阅读说明：本技术 一种基于振动数据的空压机漏气故障监测方法 (Air compressor air leakage fault monitoring method based on vibration data ) 是由 杨东伟 闫海峰 吴虹霖 岳鹏杰 王浩 牛延博 刘定 张善魁 黄建巍 李�浩 徐晓 于 2021-08-23 设计创作，主要内容包括：本发明公开了一种基于振动数据的空压机漏气故障监测方法,采用上位机和若干组布置在空压机内部管路外壁上且能够移动的故障监测装置,故障监测装置采集管路内振动数据并传输给上位机,上位机将振动数据信息与内部故障状态标签集对比,确定空压机管路的故障状态并发出警报,最后再确定故障位置,本发明可用于监测矿用空压机箱体内部漏气故障,同时能自动判断出故障发生位置,其实时性好,判断准确性高。(The invention discloses an air compressor air leakage fault monitoring method based on vibration data, which comprises the steps of adopting an upper computer and a plurality of groups of fault monitoring devices which are arranged on the outer wall of an internal pipeline of an air compressor and can move, wherein the fault monitoring devices collect vibration data in the pipeline and transmit the vibration data to the upper computer, the upper computer compares vibration data information with an internal fault state label set, determines the fault state of the pipeline of the air compressor and gives an alarm, and finally determines the fault position.)

1. The air compressor air leakage fault monitoring method based on vibration data is characterized by comprising the following steps:

s1, laying a fault monitoring device: the method comprises the following steps that a plurality of groups of fault monitoring devices which are used for detecting the vibration state inside a pipeline and can actively slide along the direction of the pipeline are arranged on the outer wall of a gas pipeline in a box body of the air compressor, and the fault monitoring devices are in communication connection with an upper computer;

s2, the upper computer obtains vibration data information of the air compressor in different fault states and converts the vibration data information into a feature vector set to obtain mapping from the feature vector set to a label set of the air compressor in the fault state;

s3, the fault monitoring device collects vibration data in the pipeline and transmits the vibration data to the upper computer, and the upper computer compares the vibration data information with an internal fault state label set to determine the fault state of the air compressor pipeline and send an alarm;

s4, determining the fault position, which comprises the following steps:

s41, correlation function R of two random signals x (t) and y (t) of each state passing through process_xy(τ) is expressed as:

wherein T is the signal period; y (t + τ) is the signal y (t) time shifted by τ; cross correlation function R_xy(τ) time shift τ corresponding to when (τ) reaches a maximum₀Reflecting the time interval when the correlation degree of the two signals is highest;

s42, determining the propagation speed v of the vibration signal generated at the air leakage point O along the direction of the pipeline₁(ii) a Receiving vibration signal x by instrument at P1 and P2 points at one end of pipeline₁And x₂，x₁And x₂With a receiving time difference of Δ t₁₂The distance between P1 and P2 is Δ d₁₂Vibration signal x₁And x₂Cross correlation functionCorresponding time shift τ when taking maximum₁₂Is Δ t₁₂I.e. Δ t₁₂＝τ₁₂The propagation velocity of the vibration signal in the pipeline is

S43, vibration signal x received by instrument at point P3 at the other end of the pipeline₃And a vibration signal x₁Performing cross-correlation analysis on the cross-correlation functionTime shift τ at maximum₁₃I.e. the time difference between P1 and P3 at which the vibration signal is received, i.e. Δ t₁₃＝τ₁₃The distance between the leak point O and the center line in the direction of the pipe between the distances P1 and P3

2. The air compressor leakage fault monitoring method based on vibration data as claimed in claim 1, wherein the fault monitoring device in step S1 includes a slide rail disposed on the outer wall of the pipeline, the slide rail is provided with a plurality of sliding blocks, the sliding blocks are provided with vibration sensors for acquiring vibration signals in the pipeline, the sliding blocks are provided with rollers engaged with the slide rail and motors for driving the rollers to operate, the motors are connected with a controller, and the controller is in communication connection with an upper computer.

