CN-121974116-A - Pipe belt machine fault diagnosis method and system

CN121974116ACN 121974116 ACN121974116 ACN 121974116ACN-121974116-A

Abstract

The invention discloses a fault diagnosis method and a fault diagnosis system for a pipe belt machine, which belong to the technical field of fault diagnosis, a plurality of vibration sensors are distributed in the conveying direction of the pipe belt machine and synchronously acquire vibration signals, target frequency components related to the running state of a carrier roller are extracted, an initial vibration phase difference map is constructed, a structural topology model reflecting a vibration propagation path is established, a structural path difference coefficient and directional propagation phase offset are calculated and are input into a diagnosis model as propagation path characteristics, a corrected vibration phase response result is obtained, the fault position of the carrier roller is finally positioned based on the correction result, and the influence of propagation interference on positioning accuracy is evaluated.

Inventors

SUN SHANJIN
CHEN MING
Zhong Baoshun
QIU LEIMING
ZHENG ZHAOZONG
HE WEIKAI
ZHANG YING
ZHANG JIANJUN
XIE MENG

Assignees

山东山矿机械有限公司

Dates

Publication Date: 20260505
Application Date: 20260126

Claims (10)

1. A fault diagnosis method of a pipe belt machine is characterized by comprising the following steps: Arranging vibration sensors at a plurality of key positions of the pipe belt machine along the conveying direction, and synchronously collecting vibration signals in the running process of the pipe belt machine; preprocessing the vibration signal, extracting a target frequency component related to the carrier roller running state, calculating phase information of different vibration sensors under the target frequency component, and constructing an initial vibration phase difference map; constructing a structural topology model reflecting a vibration signal propagation path based on the structural connection relation of the pipe belt machine; Taking the initial vibration phase difference spectrum as a node characteristic, taking a structural path difference coefficient and a directional propagation phase offset as structural propagation path characteristics, inputting the structural path difference coefficient and the directional propagation phase offset into a diagnosis model constructed based on a structural topology model, and comprehensively calculating the phase change relation of a vibration signal under the condition of multipath propagation to obtain a corrected vibration phase response result; And carrying out positioning analysis on the carrier roller fault position based on the corrected vibration phase response result, and evaluating whether the carrier roller fault positioning result generates an error exceeding a preset threshold value due to structural multipath propagation interference.
2. The method for diagnosing a pipe-line machine fault according to claim 1, wherein the vibration sensor is a triaxial acceleration sensor, a microelectromechanical vibration sensor or an industrial-level high-sensitivity accelerometer, preferably having a frequency response range of 0.5 Hz to 5,000 Hz, and the sensors are arranged at equal intervals in the conveying direction, and the intervals range of 1.5 m to 3 m.
3. The method of diagnosing a pipe-and-belt machine failure according to claim 1, wherein the preprocessing of the vibration signal includes performing a finite impulse response band-pass filtering process on the original time domain signal, wherein the filter passband range is 10 Hz to 300 Hz, the filter order is 128 th order, and the Kaiser window function is used to design the filter coefficients to filter out low frequency ground disturbance and high frequency electromagnetic noise.
4. The method for diagnosing the pipe belt machine fault according to claim 1, wherein the extraction mode of the target frequency component comprises the steps of calculating a spectrum mean value amplitude of a vibration signal, and selecting frequency components which meet the requirement that the amplitude is not lower than 1.5 times of frequency spectrum mean value and repeatedly occur in a plurality of continuous periods in a range of 20 Hz to 150 Hz as the target frequency representing the running state of the carrier roller.
5. The method for diagnosing a pipe-line machine fault according to claim 1, wherein the structural topology model is an undirected weighted graph model, nodes represent vibration sensors, edges represent structural propagation paths between the two sensors, and weights of the edges are calculated by path lengths, material propagation speeds and structural connection impedance factors.
6. The method for diagnosing a pipe-line machine fault according to claim 1, wherein the calculation of the structural path difference coefficient includes applying an excitation signal and obtaining a vibration response signal, calculating a propagation delay of a structural propagation path using a cross correlation function, and combining a path length and a connection impedance level according to the formula Calculated, wherein: a structure path difference coefficient for the i-th to j-th structure paths; Is the propagation delay of the reference path; Impedance factors are connected for the structure.
7. The method for diagnosing a pipe-line machine fault according to claim 1, wherein the calculation of the directional propagation phase offset includes constructing a polar coordinate system with a conveying direction as a reference angle, calculating phase values of sensors in different angular directions at a target frequency, and forming a directional phase function with the phase values of the reference direction, wherein the calculation expression is: Wherein DPO (thetak) is the phase offset of the kth direction relative to the reference direction, For the phase response of the kth direction at frequency ft, Is the reference direction angle.
8. The method for diagnosing the pipe-line machine fault according to claim 1, wherein the diagnosis model is a graph-note-force network model, the input layer of the graph-note-force network model receives node characteristics and edge characteristics, the graph-note-force layer dynamically adjusts weights according to adjacent nodes and edge characteristics of the adjacent nodes, and the output layer generates corrected node phase values.
9. The method for diagnosing a tube belt machine fault according to claim 1, wherein the step of performing a positioning analysis on the carrier roller fault location comprises the steps of: Extracting corrected phase values of all vibration sensors at a target frequency, constructing a corrected phase distribution sequence according to the layout sequence of the vibration sensors in the conveying direction of the pipe belt machine, and performing space mapping; Calculating a spatial first derivative of the phase distribution sequence to obtain a phase gradient value, judging whether the gradient difference value of the adjacent positions exceeds a preset mutation criterion threshold value, and if so, identifying the gradient difference value as a fault mutation point; Estimating the spatial center position of the carrier roller fault according to the mutation point sensor number and the adjacent phase change trend, and comparing the spatial center position with a set positioning error tolerance threshold; And carrying out error difference calculation on the corrected fault positioning result and an initial positioning result obtained based on the uncorrected phase response, and if the difference exceeds a tolerance threshold, judging that the fault positioning result is error caused by the structural multipath propagation interference and generating interference influence prompt information.
10. A fault diagnosis system for a pipe-and-belt machine, for implementing the fault diagnosis method for a pipe-and-belt machine according to any one of claims 1 to 9, characterized by comprising: the vibration signal acquisition module is used for arranging vibration sensors at a plurality of key positions of the pipe belt conveyor along the conveying direction and synchronously acquiring vibration signals in the running process of the pipe belt conveyor; The vibration signal processing module is used for preprocessing the vibration signal, extracting a target frequency component related to the carrier roller running state, calculating phase information of different vibration sensors under the target frequency component, and constructing an initial vibration phase difference map; the structural topology modeling module is used for constructing a structural topology model reflecting the propagation path of the vibration signal based on the structural connection relation of the pipe belt machine; The interference modeling module takes the initial vibration phase difference spectrum as node characteristics, takes a structural path difference coefficient and directional propagation phase offset as structural propagation path characteristics, inputs the structural path difference coefficient and directional propagation phase offset into a diagnosis model constructed based on a structural topology model, and comprehensively calculates the phase change relation of a vibration signal under the condition of multipath propagation to obtain a corrected vibration phase response result; And the error evaluation module is used for carrying out positioning analysis on the carrier roller fault position based on the corrected vibration phase response result and evaluating whether the carrier roller fault positioning result generates an error exceeding a preset threshold value due to structural multipath propagation interference.

