CN111707355A - Equipment operation state detection method, device, equipment and storage medium - Google Patents
Equipment operation state detection method, device, equipment and storage medium Download PDFInfo
- Publication number
- CN111707355A CN111707355A CN202010565782.9A CN202010565782A CN111707355A CN 111707355 A CN111707355 A CN 111707355A CN 202010565782 A CN202010565782 A CN 202010565782A CN 111707355 A CN111707355 A CN 111707355A
- Authority
- CN
- China
- Prior art keywords
- sound data
- deviation error
- training
- fault diagnosis
- diagnosis model
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Pending
Links
- 238000001514 detection method Methods 0.000 title claims description 25
- 238000003745 diagnosis Methods 0.000 claims abstract description 89
- 238000000034 method Methods 0.000 claims abstract description 56
- 230000008569 process Effects 0.000 claims abstract description 28
- 238000012549 training Methods 0.000 claims description 87
- 238000012545 processing Methods 0.000 claims description 44
- 230000007613 environmental effect Effects 0.000 claims description 23
- 230000001174 ascending effect Effects 0.000 claims description 5
- 238000004364 calculation method Methods 0.000 claims description 5
- 238000001914 filtration Methods 0.000 claims description 5
- 238000004590 computer program Methods 0.000 claims description 3
- 238000004891 communication Methods 0.000 description 7
- 238000010586 diagram Methods 0.000 description 6
- 238000013145 classification model Methods 0.000 description 4
- 230000006870 function Effects 0.000 description 4
- 238000012544 monitoring process Methods 0.000 description 4
- 230000005856 abnormality Effects 0.000 description 3
- 230000009471 action Effects 0.000 description 3
- 238000012423 maintenance Methods 0.000 description 3
- 241001465754 Metazoa Species 0.000 description 2
- 230000002159 abnormal effect Effects 0.000 description 2
- 230000008901 benefit Effects 0.000 description 2
- 238000007689 inspection Methods 0.000 description 2
- 238000009825 accumulation Methods 0.000 description 1
- 238000013459 approach Methods 0.000 description 1
- 238000007664 blowing Methods 0.000 description 1
- 238000010276 construction Methods 0.000 description 1
- 230000000694 effects Effects 0.000 description 1
- 238000005516 engineering process Methods 0.000 description 1
- 238000000605 extraction Methods 0.000 description 1
- 238000004519 manufacturing process Methods 0.000 description 1
- 239000011159 matrix material Substances 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 238000003062 neural network model Methods 0.000 description 1
- 230000000750 progressive effect Effects 0.000 description 1
- 230000011218 segmentation Effects 0.000 description 1
Images
Classifications
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01H—MEASUREMENT OF MECHANICAL VIBRATIONS OR ULTRASONIC, SONIC OR INFRASONIC WAVES
- G01H17/00—Measuring mechanical vibrations or ultrasonic, sonic or infrasonic waves, not provided for in the preceding groups
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01M—TESTING STATIC OR DYNAMIC BALANCE OF MACHINES OR STRUCTURES; TESTING OF STRUCTURES OR APPARATUS, NOT OTHERWISE PROVIDED FOR
- G01M99/00—Subject matter not provided for in other groups of this subclass
- G01M99/005—Testing of complete machines, e.g. washing-machines or mobile phones
-
- G—PHYSICS
- G10—MUSICAL INSTRUMENTS; ACOUSTICS
- G10L—SPEECH ANALYSIS TECHNIQUES OR SPEECH SYNTHESIS; SPEECH RECOGNITION; SPEECH OR VOICE PROCESSING TECHNIQUES; SPEECH OR AUDIO CODING OR DECODING
- G10L25/00—Speech or voice analysis techniques not restricted to a single one of groups G10L15/00 - G10L21/00
- G10L25/48—Speech or voice analysis techniques not restricted to a single one of groups G10L15/00 - G10L21/00 specially adapted for particular use
- G10L25/51—Speech or voice analysis techniques not restricted to a single one of groups G10L15/00 - G10L21/00 specially adapted for particular use for comparison or discrimination
-
- G—PHYSICS
- G10—MUSICAL INSTRUMENTS; ACOUSTICS
- G10L—SPEECH ANALYSIS TECHNIQUES OR SPEECH SYNTHESIS; SPEECH RECOGNITION; SPEECH OR VOICE PROCESSING TECHNIQUES; SPEECH OR AUDIO CODING OR DECODING
- G10L25/00—Speech or voice analysis techniques not restricted to a single one of groups G10L15/00 - G10L21/00
- G10L25/48—Speech or voice analysis techniques not restricted to a single one of groups G10L15/00 - G10L21/00 specially adapted for particular use
- G10L25/72—Speech or voice analysis techniques not restricted to a single one of groups G10L15/00 - G10L21/00 specially adapted for particular use for transmitting results of analysis
Landscapes
- Engineering & Computer Science (AREA)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Computational Linguistics (AREA)
- Signal Processing (AREA)
- Health & Medical Sciences (AREA)
- Audiology, Speech & Language Pathology (AREA)
- Human Computer Interaction (AREA)
- Acoustics & Sound (AREA)
- Multimedia (AREA)
- Measurement Of Mechanical Vibrations Or Ultrasonic Waves (AREA)
Abstract
The application discloses a method, a device, equipment and a storage medium for detecting the running state of the equipment, wherein a fault diagnosis model is configured in advance, the fault diagnosis model can process sound data collected by the equipment and output deviation errors for measuring the degree of the sound data deviating from the normal running state, on the basis, the application can acquire the sound data collected by target equipment, and further process the sound data by using the fault diagnosis model to obtain the deviation errors for measuring the degree of the sound data deviating from the normal running state, and further can determine the running state of the target equipment based on the deviation errors and a preset deviation error threshold value, namely determine whether the target equipment breaks down. Based on this, the running state of the target equipment can be detected efficiently and timely, namely, the fault can be found timely when the target equipment is in fault, and the loss caused by the fault duration time is reduced.