3. The air compressor air leakage fault monitoring method based on vibration data as claimed in claim 1, wherein the specific process of obtaining the mapping from the feature vector set to the air compressor fault state label set in step S2 is as follows:

step 1, collecting vibration data of an air compressor within a first preset time length at a first sampling rate under different fault states of the air compressor to obtain training vibration data of the air compressor under different states;

step 2, dividing training vibration data of the air compressor under each fault state into a plurality of training vibration data groups according to a second preset time length;

step 3, grouping each training vibration data and respectively extracting training characteristic vectors, taking the fault state of the air compressor corresponding to the training characteristic vectors as labels of the training characteristic vectors, and forming a training data set by all the training characteristic vectors with the labels;

and 4, training the machine learning classifier by using the training data set to obtain the mapping from the feature vector set to the air compressor fault state label set.

4. The air compressor air leakage fault monitoring method based on vibration data as claimed in claim 2, wherein the specific method for determining the fault state of the air compressor pipeline in step S3 is as follows:

step 1, collecting vibration data of the air compressor within a third preset time length at a second sampling rate to obtain air compressor detection vibration data;

step 2, dividing the detected vibration data of the air compressor into a plurality of detected vibration data groups according to a second preset time length;

step 3, extracting the detection characteristic vectors of each detected vibration data packet by adopting the method same as the method for extracting the training characteristic vectors in the step S2 to obtain a detection characteristic vector set;

and 4, inputting the detected feature vector set into the trained machine learning classifier, and outputting a fault state label of the air compressor, so that the fault state of the air compressor is determined.

5. The air compressor air leakage fault monitoring method based on vibration data as claimed in claim 1, wherein the fault state in step S3 includes a normal state and an abnormal state, the normal state is normal operation of the air compressor, and the abnormal state includes a vibration abnormal state and an air leakage state.

Technical Field

The invention relates to the field of air compressor air leakage fault monitoring methods based on vibration data, in particular to an air compressor air leakage fault monitoring method based on vibration data.

Background

The air compressor is a compressed gas generating device, is generally used for providing power for pneumatic equipment, and compresses air into gas with a certain pressure, and the gas is connected to the pneumatic equipment to enable the pneumatic equipment to work and operate, so that energy conversion is realized. The compressed gas is safe, reliable and pollution-free, and is widely applied to various links of life and production. In the production process of mines, the air compressor is a power source of pneumatic machinery, can provide air to the underground working face, dilute and remove various harmful gases and mine dust, and is one of important devices for mine production and safety guarantee.

According to incomplete statistics, 95% of mine enterprises in China use compressed air, and continuous production is required, and the combined configuration of multiple groups of air compressors is usually adopted. Because the air compressor machine operational environment noise is big, and the temperature is high, hardly follow the box outside discovery when air leakage trouble takes place for the inside pipeline of air compressor machine box to cause phenomenons such as wearing and tearing, corruption pipeline, wasting of resources, cause huge economic loss to the colliery.

At present, a remote monitoring system for the air compressor is mainly based on PLC control, the air compressor is remotely controlled and managed, some progress is made in the aspect of fault monitoring, and whether the air compressor breaks down or not is monitored by installing pressure sensors on the air tank and the air compressor and installing temperature sensors on an air outlet pipeline of the air tank. However, such a solution has the following disadvantages:

1. because temperature sensor belongs to contact sensor, the spoilage is high, overhauls inconveniently.

2. Because temperature sensor belongs to point contact sensor, can not carry out all-round monitoring to the temperature of air compressor machine, so difficult discovery when non-contact point breaks down, and can't accurate judgement gas leakage fault degree and trouble position, so coal mine enterprise still mainly relies on the manual work to patrol and examine regularly to the fault monitoring of air compressor machine now.

Disclosure of Invention

In view of the technical deficiencies, the invention aims to provide a vibration data-based air compressor air leakage fault monitoring method, wherein a movable vibration sensor is arranged on an inner pipeline of an air compressor box body to acquire vibration conditions inside the air compressor, and the acquired vibration signals are analyzed and processed to acquire whether air leakage faults exist and fault occurrence positions through an algorithm, so that the method is good in effect, low in cost and good in real-time performance.

