CN-122020019-A - Intelligent diagnosis and early warning system and method for driving faults of multi-mode belt conveyor

CN122020019ACN 122020019 ACN122020019 ACN 122020019ACN-122020019-A

Abstract

The invention relates to an intelligent diagnosis and early warning system and method for a driving fault of a multi-mode belt conveyor, and belongs to the technical field of fault diagnosis of driving systems of belt conveyors. The system comprises a substation acquisition unit, a field sensing unit, an industrial audio acquisition unit, an industrial server and a system client. The method comprises the steps of collecting current, voltage, vibration, temperature, audio and other multi-mode signals of key components in a driving system, transmitting the signals to an industrial server through an industrial Ethernet to perform multi-mode signal noise reduction processing, constructing a fusion matrix, correcting a dynamic threshold value in real time based on load rate and ambient temperature to perform local decision, and adopting a D-S evidence theory to achieve accurate tracing of fault components. The invention has the advantages of strong anti-interference capability, dynamic self-adaption of threshold value, high decision interpretation, high response speed and the like, can effectively realize real-time monitoring and intelligent pre-diagnosis alarming of the driving fault of the belt conveyor, remarkably improves the early warning accuracy and efficiency, and reduces the maintenance cost.

Inventors

SUN YE
LI NAN
YANG HONGCHAO
WANG ZHEN
GUO LIN
ZHAN SHUHONG
QI JIALONG
WANG XIN
Bu Huifeng
LI YANCHENG
WANG BAOJUN
ZHANG SHUQI
SUN JIAN

Assignees

秦皇岛港股份有限公司

Dates

Publication Date: 20260512
Application Date: 20251225

Claims (10)

1. The intelligent diagnosis and early warning system for the driving faults of the multi-mode belt conveyor is characterized by comprising a power substation acquisition unit, a field sensing unit, an industrial audio acquisition unit, an industrial server and a system client, wherein the power substation acquisition unit, the field sensing unit and the industrial audio acquisition unit are respectively connected with the industrial server through an industrial Ethernet, the system client is in communication connection with the industrial server, the power substation acquisition unit is used for acquiring current and voltage signals of a driving motor, the field sensing unit is used for acquiring vibration and temperature signals of the driving motor, a speed reducer and a driving roller, the industrial audio acquisition unit is used for acquiring equipment audio signals of the driving motor, the speed reducer and the driving roller and environment audio signals in a driving station, the industrial server is in communication connection with the power substation acquisition unit, the field sensing unit and the industrial audio acquisition unit through an industrial Ethernet, is used for conducting noise reduction treatment on the received multi-mode signals, constructing a feature fusion matrix, correcting dynamic thresholds in real time based on load rates and environment temperatures, conducting local decision making and conducting fault component tracing based on D-S evidence theory, and the system client is in communication connection with the industrial server, so that a user interface is provided for realizing the interactive interface access and the use of the system.
2. The intelligent diagnosis and early warning system for the driving faults of the multi-mode belt conveyor according to claim 1 is characterized in that the power substation acquisition unit comprises a current transformer CT and a voltage transformer PT which are arranged in a driving motor inlet wire cabinet, and the current and voltage data of a driving motor of the driving system are detected and acquired respectively.
3. The intelligent diagnosis and early warning system for driving faults of a multi-mode belt conveyor according to claim 1, wherein the on-site sensing unit comprises a plurality of Wen Zhen sensors which are respectively arranged on a driving end and a non-driving end of a driving motor, an input shaft and an output shaft of a speed reducer and bearing boxes on two sides of a driving roller, and are arranged in a manner of pointing to the rotation center of a measured shaft in the axial direction.
4. The intelligent diagnosis and early warning system for the driving faults of the multi-mode belt conveyor is characterized in that the industrial audio acquisition unit comprises a plurality of contact audio sensors and an environment pickup sensor, wherein the contact audio sensors are attached to the driving end of a driving motor, the output shaft end of a speed reducer and the surfaces of bearing boxes on two sides of a driving roller, and the environment pickup sensor is hung on a ceiling of a driving station.
5. The intelligent diagnosis and early warning system for driving faults of a multi-mode belt conveyor according to claim 1, wherein the industrial server adopts a dual-CPU architecture, wherein a first CPU is dedicated to wavelet noise reduction, blind source separation and feature extraction of multi-mode signals, and a second CPU is dedicated to performing weighted voting decision based on dynamic threshold and fault matching based on D-S evidence theory.
6. A method for providing a multi-modal belt conveyor drive failure intelligent diagnostic and warning system as claimed in any one of claims 1 to 5, comprising the steps of: Step 1, an industrial server receives multi-mode signals transmitted by a substation acquisition unit, a field sensing unit and an industrial audio unit through an industrial Ethernet, and the multi-mode signals are preprocessed through a first CPU to eliminate power frequency interference, environmental noise and transmission noise; step 2, the industrial server extracts fault sensitive features from the noise-reduced multi-mode signals through a first CPU, and a feature fusion matrix is constructed; Step 3, the industrial server takes the feature fusion matrix as input through a second CPU, corrects the dynamic threshold value in real time according to the load rate and the ambient temperature of the belt conveyor drive, and adopts a weighted voting mechanism to make a local decision according to the fault probability of multi-feature combination; And 4, the industrial server uses a local decision result as input through a second CPU, and identifies the equipment fault component and type through a D-S evidence theory so as to realize fault positioning and tracing.
7. The method for early warning of the intelligent diagnosis and early warning system for the driving faults of the multi-mode belt conveyor is characterized in that the specific implementation method of the step 1 is that an adaptive wave trap is adopted to eliminate power frequency interference in electric signals, a wavelet threshold noise reduction method is adopted to process vibration signals, the shrinkage threshold of the vibration signals is adaptively adjusted along with the kurtosis of the signals, and a blind source separation method is adopted to separate audio signals so as to inhibit steady background noise.
8. The method for early warning of the intelligent diagnosis and early warning system for the driving faults of the multi-mode belt conveyor is characterized in that the specific implementation method of the step 2 is that fault sensitive features of preset dimensions are respectively extracted from current/voltage signals, vibration signals, temperature signals and audio signals, feature vectors are aligned in time through an interpolation method for time domain asynchronous problems caused by different sampling rates, and a feature fusion matrix comprising 15-dimensional basic features and 45-dimensional window statistic features is constructed.
9. The method for intelligent diagnosis and early warning of driving failure of multi-modal belt conveyor as in claim 6 wherein the specific implementation method of step 3 is to correct dynamic threshold in real time according to equipment load rate and ambient temperature Under the rated working condition, the characteristic value distribution of the historical normal data is counted, and the percentage is used as a reference threshold value Is set to 99%, in Can be obtained through regression of historical data of 1 ten thousand groups of port belt conveyors Ensuring that the threshold correction accords with the actual working condition of the port belt conveyor, and then using a weighted voting mechanism to combine the fault probability of the multi-feature combination A local decision is made.
10. The early warning method of the intelligent diagnosis early warning system for the driving faults of the multi-mode belt conveyor is characterized in that the specific implementation method of the step 4 is that basic probability distribution of each sensor characteristic to each preset fault type is calculated, evidence of a plurality of sensors is synthesized according to a Dempster combination rule or a weighted synthesis rule, and the fault type with the highest confidence after synthesis is selected as a final diagnosis result, so that fault tracing is achieved.