Description
Technical Field
The present disclosure relates to the field of industrial detection technologies, and in particular, to a method, an apparatus, a device, and a storage medium for detecting an operating state of a device.
Background
The equipment running state detection is mainly used for detecting the running state of the equipment, such as whether the equipment runs normally or has fault abnormality. Taking industrial equipment as an example, if the industrial equipment has a fault abnormality, decision and maintenance need to be carried out in time, so that the equipment maintenance efficiency of industrial enterprises is improved, and the abnormal shutdown loss of a production line is reduced.
In the prior art, an experienced device operation and maintenance worker typically inspects the industrial device to determine whether the industrial device has a fault abnormality. However, because industrial equipment in the internet of things is numerous, and the operating conditions of different industrial equipment are various, the manual inspection is not only inefficient, but also has the problem that the abnormal fault is not found timely.
Disclosure of Invention
In view of the above problems, the present application is provided to provide a method, an apparatus, a device and a storage medium for detecting an operating state of a device, so as to solve the problems of low efficiency and untimely fault exception discovery in the conventional manual inspection. The specific scheme is as follows:
an equipment running state detection method comprises the following steps:
acquiring sound data collected by target equipment;
processing the sound data by using a preset fault diagnosis model, and outputting a deviation error for measuring the degree of the sound data deviating from a normal operation state;
and determining the running state of the target equipment based on the deviation error and a preset deviation error threshold value.
Preferably, the training process of the fault diagnosis model includes:
acquiring training sound data of the target equipment in a normal operation state;
processing the training sound data by using a fault diagnosis model to obtain a deviation error for measuring the deviation degree of the training sound data from a normal running state;
and training the fault diagnosis model by taking the deviation error approaching zero as a training target to obtain the trained fault diagnosis model.
Preferably, the processing the training sound data by using the fault diagnosis model to obtain a deviation error for measuring the degree of deviation of the training sound data from a normal operation state includes:
extracting acoustic features of the training sound data;
and processing the acoustic features by using a fault diagnosis model to obtain hidden layer reconstruction features, and calculating a deviation error between the acoustic features and the hidden layer reconstruction features, wherein the deviation error is used for measuring the degree of the training sound data deviating from a normal running state.
Preferably, the processing the sound data by using a preset fault diagnosis model and outputting a deviation error for measuring the degree of the sound data deviating from a normal operation state includes:
receiving acoustic features of the acoustic data using an input layer of a fault diagnosis model;
processing the acoustic features and generating hidden layer reconstruction features by utilizing a feature processing layer of a fault diagnosis model;
and calculating the deviation error between the acoustic characteristic and the hidden layer reconstruction characteristic by utilizing a deviation error calculation layer of the fault diagnosis model, wherein the deviation error is used for measuring the degree of the sound data deviating from the normal operation state.
Preferably, the determining of the deviation error threshold includes:
processing each piece of training sound data by using a preset fault diagnosis model to obtain a deviation error corresponding to each piece of training sound data, wherein the training sound data is the collected sound of the target equipment in a normal operation state;
arranging deviation errors of all pieces of training sound data in an ascending order or a descending order according to the sizes;
and selecting a target deviation error in the deviation error sequence as a deviation error threshold according to the recall rate or the false alarm rate set by the user.
Preferably, when the operation state of the target device is determined to be a fault, the method further includes:
based on the sound data, a fault type of the target device is identified.
Preferably, the acquiring sound data collected by the target device includes:
acquiring original sound data acquired by target equipment;
and filtering the original sound data according to the specific frequency band where the target equipment emits the sound to obtain the sound data of the specific frequency band.
Preferably, the method further comprises the following steps:
taking the sound data of other frequency bands except the specific frequency band in the original sound data as environmental sound data;
and carrying out sound type identification on the environmental sound data to obtain the target type environmental sound contained in the environmental sound data.
An apparatus operation state detection device comprising:
the voice data acquisition unit is used for acquiring voice data acquired by the target equipment;
the deviation error determining unit is used for processing the sound data by utilizing a preset fault diagnosis model and outputting a deviation error for measuring the degree of the sound data deviating from a normal running state;
and the state determining unit is used for determining the running state of the target equipment based on the deviation error and a preset deviation error threshold value.
Preferably, the method further comprises the following steps: a deviation error threshold determination unit for determining a deviation error threshold, the process may include:
processing each piece of training sound data by using a preset fault diagnosis model to obtain a deviation error corresponding to each piece of training sound data, wherein the training sound data is the collected sound of the target equipment in a normal operation state;
arranging deviation errors of all pieces of training sound data in an ascending order or a descending order according to the sizes;
and selecting a target deviation error in the deviation error sequence as a deviation error threshold according to the recall rate or the false alarm rate set by the user.