In order to solve the technical problems, the invention adopts the following technical scheme:

the invention provides an air compressor air leakage fault monitoring method based on vibration data, which specifically comprises the following steps:

s1, laying a fault monitoring device: the method comprises the following steps that a plurality of groups of fault monitoring devices which are used for detecting the vibration state inside a pipeline and can actively slide along the direction of the pipeline are arranged on the outer wall of a gas pipeline in a box body of the air compressor, and the fault monitoring devices are in communication connection with an upper computer;

s2, the upper computer obtains vibration data information of the air compressor in different fault states and converts the vibration data information into a feature vector set to obtain mapping from the feature vector set to a label set of the air compressor in the fault state;

s3, the fault monitoring device collects vibration data in the pipeline and transmits the vibration data to the upper computer, and the upper computer compares the vibration data information with an internal fault state label set to determine the fault state of the air compressor pipeline and send an alarm;

s4, determining the fault position, which comprises the following steps:

s41, correlation function R of two random signals x (t) and y (t) of each state passing through process_xy(τ) is expressed as:

wherein T is the signal period; y (t + τ) is the signal y (t) time shifted by τ; cross correlation function R_xy(τ) time shift τ corresponding to when (τ) reaches a maximum₀Reflecting the time interval when the correlation degree of the two signals is highest;

s42, determining the propagation speed v of the vibration signal generated at the air leakage point O along the direction of the pipeline₁(ii) a Receiving vibration signal x by instrument at P1 and P2 points at one end of pipeline₁And x₂，x₁And x₂With a receiving time difference of Δ t₁₂The distance between P1 and P2 is Δ d_12，Vibration signal x₁And x₂Cross correlation functionCorresponding time shift τ when taking maximum₁₂Is Δ t₁₂I.e. Δ t₁₂＝τ₁₂The propagation velocity of the vibration signal in the pipeline is

S43, vibration signal x received by instrument at point P3 at the other end of the pipeline₃And a vibration signal x₁A cross-correlation analysis is performed and,in the cross-correlation functionTime shift τ at maximum₁₃I.e. the time difference between P1 and P3 at which the vibration signal is received, i.e. Δ t₁₃＝τ₁₃The distance between the leak point O and the center line in the direction of the pipe between the distances P1 and P3

Preferably, fault monitoring device is including establishing slide rail 1 on the pipeline outer wall in step S1, be equipped with a plurality of sliders 2 that can slide on the slide rail 1, be equipped with the vibration sensor 3 that is used for acquireing intraductal vibration signal on the slider 2, be equipped with the gyro wheel that meshes with slide rail 1 and the motor of drive roller work in the slider 2, the motor is connected with the controller, and the controller is connected with the host computer communication.

Preferably, the specific process of obtaining the mapping from the feature vector set to the air compressor fault state tag set in step S2 is as follows:

step 1, collecting vibration data of an air compressor within a first preset time length at a first sampling rate under different fault states of the air compressor to obtain training vibration data of the air compressor under different states;

step 2, dividing training vibration data of the air compressor under each fault state into a plurality of training vibration data groups according to a second preset time length;

step 3, grouping each training vibration data and respectively extracting training characteristic vectors, taking the fault state of the air compressor corresponding to the training characteristic vectors as labels of the training characteristic vectors, and forming a training data set by all the training characteristic vectors with the labels;

and 4, training the machine learning classifier by using the training data set to obtain the mapping from the feature vector set to the air compressor fault state label set.

Preferably, the specific method for determining the fault state of the air compressor pipeline in step S3 is as follows:

step 1, collecting vibration data of the air compressor within a third preset time length at a second sampling rate to obtain air compressor detection vibration data;

step 2, dividing the detected vibration data of the air compressor into a plurality of detected vibration data groups according to a second preset time length;

step 3, extracting the detection characteristic vectors of each detected vibration data packet by adopting the method same as the method for extracting the training characteristic vectors in the step S2 to obtain a detection characteristic vector set;

and 4, inputting the detected feature vector set into the trained machine learning classifier, and outputting a fault state label of the air compressor, so that the fault state of the air compressor is determined.

