CN116517820A - Compressor abnormality detection method, device and computer readable storage medium - Google Patents
Compressor abnormality detection method, device and computer readable storage medium Download PDFInfo
- Publication number
- CN116517820A CN116517820A CN202310420187.XA CN202310420187A CN116517820A CN 116517820 A CN116517820 A CN 116517820A CN 202310420187 A CN202310420187 A CN 202310420187A CN 116517820 A CN116517820 A CN 116517820A
- Authority
- CN
- China
- Prior art keywords
- operation data
- field operation
- working condition
- anomaly detection
- abnormality detection
- 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 abstract description 249
- 230000005856 abnormality Effects 0.000 title claims abstract description 108
- 238000012549 training Methods 0.000 claims abstract description 93
- 238000000034 method Methods 0.000 claims abstract description 46
- 230000002159 abnormal effect Effects 0.000 claims abstract description 30
- 238000003062 neural network model Methods 0.000 claims description 30
- 238000006073 displacement reaction Methods 0.000 claims description 21
- 238000000513 principal component analysis Methods 0.000 claims description 20
- 239000000725 suspension Substances 0.000 claims description 20
- 238000005339 levitation Methods 0.000 claims description 10
- 238000004590 computer program Methods 0.000 claims description 6
- 238000007405 data analysis Methods 0.000 claims description 6
- 230000001537 neural effect Effects 0.000 claims description 6
- 238000000605 extraction Methods 0.000 claims description 4
- 238000010606 normalization Methods 0.000 claims description 4
- 238000002474 experimental method Methods 0.000 abstract description 5
- 238000003745 diagnosis Methods 0.000 abstract description 3
- 230000008569 process Effects 0.000 description 14
- 239000011159 matrix material Substances 0.000 description 7
- 238000013178 mathematical model Methods 0.000 description 4
- 238000005516 engineering process Methods 0.000 description 3
- 230000002411 adverse Effects 0.000 description 2
- 238000004891 communication Methods 0.000 description 2
- 238000011161 development Methods 0.000 description 2
- 230000006870 function Effects 0.000 description 2
- 238000012545 processing Methods 0.000 description 2
- 238000012216 screening Methods 0.000 description 2
- 238000012795 verification Methods 0.000 description 2
- 238000004458 analytical method Methods 0.000 description 1
- 238000010586 diagram Methods 0.000 description 1
- 230000000694 effects Effects 0.000 description 1
- 239000000463 material Substances 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 230000003287 optical effect Effects 0.000 description 1
- 230000009467 reduction Effects 0.000 description 1
- 238000007619 statistical method Methods 0.000 description 1
- 230000009466 transformation Effects 0.000 description 1
- 230000007723 transport mechanism Effects 0.000 description 1
- XLYOFNOQVPJJNP-UHFFFAOYSA-N water Substances O XLYOFNOQVPJJNP-UHFFFAOYSA-N 0.000 description 1
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/24—Classification techniques
- G06F18/243—Classification techniques relating to the number of classes
- G06F18/2433—Single-class perspective, e.g. one-against-all classification; Novelty detection; Outlier detection
-
- F—MECHANICAL ENGINEERING; LIGHTING; HEATING; WEAPONS; BLASTING
- F04—POSITIVE - DISPLACEMENT MACHINES FOR LIQUIDS; PUMPS FOR LIQUIDS OR ELASTIC FLUIDS
- F04B—POSITIVE-DISPLACEMENT MACHINES FOR LIQUIDS; PUMPS
- F04B51/00—Testing machines, pumps, or pumping installations
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/21—Design or setup of recognition systems or techniques; Extraction of features in feature space; Blind source separation
- G06F18/217—Validation; Performance evaluation; Active pattern learning techniques
Landscapes
- Engineering & Computer Science (AREA)
- Data Mining & Analysis (AREA)
- Theoretical Computer Science (AREA)
- General Engineering & Computer Science (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Bioinformatics & Computational Biology (AREA)
- Bioinformatics & Cheminformatics (AREA)
- Evolutionary Biology (AREA)
- Evolutionary Computation (AREA)
- Physics & Mathematics (AREA)
- Artificial Intelligence (AREA)
- General Physics & Mathematics (AREA)
- Life Sciences & Earth Sciences (AREA)
- Mechanical Engineering (AREA)
- Testing And Monitoring For Control Systems (AREA)
Abstract
The embodiment of the application discloses a compressor abnormality detection method, a device and a computer readable storage medium, wherein the method comprises the following steps: acquiring first field operation data of equipment to be detected, and identifying a working condition to which the first field operation data belongs; under the condition that an abnormality detection model corresponding to the working condition to which the first field operation data belongs exists, inputting the first field operation data into the abnormality detection model corresponding to the working condition to perform abnormality detection, and outputting a detection result; under the condition that the detection result accords with the preset condition, adding the first field operation data into a training data set, and updating an abnormality detection model corresponding to the working condition to which the first field operation data belongs; the preset conditions comprise: the abnormal value x in the detection result is larger than or equal to the abnormal alarm threshold t and smaller than or equal to N times of t. According to the embodiment, the diagnosis model has higher accuracy under the full working condition without adding additional sensors or additional fault experiments, the training data quality can be improved, and the operation amount is reduced.