Preferably, the method further comprises the following steps:
and the fault type determining unit is used for identifying the fault type of the target equipment based on the sound data when the running state of the target equipment is determined to be the fault.
Preferably, the process of acquiring the sound data acquired by the sound data acquiring unit for the target device may include:
acquiring original sound data acquired by target equipment;
and filtering the original sound data according to the specific frequency band where the target equipment emits the sound to obtain the sound data of the specific frequency band.
Preferably, the method further comprises the following steps: an ambient sound type determination unit for:
taking the sound data of other frequency bands except the specific frequency band in the original sound data as environmental sound data; and carrying out sound type identification on the environmental sound data to obtain the target type environmental sound contained in the environmental sound data.
An apparatus operation state detection apparatus comprising: a memory and a processor;
the memory is used for storing programs;
the processor is configured to execute the program to implement the steps of the above method for detecting the operating state of the device.
A storage medium having stored thereon a computer program which, when executed by a processor, implements the steps of the above method of detecting an operational state of a device.
By means of the technical scheme, the fault diagnosis model is configured in advance, the fault diagnosis model can process sound data collected by the equipment and output deviation errors for measuring the degree of the sound data deviating from the normal operation state, on the basis, the sound data collected by the target equipment can be obtained, the sound data are further processed by the fault diagnosis model, the deviation errors for measuring the degree of the sound data deviating from the normal operation state are obtained, and the operation state of the target equipment can be determined further based on the deviation errors and a preset deviation error threshold value, namely whether the target equipment breaks down or not is determined. Based on this, the running state of the target equipment can be detected efficiently and timely, namely, the fault can be found timely when the target equipment is in fault, and the loss caused by the fault duration time is reduced.
Further, because the industrial equipment is operated in a normal state in most of time and only a very small amount of time fails, the fault diagnosis model determines whether the target equipment fails or not by outputting deviation errors for measuring the degree of the sound data deviating from the normal operation state, namely, the normal operation state of the equipment is taken as reference, so that the actual operation state rule of the equipment is better met, and the finally obtained operation state of the target equipment is more accurate.
Drawings
Various other advantages and benefits will become apparent to those of ordinary skill in the art upon reading the following detailed description of the preferred embodiments. The drawings are only for purposes of illustrating the preferred embodiments and are not to be construed as limiting the application. Also, like reference numerals are used to refer to like parts throughout the drawings. In the drawings:
fig. 1 is a schematic flow chart of a method for detecting an operating state of a device according to an embodiment of the present disclosure;
FIG. 2 illustrates a schematic diagram of a training process for a fault diagnosis model;
fig. 3 is a schematic structural diagram of an apparatus operation state detection device according to an embodiment of the present disclosure;
fig. 4 is a schematic structural diagram of a detection apparatus provided in an embodiment of the present application.
Detailed Description
The technical solutions in the embodiments of the present application will be clearly and completely described below with reference to the drawings in the embodiments of the present application, and it is obvious that the described embodiments are only a part of the embodiments of the present application, 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 application.
The application provides an equipment running state detection scheme, which can timely and accurately detect the running state of target equipment, namely find whether the target equipment breaks down.
The scheme can be realized based on a terminal with data processing capacity, and the terminal can be a mobile phone, a computer, a server, a cloud terminal and the like.
Next, as described with reference to fig. 1, the method for detecting the operating state of the device of the present application may include the following steps:
and step S100, acquiring sound data collected by the target equipment.
Specifically, in order to implement detection on the target device, at least one monitoring point may be set at the target device in advance, and the monitoring point is provided with a sensor for collecting sound data from the target device.
The number of the monitoring points and the setting direction of the target equipment can be set according to needs, and the sensors at the monitoring points in different directions are used for acquiring sound data transmitted by the target equipment in corresponding directions. The sensor can adopt a microphone, a microphone matrix or other sensors with sound receiving capability.
And step S110, processing the sound data by using a preset fault diagnosis model, and outputting a deviation error for measuring the degree of the sound data deviating from the normal operation state.
Specifically, in the embodiment of the present application, a fault diagnosis model may be trained in advance, and the fault diagnosis model may process input sound data and output a deviation error, where the deviation error is used to measure a degree of deviation of the input sound data from a normal operation state of a target device.
It can be understood that, because the industrial equipment is operated in the normal state most of the time, and only a very small amount of time is in failure, the failure diagnosis model of the application determines whether the target equipment is in failure by outputting a deviation error for measuring the degree of the sound data deviating from the normal operation state, that is, the failure diagnosis model more conforms to the law of the actual operation state of the equipment by taking the normal operation state of the equipment as a reference.
And step S120, determining the running state of the target equipment based on the deviation error and a preset deviation error threshold value.
Specifically, the deviation error threshold value may be configured in advance according to user needs, and after the deviation error is determined in the previous step, the magnitude relationship between the deviation error and the deviation error threshold value may be compared, if the deviation error threshold value is exceeded, it is indicated that the degree of deviation of the sound data from the normal operation state is too serious, that is, it indicates that the target device has a fault, and if the deviation error threshold value is not exceeded, it indicates that the target device is operating normally.