Preferably, the fault state in step S3 includes a normal state and an abnormal state, the normal state is a normal operation of the air compressor, and the abnormal state includes a vibration abnormal state and an air leakage state.

The invention has the beneficial effects that: according to the invention, the movable vibration sensor is arranged on the pipeline in the air compressor box body to acquire the vibration condition in the air compressor, the acquired vibration signal is analyzed and processed, whether an air leakage fault exists and the fault occurrence position are acquired through an algorithm, the sliding block periodically slides, and the signal is acquired at a fixed point, so that the problem of inaccurate fault identification caused by weakening of the vibration signal due to the fact that the air leakage point is too far away from the sensor can be effectively avoided, the effect is good, the cost is low, and better real-time performance can be achieved.

Drawings

In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings used in the description of the embodiments or the prior art will be briefly described below, it is obvious that the drawings in the following description are only some embodiments of the present invention, and for those skilled in the art, other drawings can be obtained according to the drawings without creative efforts.

Fig. 1 is a schematic view illustrating an installation of a fault monitoring apparatus on an outer wall of a gas pipeline according to an embodiment of the present invention;

fig. 2 is a schematic structural diagram of a fault monitoring apparatus according to an embodiment of the present invention;

description of reference numerals:

1. a slide rail; 2. a slider; 3. a vibration sensor.

Detailed Description

The technical solutions in the embodiments of the present invention will be clearly and completely described below with reference to the drawings in the embodiments of the present invention, and it is obvious that the described embodiments are only a part of the embodiments of the present invention, and not all of the embodiments. All other embodiments, which can be derived by a person skilled in the art from the embodiments given herein without making any creative effort, shall fall within the protection scope of the present invention.

As shown in fig. 1 to 2, a method for monitoring air compressor air leakage fault based on vibration data specifically includes the following steps:

s1, laying a fault monitoring device: the method comprises the following steps that a plurality of groups of fault monitoring devices which are used for detecting the vibration state inside a pipeline and can actively slide along the direction of the pipeline are arranged on the outer wall of a certain section of gas pipeline in a box body of the air compressor, the fault monitoring devices are in communication connection with an upper computer, and the upper computer selects a computer terminal or a mobile terminal;

fault monitoring device is including establishing slide rail 1 on the pipeline outer wall, be equipped with 3 sliders 2 that can slide on the slide rail 1, be equipped with the vibration sensor 3 that is used for obtaining intraductal vibration signal on the slider 2, be equipped with the gyro wheel with 1 meshing of slide rail and the motor of drive roller work in the slider 2, the motor is connected with the controller, the controller is connected with the host computer communication, vibration sensor 3 includes vibration pickup and measuring circuit, wherein vibration pickup is used for the vibration of perception gas pipeline, measuring circuit converts the parameter of engineering vibration into the signal of telecommunication, it shows and the record to transmit into the host computer after electronic circuit enlargies.

S2, the upper computer obtains vibration data information of the air compressor in different fault states and converts the vibration data information into a feature vector set to obtain mapping from the feature vector set to a label set of the air compressor in the fault state;

the specific process of obtaining the mapping from the feature vector set to the air compressor fault state label set comprises the following steps:

step 1, collecting vibration data of an air compressor within a first preset time length at a first sampling rate under different fault states of the air compressor to obtain training vibration data of the air compressor under different states;

step 2, dividing training vibration data of the air compressor under each fault state into a plurality of training vibration data groups according to a second preset time length;

step 3, grouping each training vibration data and respectively extracting training characteristic vectors, taking the fault state of the air compressor corresponding to the training characteristic vectors as labels of the training characteristic vectors, and forming a training data set by all the training characteristic vectors with the labels;

and 4, training the machine learning classifier by using the training data set to obtain the mapping from the feature vector set to the air compressor fault state label set.

