collected by the GPS module,and for processing the spindle rotation speed and waterfall block spectrum to produce a waterfall-block adjustment spectrum;and then to generate a time domain encoding for each of the order noise levels in order to obtain permanent status components for each order noise level after the time domain encoding passes the IIR low pass filter of a high order;the microprocessor is further trained to compare the durable state components for each level of order noise with the normal operating spectral database of the train and to determine whether a deviation exceeds a standard, corresponding to comparative deviations if the deviation exceeds the standard,the microprocessor performs an error alarm and compares the deviation with the error spectrum, if the deviation is not in the error spectrum, the microprocessor uses it as suspect error spectrum data;the microprocessor is further trained to enter the alarmed suspect error spectrum data into the database and to manage the database;and to determine that the suspect spectrum data is added to the error spectrum database or to the normal operational spectrum database of the train by filtering and determining by an intervening and judging person,to achieve continuous updating and improvement of the train's fault spectral database and normal operational spectral database;the microprocessor is further trained to perform a data interaction with a soil information centre via the 3G module;the status indicator is trained to perform an error;the warning tonator is trained to issue an alarm.