The method comprises the steps that a fault diagnosis model is configured in advance, the fault diagnosis model can process sound data collected by equipment and output deviation errors for measuring the degree of the sound data deviating from the normal operation state, on the basis, the sound data collected by target equipment can be obtained, the sound data are further processed by the fault diagnosis model, the deviation errors for measuring the degree of the sound data deviating from the normal operation state are obtained, further, the operation state of the target equipment can be determined based on the deviation errors and preset deviation error thresholds, and whether the target equipment breaks down or not is determined. Based on this, the running state of the target equipment can be detected efficiently and timely, namely, the fault can be found timely when the target equipment is in fault, and the loss caused by the fault duration time is reduced.
Further, because the industrial equipment is operated in a normal state in most of time and only a very small amount of time fails, the fault diagnosis model determines whether the target equipment fails or not by outputting deviation errors for measuring the degree of the sound data deviating from the normal operation state, namely, the normal operation state of the equipment is taken as reference, so that the actual operation state rule of the equipment is better met, and the finally obtained operation state of the target equipment is more accurate.
Optionally, in step S120, when the operating state of the target device is determined to be a fault based on the deviation error and a preset deviation error threshold, the following processing steps may be further added in the scheme of the application:
based on the sound data, a fault type of the target device is identified.
Specifically, in the present application, the sound data of the target device during each type of fault may be collected in advance, and then the fault type classification model may be trained based on the sound data of each type of fault, so as to obtain the trained fault type classification model.
On the basis, the sound data collected by the target equipment can be sent to the fault type classification model to obtain the specific fault type output by the model.
It can be understood that, with the accumulation of time, the method and the device can continuously collect various types of faults which may occur to the target device, for example, for an unknown fault, the unknown fault can be calibrated by an expert, and then a fault type classification model is iteratively trained by using data calibrated by the expert, so that the model can be continuously optimized.
In another embodiment of the present application, a process for training a fault diagnosis model is described.
It can be understood that, since the industrial equipment is in normal operation for most of the time, and only a very small amount of time is in failure, most of the sound data collected by the target equipment is sound data of the target equipment in normal operation, and only a very small amount of sound data in failure state is available. In order to avoid the problem of poor model training effect caused by imbalance of the positive sample and the negative sample, a scheme for training the fault diagnosis model based on normal state sound data is creatively provided in the embodiment, that is, only the positive sample is used for training the fault diagnosis model. Next, a training process of the fault diagnosis model is described, which may include:
and S1, acquiring training sound data when the target equipment is in a normal operation state.
And S2, processing the training sound data by using a fault diagnosis model to obtain deviation errors for measuring the deviation degree of the training sound data from the normal operation state.
Specifically, the fault diagnosis model may be a deep neural network model that is capable of processing input training sound data and generating a deviation error that measures the degree to which the training sound data deviates from a normal operating state.
And S3, training the fault diagnosis model by taking the deviation error approaching zero as a training target to obtain the trained fault diagnosis model.
It can be understood that, for the training sound data in the normal state, the degree of deviation from the normal operation state cannot be too large, and it is expected that the smaller deviation error of the model for the training sound data is, the better, that is, the deviation error can be taken as a training target, the fault diagnosis model is trained, and finally the trained fault diagnosis model is obtained.
Optionally, it may be set that when the deviation errors of the fault diagnosis model for the output of the consecutive N pieces of training sound data approach to be stable, for example, when the difference between the deviation errors for the output of the consecutive N pieces of training sound data does not exceed the set difference threshold, the fault diagnosis model may be considered to end the training.
For S2, the process of processing the training sound data by using the fault diagnosis model to obtain the deviation error for measuring the deviation degree of the training sound data from the normal operation state may be implemented as follows:
and S21, extracting the acoustic features of the training sound data.
And S22, processing the acoustic features by using a fault diagnosis model to obtain hidden layer reconstruction features, and calculating deviation errors between the acoustic features and the hidden layer reconstruction features, wherein the deviation errors are used for measuring the degree of the training sound data deviating from a normal operation state.
Specifically, the fault diagnosis model can learn some key features of the normal state sound data based on the acoustic features of the normal state training sound data, and perform feature reconstruction based on the key features, that is, obtain hidden layer reconstruction features, and further calculate a deviation error between the acoustic features and the hidden layer reconstruction features by using the hidden layer reconstruction features as a reference, wherein the deviation error measures the degree of deviation of the training sound data from a normal operation state.
Optionally, for the above-mentioned offset error LOSS, it may choose a mean square error MSE, that is:
wherein Y represents an acoustic feature of the training sound data,
representing the hidden layer reconstruction feature.
Referring to FIG. 2, a schematic diagram of a training process for a fault diagnosis model is illustrated.
The method comprises the steps of carrying out segmentation processing on collected sound data to obtain a plurality of audio segments, carrying out acoustic feature extraction on each audio segment, sending the extracted acoustic features to a feature processing layer of a fault diagnosis model, wherein the feature processing layer can comprise a plurality of convolution coding layers, rebuilding the acoustic features, and finally obtaining rebuilt hidden layer rebuilding features. And finally, calculating the deviation error between the acoustic characteristic and the hidden layer reconstruction characteristic by a deviation error calculation layer of the fault diagnosis model.
Further, based on the trained fault diagnosis model, the sound data may be processed, and a deviation error for measuring a degree of deviation of the sound data from a normal operation state may be output, where the process specifically includes:
receiving acoustic features of the acoustic data using an input layer of a fault diagnosis model.