S3, the fault monitoring device collects vibration data in the pipeline and transmits the vibration data to the upper computer, and the upper computer compares the vibration data information with an internal fault state label set to determine the fault state of the air compressor pipeline and send an alarm;

the specific method for determining the fault state of the air compressor pipeline comprises the following steps:

step 1, collecting vibration data of the air compressor within a third preset time length at a second sampling rate to obtain air compressor detection vibration data;

step 2, dividing the detected vibration data of the air compressor into a plurality of detected vibration data groups according to a second preset time length;

step 3, extracting the detection characteristic vectors of each detected vibration data packet by adopting the method same as the method for extracting the training characteristic vectors in the step S2 to obtain a detection characteristic vector set;

step 4, inputting the detected feature vector set into the trained machine learning classifier, and outputting a fault state label of the air compressor, so as to determine the fault state of the air compressor;

the fault state comprises a normal state and an abnormal state, the normal state is the normal operation of the air compressor, and the abnormal state comprises a vibration abnormal state and an air leakage state.

S4, determining the fault position, which comprises the following steps:

s41, correlation function R of two random signals x (t) and y (t) of each state passing through process_xy(τ) is expressed as:

wherein T is the signal period; y (t + τ) is the signal y (t) time shifted by τ; cross correlation function R_xy(τ) time shift τ corresponding to when (τ) reaches a maximum₀Reflecting the time interval when the correlation degree of the two signals is highest;

s42, determining the propagation speed v of the vibration signal generated at the air leakage point O along the direction of the pipeline₁(ii) a At the positions of two vibration sensors 3 (P1, P2) at one end of the pipeline, a vibration signal x is received by an instrument₁And x₂，x₁And x₂With a receiving time difference of Δ t₁₂The distance between P1 and P2 is Δ d₁₂Vibration signal x₁And x₂Cross correlation functionCorresponding time shift τ when taking maximum₁₂Is Δ t₁₂I.e. Δ t₁₂＝τ₁₂The propagation velocity of the vibration signal in the pipeline is

S43, using the vibration signal x received by the instrument at the position of the third vibration sensor 3 at the other end of the pipeline (point P3)₃And a vibration signal x₁Performing cross-correlation analysis on the cross-correlation functionTime shift τ at maximum₁₃I.e. the time difference between P1 and P3 at which the vibration signal is received, i.e. Δ t₁₃＝τ₁₃The distance between the leak point O and the center line in the direction of the pipe between the distances P1 and P3

The built-in vibration signal fault identification of having and position location algorithm model in the host computer, it is through the reciprocating motion of control slider 2, vibration sensor 3 on the slider 2 acquires the pipeline vibration signal and transmits for the host computer simultaneously, the host computer carries out analysis and discernment to the signal that comes from vibration sensor 3, discern the pipeline trouble through machine learning, obtain the propagation velocity of vibration signal on the pipeline of place through relevant analysis, and then obtain the air compressor machine fault location, then deem the air compressor machine to be in abnormal operating condition when the change of vibration signal satisfies the vibration change characteristic that sets for.

The method can be used for monitoring the air leakage fault of the gas pipeline of the mining air compressor, and can automatically judge the occurrence position of the air leakage fault of the pipeline, so that the method is good in real-time performance and accurate in judgment.

In the second embodiment, in step S1, 2 sliders are disposed on the slide rail 1, 2 vibration sensors 3 are disposed on the first slider 2, the positions of the 2 vibration sensors 3 are points P1 and P2, the position of the third vibration sensor 3 is disposed on the second slider 2, the position of the vibration sensor 3 is point P3, the 2 vibration sensors 3 at points P1 and P2 are used to determine the propagation speed of the vibration signal, and the two vibration sensors 3 at positions P3 and P1 are used to locate the fault occurrence position in cooperation, and the rest steps are the same as those in embodiment 1.

It will be apparent to those skilled in the art that various changes and modifications may be made in the present invention without departing from the spirit and scope of the invention. Thus, if such modifications and variations of the present invention fall within the scope of the claims of the present invention and their equivalents, the present invention is also intended to include such modifications and variations.