Description

Pipe belt machine fault diagnosis method and system Technical Field The invention relates to the technical field of fault diagnosis, in particular to a fault diagnosis method and system for a pipe belt machine. Background The pipe belt machine is a device for continuously conveying bulk materials, and utilizes a closed rubber pipe and a belt conveying structure to realize efficient and environment-friendly conveying of the materials. However, in the prior art, when the abnormal carrier roller of the pipe belt conveyor is positioned by using the phase difference spectrum, serious problems of structural multipath propagation interference are often faced. Because vibration signals can propagate along multiple paths such as a main frame and a supporting diagonal brace in a complex structure of the pipe-and-belt machine, signal superposition of different paths can cause abnormal deviation of a phase angle or abrupt change of a pseudo phase, so that a fault position is misled to a system to judge. Especially in asymmetric structure or uneven scene of ground, this problem easily leads to erroneous judgement distance to reach more than several meters, not only influences maintenance efficiency, still probably misses the processing opportunity of true trouble, reduces diagnostic system's reliability and practicality. Disclosure of Invention The invention aims to provide a fault diagnosis method and system for a pipe belt machine, which are used for solving the defects in the background technology. In order to achieve the purpose, the invention provides the following technical scheme that the fault diagnosis method of the pipe belt machine comprises the following steps: Arranging vibration sensors at a plurality of key positions of the pipe belt machine along the conveying direction, and synchronously collecting vibration signals in the running process of the pipe belt machine; preprocessing the vibration signal, extracting a target frequency component related to the carrier roller running state, calculating phase information of different vibration sensors under the target frequency component, and constructing an initial vibration phase difference map; constructing a structural topology model reflecting a vibration signal propagation path based on the structural connection relation of the pipe belt machine; Taking the initial vibration phase difference spectrum as a node characteristic, taking a structural path difference coefficient and a directional propagation phase offset as structural propagation path characteristics, inputting the structural path difference coefficient and the directional propagation phase offset into a diagnosis model constructed based on a structural topology model, and comprehensively calculating the phase change relation of a vibration signal under the condition of multipath propagation to obtain a corrected vibration phase response result; And carrying out positioning analysis on the carrier roller fault position based on the corrected vibration phase response result, and evaluating whether the carrier roller fault positioning result generates an error exceeding a preset threshold value due to structural multipath propagation interference. Preferably, the vibration sensor is a triaxial acceleration sensor, a microelectromechanical vibration sensor or an industrial-level high-sensitivity accelerometer, the frequency response range is preferably 0.5 Hz to 5,000 Hz, the sensors are distributed in an equidistant manner along the conveying direction, and the distance range is 1.5 to 3 meters. Preferably, the preprocessing of the vibration signal includes performing finite impulse response band-pass filtering on the original time domain signal, wherein the filter passband range is 10 Hz to 300 Hz, the filter order is 128, and the filter coefficient is designed by using a Kaiser window function so as to filter low-frequency ground disturbance and high-frequency electromagnetic noise. Preferably, the extraction method of the target frequency component includes calculating a spectrum mean amplitude of the vibration signal, and selecting a frequency component which satisfies the frequency multiplication spectrum mean of not less than 1.5 and repeatedly appears in a plurality of continuous periods within a range of 20 Hz to 150 Hz as the target frequency representing the carrier roller running state. Preferably, the structural topology model is an undirected weighted graph model, the nodes represent vibration sensors, the edges represent structural propagation paths between the two sensors, and the weights of the edges are calculated by path lengths, material propagation speeds and structural connection impedance factors. Preferably, the calculation of the structural path difference coefficient includes applying an excitation signal and obtaining a vibration response signal, calculating the propagation delay of the structural propagation path by using a cross correlation function, and combining the p