And processing the acoustic features and generating hidden layer reconstruction features by utilizing a feature processing layer of the fault diagnosis model.
And calculating the deviation error between the acoustic characteristic and the hidden layer reconstruction characteristic by utilizing a deviation error calculation layer of the fault diagnosis model, wherein the deviation error is used for measuring the degree of the sound data deviating from the normal operation state.
Still further, the process of determining the preset deviation error threshold is described in conjunction with the above description of the fault diagnosis model training process.
Based on the training process of the fault diagnosis model introduced in the above embodiment, a trained fault diagnosis model can be obtained. On the basis, the fault diagnosis model can be used for processing each piece of training sound data, and further obtaining the deviation error corresponding to each piece of training sound data. It should be noted that the training sound data described herein is a sound acquired when the target device is in a normal operation state.
Further, the deviation errors of the training audio data may be sorted in ascending order or descending order according to magnitude to obtain a deviation error sequence.
And finally, selecting a target deviation error in the deviation error sequence as a deviation error threshold according to the recall rate or the false alarm rate set by the user.
Specifically, the recall rate is the lowest probability set by the user for the correct diagnosis of the running state of the equipment, and the false alarm rate is the highest probability set by the user for misdiagnosing normal state sound data as fault sound data.
Taking the deviation error sequence as a descending sorting result as an example, the process of determining the deviation error threshold value according to the false alarm rate is explained as follows:
TABLE 1
Assuming that 1000 pieces of data are common in the training sound data set, table 1 illustrates the deviation errors of the first 13 pieces of training sound data in descending order.
Defining the false alarm rate set by a user to be 1%, determining that the target equipment has a fault when the deviation error exceeds the deviation error threshold value, namely performing fault alarm, and when the false alarm rate is 1%, generating 1000 × 1% to 10 training sound data in the training sound data set, and performing fault diagnosis by the fault diagnosis model to be fault alarm.
Therefore, the 10 th deviation error in descending order may be chosen as the deviation error threshold. The model may be diagnosed as faulty for training sound data having a deviation error greater than the deviation error threshold, and normal for training sound data having a deviation error less than the deviation error threshold.
In another embodiment of the present application, an alternative implementation of the step S100 for acquiring the sound data collected by the target device is introduced.
In an alternative manner, the raw sound data collected by the target device may be obtained, for example, the raw sound data collected by a sensor disposed at the target device may be obtained, and the raw sound data may be used as the sound data collected by the target device for subsequent processing.
In another optional manner, in this embodiment, the original sound data may be further subjected to filtering processing, that is:
it is considered that the raw sound data collected by the sensor may include sounds emitted by other objects in the complex environment, such as environmental sounds like thunderstorms, wind-blowing sounds, human voices, construction site sounds, animal calls, vehicle traffic sounds, and the like. This part of the ambient sound may interfere with the detection of the operational status of the target device.
In this embodiment, considering that the frequency band of the sound emitted by the target device may be different from that of other environmental sounds, based on this, the original sound data may be filtered according to the specific frequency band in which the sound emitted by the target device is located, so as to obtain the sound data in the specific frequency band. Therefore, part of the environmental sound can be filtered, and the accuracy of the subsequent detection of the running state of the target equipment based on the sound data is improved.
The sound data of the other frequency bands except the specific frequency band in the original sound data can be used as the environmental sound data.
Furthermore, the sound type of the environmental sound data can be identified, and the target type environmental sound contained in the environmental sound data is obtained.
On the basis, the obtained target type environment sound can be prompted to the user, so that the user can know the condition of the environment around the target device, for example, the user can know whether animals exist around the target device or whether people enter the target device.
The device operating state detection apparatus provided in the embodiment of the present application is described below, and the device operating state detection apparatus described below and the device operating state detection method described above may be referred to in correspondence with each other.
Referring to fig. 3, fig. 3 is a schematic structural diagram of an apparatus operation state detection device disclosed in the embodiment of the present application.
As shown in fig. 3, the apparatus may include:
a sound data acquisition unit 11 configured to acquire sound data acquired by a target device;
a deviation error determination unit 12, configured to process the sound data by using a preset fault diagnosis model, and output a deviation error for measuring a degree of deviation of the sound data from a normal operating state;
a state determination unit 13, configured to determine an operation state of the target device based on the deviation error and a preset deviation error threshold.
The device running state detection device provided by the embodiment of the application can acquire the sound data collected by the target device, further process the sound data by using the fault diagnosis model, obtain the deviation error for measuring the degree of the sound data deviating from the normal running state, and further determine the running state of the target device based on the deviation error and the preset deviation error threshold value, namely determine whether the target device breaks down. Based on this, the running state of the target equipment can be detected efficiently and timely, namely, the fault can be found timely when the target equipment is in fault, and the loss caused by the fault duration time is reduced.
Further, because the industrial equipment is operated in a normal state in most of time and only a very small amount of time fails, the fault diagnosis model determines whether the target equipment fails or not by outputting deviation errors for measuring the degree of the sound data deviating from the normal operation state, namely, the normal operation state of the equipment is taken as reference, so that the actual operation state rule of the equipment is better met, and the finally obtained operation state of the target equipment is more accurate.