Description

Intelligent diagnosis and early warning system and method for driving faults of multi-mode belt conveyor Technical Field The invention belongs to the technical field of state monitoring and fault diagnosis of a belt conveyor driving system, and particularly relates to an intelligent diagnosis and early warning system and method for a driving fault of a multi-mode belt conveyor. Background In the transportation of large-scale dry bulk ports such as coal, ore, etc., belt conveyors are widely used because of the characteristics of long conveying distance, simple structure, convenient maintenance, strong continuous operation capability, easy realization of programmed control and automatic operation, etc. The destacking equipment and the taking and loading equipment in the port are connected through the belt conveyor, so that long-distance transfer and stacking of operational materials after dumping and material taking, transfer, loading or loading of the stacked materials are realized. The operation of the belt conveyor is mainly driven by the combination of all parts of the driving system, if the parts in the driving system fail during operation, the belt conveyor can stop running, the driving failure needs to be checked, the driving failure can be restarted after the failure is recovered, the failure checking process has a plurality of uncertain factors, the checking time is not predicted, and the production operation efficiency can be seriously influenced if the long-time failure stops. At present, a port belt conveyor driving fault diagnosis system basically depends on a traditional signal analysis or a single machine learning model, accuracy rate is reduced when multiple faults are concurrent under complex working conditions, pre-diagnosis effect is poor, false alarm rate is more than 15%, average fault positioning time is more than 1 hour, most faults can be maintained and recovered aiming at the faults after the faults occur, spare part damage degree is large, spare part repair difficulty is relatively increased, fault maintenance time is increased along with the fault maintenance time, operation efficiency is greatly influenced, and intelligent development requirements of ports cannot be met. Disclosure of Invention The invention aims to overcome the defects of the prior art, and provides an intelligent diagnosis and early warning system and method for a multi-mode belt conveyor driving fault, wherein the diagnosis system collects detection signals such as voltage, current, temperature, vibration, sound and the like of key components in a driving system through a substation acquisition unit, a field sensing unit and an industrial audio acquisition unit, the detection signals are transmitted into an industrial server through an industrial Ethernet, noise reduction of multi-mode signal data, feature fusion matrix construction, dynamic threshold self-adaptive adjustment decision and D-S evidence theory fault tracing are completed in the industrial server, and hierarchical access and use of the system are completed in a client man-machine interaction mode, so that the real-time monitoring and intelligent pre-diagnosis and alarm functions of the belt conveyor driving fault are realized. The invention solves the technical problems by adopting the following technical scheme: A multi-mode belt conveyor driving fault intelligent diagnosis early warning system is characterized by comprising a power substation acquisition unit, a field sensing unit, an industrial audio acquisition unit, an industrial server and a system client, wherein the power substation acquisition unit, the field sensing unit and the industrial audio acquisition unit are respectively connected with the industrial server through an industrial Ethernet, the system client is in communication connection with the industrial server, the power substation acquisition unit, the field sensing unit and the industrial audio acquisition unit respectively conduct data acquisition on current, voltage, vibration, temperature and audio detection signals of key components of a driving system, the acquired signals are transmitted to the industrial server through the industrial Ethernet, the industrial server is used for executing multi-mode signal noise reduction, 15-dimensional basic characteristics and 45-dimensional statistic characteristic fusion matrix construction, and dynamic threshold values are corrected in real time based on load rates and environment temperatures The local decision and D-S evidence theory fault component tracing, and the client side completes final hierarchical access and use of the system through an interactive interface. And the power substation acquisition unit is composed of a current transformer CT for detecting the current of the driving motor and a voltage transformer PT for detecting the voltage of the driving motor in the power substation driving motor incoming line cabinet, and is used for respectively detecting and acquirin