Optionally, the apparatus of the present application may further include: the model training unit is used for training a fault diagnosis model, and specifically may include:
acquiring training sound data of the target equipment in a normal operation state;
processing the training sound data by using a fault diagnosis model to obtain a deviation error for measuring the deviation degree of the training sound data from a normal running state;
and training the fault diagnosis model by taking the deviation error approaching zero as a training target to obtain the trained fault diagnosis model.
Optionally, the process of processing the training sound data by the model training unit using the fault diagnosis model to obtain a deviation error for measuring a degree of deviation of the training sound data from a normal operating state may include:
extracting acoustic features of the training sound data;
and processing the acoustic features by using a fault diagnosis model to obtain hidden layer reconstruction features, and calculating a deviation error between the acoustic features and the hidden layer reconstruction features, wherein the deviation error is used for measuring the degree of the training sound data deviating from a normal running state.
Optionally, the process of processing the sound data by the deviation error determination unit using a preset fault diagnosis model and outputting a deviation error for measuring a degree of deviation of the sound data from a normal operating state may include:
receiving acoustic features of the acoustic data using an input layer of a fault diagnosis model;
processing the acoustic features and generating hidden layer reconstruction features by utilizing a feature processing layer of a fault diagnosis model;
and calculating the deviation error between the acoustic characteristic and the hidden layer reconstruction characteristic by utilizing a deviation error calculation layer of the fault diagnosis model, wherein the deviation error is used for measuring the degree of the sound data deviating from the normal operation state.
Optionally, the apparatus of the present application may further include: a deviation error threshold determination unit for determining a deviation error threshold, the process may include:
processing each piece of training sound data by using a preset fault diagnosis model to obtain a deviation error corresponding to each piece of training sound data, wherein the training sound data is the collected sound of the target equipment in a normal operation state;
arranging deviation errors of all pieces of training sound data in an ascending order or a descending order according to the sizes;
and selecting a target deviation error in the deviation error sequence as a deviation error threshold according to the recall rate or the false alarm rate set by the user.
Optionally, the apparatus of the present application may further include:
and the fault type determining unit is used for identifying the fault type of the target equipment based on the sound data when the running state of the target equipment is determined to be the fault.
Optionally, the process of acquiring the sound data acquired by the target device by the sound data acquiring unit may include:
acquiring original sound data acquired by target equipment;
and filtering the original sound data according to the specific frequency band where the target equipment emits the sound to obtain the sound data of the specific frequency band.
Optionally, the apparatus of the present application may further include: an ambient sound type determination unit for:
taking the sound data of other frequency bands except the specific frequency band in the original sound data as environmental sound data; and carrying out sound type identification on the environmental sound data to obtain the target type environmental sound contained in the environmental sound data.
The device running state detection device provided by the embodiment of the application can be applied to detection equipment, such as a terminal: mobile phones, computers, etc. Optionally, fig. 4 shows a block diagram of a hardware structure of the detection device, and referring to fig. 4, the hardware structure of the detection device may include: at least one processor 1, at least one communication interface 2, at least one memory 3 and at least one communication bus 4;
in the embodiment of the application, the number of the processor 1, the communication interface 2, the memory 3 and the communication bus 4 is at least one, and the processor 1, the communication interface 2 and the memory 3 complete mutual communication through the communication bus 4;
the processor 1 may be a central processing unit CPU, or an application specific Integrated circuit asic, or one or more Integrated circuits configured to implement embodiments of the present invention, etc.;
the memory 3 may include a high-speed RAM memory, and may further include a non-volatile memory (non-volatile memory) or the like, such as at least one disk memory;
wherein the memory stores a program and the processor can call the program stored in the memory, the program for:
acquiring sound data collected by target equipment;
processing the sound data by using a preset fault diagnosis model, and outputting a deviation error for measuring the degree of the sound data deviating from a normal operation state;
and determining the running state of the target equipment based on the deviation error and a preset deviation error threshold value.
Alternatively, the detailed function and the extended function of the program may be as described above.
Embodiments of the present application further provide a storage medium, where a program suitable for execution by a processor may be stored, where the program is configured to:
acquiring sound data collected by target equipment;
processing the sound data by using a preset fault diagnosis model, and outputting a deviation error for measuring the degree of the sound data deviating from a normal operation state;
and determining the running state of the target equipment based on the deviation error and a preset deviation error threshold value.
Alternatively, the detailed function and the extended function of the program may be as described above.
Finally, it should also be noted that, herein, relational terms such as first and second, and the like may be used solely to distinguish one entity or action from another entity or action without necessarily requiring or implying any actual such relationship or order between such entities or actions. Also, the terms "comprises," "comprising," or any other variation thereof, are intended to cover a non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements does not include only those elements but may include other elements not expressly listed or inherent to such process, method, article, or apparatus. Without further limitation, an element defined by the phrase "comprising an … …" does not exclude the presence of other identical elements in a process, method, article, or apparatus that comprises the element.
The embodiments in the present description are described in a progressive manner, each embodiment focuses on differences from other embodiments, the embodiments may be combined as needed, and the same and similar parts may be referred to each other.
The previous description of the disclosed embodiments is provided to enable any person skilled in the art to make or use the present application. Various modifications to these embodiments will be readily apparent to those skilled in the art, and the generic principles defined herein may be applied to other embodiments without departing from the spirit or scope of the application. Thus, the present application is not intended to be limited to the embodiments shown herein but is to be accorded the widest scope consistent with the principles and novel features disclosed herein.
Claims (11)
1. An equipment operation state detection method is characterized by comprising the following steps:
acquiring sound data collected by target equipment;
processing the sound data by using a preset fault diagnosis model, and outputting a deviation error for measuring the degree of the sound data deviating from a normal operation state;
and determining the running state of the target equipment based on the deviation error and a preset deviation error threshold value.
2. The method of claim 1, wherein the training process of the fault diagnosis model comprises:
acquiring training sound data of the target equipment in a normal operation state;
processing the training sound data by using a fault diagnosis model to obtain a deviation error for measuring the deviation degree of the training sound data from a normal running state;
and training the fault diagnosis model by taking the deviation error approaching zero as a training target to obtain the trained fault diagnosis model.
3. The method of claim 2, wherein processing the training acoustic data using a fault diagnosis model to derive a deviation error that measures how far the training acoustic data deviates from normal operating conditions comprises:
extracting acoustic features of the training sound data;
and processing the acoustic features by using a fault diagnosis model to obtain hidden layer reconstruction features, and calculating a deviation error between the acoustic features and the hidden layer reconstruction features, wherein the deviation error is used for measuring the degree of the training sound data deviating from a normal running state.
4. The method of claim 1, wherein processing the acoustic data using a preset fault diagnosis model and outputting a deviation error measuring a degree of deviation of the acoustic data from a normal operation state comprises:
receiving acoustic features of the acoustic data using an input layer of a fault diagnosis model;
processing the acoustic features and generating hidden layer reconstruction features by utilizing a feature processing layer of a fault diagnosis model;
and calculating the deviation error between the acoustic characteristic and the hidden layer reconstruction characteristic by utilizing a deviation error calculation layer of the fault diagnosis model, wherein the deviation error is used for measuring the degree of the sound data deviating from the normal operation state.
5. The method of claim 1, wherein the determining the deviation error threshold comprises:
processing each piece of training sound data by using a preset fault diagnosis model to obtain a deviation error corresponding to each piece of training sound data, wherein the training sound data is the collected sound of the target equipment in a normal operation state;
arranging deviation errors of all pieces of training sound data in an ascending order or a descending order according to the sizes;
and selecting a target deviation error in the deviation error sequence as a deviation error threshold according to the recall rate or the false alarm rate set by the user.
6. The method of claim 1, wherein upon determining that the operational status of the target device is a fault, the method further comprises:
based on the sound data, a fault type of the target device is identified.
7. The method of claim 1, wherein the obtaining sound data collected for a target device comprises:
acquiring original sound data acquired by target equipment;
and filtering the original sound data according to the specific frequency band where the target equipment emits the sound to obtain the sound data of the specific frequency band.
8. The method of claim 7, further comprising:
taking the sound data of other frequency bands except the specific frequency band in the original sound data as environmental sound data;
and carrying out sound type identification on the environmental sound data to obtain the target type environmental sound contained in the environmental sound data.
9. An apparatus operation state detection device, characterized by comprising:
the voice data acquisition unit is used for acquiring voice data acquired by the target equipment;
the deviation error determining unit is used for processing the sound data by utilizing a preset fault diagnosis model and outputting a deviation error for measuring the degree of the sound data deviating from a normal running state;
and the state determining unit is used for determining the running state of the target equipment based on the deviation error and a preset deviation error threshold value.
10. An apparatus operation state detection apparatus characterized by comprising: a memory and a processor;
the memory is used for storing programs;
the processor is configured to execute the program to implement the steps of the method for detecting the operation state of the device according to any one of claims 1 to 8.
11. A storage medium having a computer program stored thereon, wherein the computer program, when executed by a processor, implements the steps of the method for detecting an operational state of an apparatus according to any one of claims 1 to 8.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202010565782.9A CN111707355A (en) | 2020-06-19 | 2020-06-19 | Equipment operation state detection method, device, equipment and storage medium |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202010565782.9A CN111707355A (en) | 2020-06-19 | 2020-06-19 | Equipment operation state detection method, device, equipment and storage medium |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN111707355A true CN111707355A (en) | 2020-09-25 |
Family
ID=72541306
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202010565782.9A Pending CN111707355A (en) | 2020-06-19 | 2020-06-19 | Equipment operation state detection method, device, equipment and storage medium |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN111707355A (en) |
Citations (11)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| JP2013200144A (en) * | 2012-03-23 | 2013-10-03 | Mitsubishi Electric Corp | Abnormal sound diagnosis device |
| CN106596123A (en) * | 2016-11-22 | 2017-04-26 | 东软集团股份有限公司 | Device fault diagnosis method, device fault diagnosis device, and device fault diagnosis system |
| CN106656637A (en) * | 2017-02-24 | 2017-05-10 | 国网河南省电力公司电力科学研究院 | Anomaly detection method and device |
| CN108052974A (en) * | 2017-12-12 | 2018-05-18 | 苏州大学 | A kind of method for diagnosing faults, system, equipment and storage medium |
| EP3367908A1 (en) * | 2015-10-30 | 2018-09-05 | The Johns Hopkins University | Programmable electronic stethoscope devices, algorithms, systems, and methods |
| CN109035749A (en) * | 2018-08-02 | 2018-12-18 | 浙江讯飞智能科技有限公司 | A kind of road thunders sound detection early warning application system extremely |
| US20190114489A1 (en) * | 2017-10-17 | 2019-04-18 | Toyota Research Institute, Inc. | Systems and methods for identification of objects using audio and sensor data |
| CN110364141A (en) * | 2019-06-04 | 2019-10-22 | 杭州电子科技大学 | Elevator typical case's abnormal sound alarm method based on depth single classifier |
| CN110443274A (en) * | 2019-06-28 | 2019-11-12 | 平安科技（深圳）有限公司 | Method for detecting abnormality, device, computer equipment and storage medium |
| CN110779745A (en) * | 2019-10-12 | 2020-02-11 | 杭州安脉盛智能技术有限公司 | Heat exchanger early fault diagnosis method based on BP neural network |
| CN110929934A (en) * | 2019-11-22 | 2020-03-27 | 深圳市通用互联科技有限责任公司 | Equipment failure prediction method and device, computer equipment and storage medium |
-
2020
- 2020-06-19 CN CN202010565782.9A patent/CN111707355A/en active Pending
Patent Citations (11)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| JP2013200144A (en) * | 2012-03-23 | 2013-10-03 | Mitsubishi Electric Corp | Abnormal sound diagnosis device |
| EP3367908A1 (en) * | 2015-10-30 | 2018-09-05 | The Johns Hopkins University | Programmable electronic stethoscope devices, algorithms, systems, and methods |
| CN106596123A (en) * | 2016-11-22 | 2017-04-26 | 东软集团股份有限公司 | Device fault diagnosis method, device fault diagnosis device, and device fault diagnosis system |
| CN106656637A (en) * | 2017-02-24 | 2017-05-10 | 国网河南省电力公司电力科学研究院 | Anomaly detection method and device |
| US20190114489A1 (en) * | 2017-10-17 | 2019-04-18 | Toyota Research Institute, Inc. | Systems and methods for identification of objects using audio and sensor data |
| CN108052974A (en) * | 2017-12-12 | 2018-05-18 | 苏州大学 | A kind of method for diagnosing faults, system, equipment and storage medium |
| CN109035749A (en) * | 2018-08-02 | 2018-12-18 | 浙江讯飞智能科技有限公司 | A kind of road thunders sound detection early warning application system extremely |
| CN110364141A (en) * | 2019-06-04 | 2019-10-22 | 杭州电子科技大学 | Elevator typical case's abnormal sound alarm method based on depth single classifier |
| CN110443274A (en) * | 2019-06-28 | 2019-11-12 | 平安科技（深圳）有限公司 | Method for detecting abnormality, device, computer equipment and storage medium |
| CN110779745A (en) * | 2019-10-12 | 2020-02-11 | 杭州安脉盛智能技术有限公司 | Heat exchanger early fault diagnosis method based on BP neural network |
| CN110929934A (en) * | 2019-11-22 | 2020-03-27 | 深圳市通用互联科技有限责任公司 | Equipment failure prediction method and device, computer equipment and storage medium |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN110839016B (en) | Abnormal flow monitoring method, device, equipment and storage medium | |
| TWI706318B (en) | Solder paste printing quality detecting method, data processing device and computer storage medium | |
| CN106951925B (en) | Data processing method, device, server and system | |
| CN109960232B (en) | Method for selecting leading auxiliary parameter and method for equipment maintenance pre-diagnosis | |
| CN111722952A (en) | Fault analysis method, system, equipment and storage medium of business system | |
| CN111811567B (en) | Equipment detection method based on curve inflection point comparison and related device | |
| CN110837852A (en) | Fault diagnosis method and device for rolling mill gearbox and terminal equipment | |
| CN109408383B (en) | Java memory leak analysis method and device | |
| CN111739557A (en) | Equipment fault positioning method, device, equipment and storage medium | |
| CN116302809A (en) | Edge end data analysis and calculation device | |
| KR101960755B1 (en) | Method and apparatus of generating unacquired power data | |
| CN114356703A (en) | Root cause analysis method and device | |
| CN116628564B (en) | Model training method and system for detecting generator state | |
| CN111707355A (en) | Equipment operation state detection method, device, equipment and storage medium | |
| CN111581425A (en) | Equipment sound classification method based on deep learning | |
| CN114564349B (en) | Monitoring method and device of server, electronic equipment and storage medium | |
| WO2022059720A1 (en) | Structure diagnosis system, structure diagnosis method, and structure diagnosis program | |
| CN111695829B (en) | Index fluctuation period calculation method and device, storage medium and electronic equipment | |
| CN111967364B (en) | Composite fault diagnosis method, device, electronic equipment and storage medium | |
| JP2001217169A (en) | Data variation monitoring method and monitoring device | |
| CN114330569A (en) | Method and equipment for detecting fan unit component fault and storage medium | |
| CN112069037A (en) | Method and device for detecting no threshold value of cloud platform | |
| CN111798237A (en) | Abnormal transaction diagnosis method and system based on application log | |
| CN114936113B (en) | Task avalanche recovery method and device, electronic equipment and storage medium | |
| CN111522678B (en) | Fault detection method and device |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| RJ01 | Rejection of invention patent application after publication | ||
| RJ01 | Rejection of invention patent application after publication |
Application publication date: 20200925 |