Description
Technical Field
The embodiment of the application relates to an electrical equipment abnormality detection technology, in particular to a compressor abnormality detection method, a device and a computer readable storage medium.
Background
The magnetic suspension compressor is core equipment of the water chilling unit, and the abnormal condition of the magnetic suspension compressor can be found in time, so that the reliability of the equipment can be effectively improved. The current common anomaly detection methods can be divided into data sources: abnormality detection using a vibration signal, abnormality detection using temperature, pressure, etc.; the detection method can be divided into: methods based on signal processing, based on physical model, or based on data driving, etc. The use of vibration sensors undoubtedly increases the cost of fault early warning, and development and application of sufficiently accurate physical models are also difficult.
Disclosure of Invention
The embodiment of the application provides a compressor anomaly detection method, a compressor anomaly detection device and a computer-readable storage medium, which can avoid adding additional sensors and additional fault experiments, so that a diagnosis model has higher accuracy under all working conditions, the quality of training data can be improved, and the operation amount is reduced.
The embodiment of the application provides a compressor abnormality detection method, which can include:
acquiring first field operation data in the operation process of equipment to be detected, and identifying the working condition of the first field operation data;
under the condition that an abnormality detection model corresponding to the working condition of the first field operation data exists, inputting the first field operation data into the abnormality detection model corresponding to the working condition to perform abnormality detection, and outputting a detection result;
when the detection result accords with the preset condition, the first field operation data is added into a training data set, and an abnormality detection model corresponding to the working condition of the input first field operation data can be updated, so that the updated abnormality detection model can be used for abnormality detection in the subsequent abnormality detection; the preset conditions include: the abnormal value x contained in the detection result is larger than or equal to an abnormal alarm threshold t and smaller than or equal to N times of t; n is a positive integer greater than 1; wherein N is a positive integer greater than 1.
In an exemplary embodiment of the present application, the automatically updating the obtained anomaly detection model corresponding to the corresponding working condition of the input first field operation data may include:
training data added with first field operation data is adopted to retrain the abnormality detection model corresponding to the corresponding working condition, so that the abnormality detection model is automatically updated, and the retrained abnormality detection model is used as the updated abnormality detection model corresponding to the working condition.
In an exemplary embodiment of the present application, the method may further include: after each abnormality detection of the abnormality detection model, outputting the abnormality alarm threshold t and the multiple N; the detection result comprises the abnormal alarm threshold t and the multiple N.
In an exemplary embodiment of the present application, the inputting the first field operation data into an anomaly detection model corresponding to the operating condition to perform anomaly detection may include:
and after feature extraction and normalization are carried out on the first field operation data in the abnormality detection model, a preset principal component analysis algorithm is adopted to carry out data analysis on the first field operation data, so that abnormality detection is realized.
In an exemplary embodiment of the present application, the principal component analysis algorithm may be configured to detect whether one or more of the first field operation data input to the anomaly detection model is anomaly data;
in an exemplary embodiment of the present application, the method may further include obtaining an anomaly detection model corresponding to a working condition to which the first field operation data belongs according to the following manner:
invoking an anomaly detection model corresponding to the working condition of the first field operation data which is created in advance and trained; or alternatively, the process may be performed,
and directly creating and training an abnormality detection model corresponding to the working condition to which the first field operation data belong.
In an exemplary embodiment of the present application, creating and training an anomaly detection model under different conditions may include:
establishing anomaly detection neural network models corresponding to different working conditions;
acquiring second field operation data of equipment to be detected in the operation process, identifying the working condition of the second field operation data, and taking the acquired second field operation data as training data corresponding to an anomaly detection neural network model corresponding to the identified working condition;
for the anomaly detection neural network model under each working condition, the following training steps are respectively carried out: and inputting training data corresponding to the working condition into an anomaly detection neural network model corresponding to the working condition, training the anomaly detection neural model, and taking the trained anomaly detection neural model as the anomaly detection model corresponding to the working condition.
In an exemplary embodiment of the present application, before the training data corresponding to the working condition is input into the anomaly detection neural network model under the corresponding working condition for training, the method may further include:
after adding new training data each time, detecting whether the quantity of the training data under the corresponding working condition reaches a preset quantity threshold value or not;
when the number of the training data under the corresponding working conditions reaches the number threshold, the training data under the corresponding working conditions can be input into the anomaly detection neural network model under the corresponding working conditions for training;
and when the quantity of the training data under the corresponding working condition does not reach the quantity threshold, continuing to wait for the addition of new training data under the working condition.
In an exemplary embodiment of the present application, the apparatus to be detected may include a magnetic levitation compressor; the first field operation data and the second field operation data are operation data obtained from sensors when the magnetic suspension compressor normally operates; the first field operation data and the second field operation data may each include, but are not limited to, any one or more of the following:
the method comprises the following steps of feeding back the opening degree of a guide vane, feeding back the pressure of secondary exhaust, feeding back the frequency of a frequency converter, sucking pressure, AZ current, FY current, FX current, RY current, RX current, FY displacement, FX displacement, RY displacement, RX displacement and AZ displacement.
In an exemplary embodiment of the present application, the method may further include:
when the abnormality detection model corresponding to the working condition to which the first field operation data belongs does not exist in all the abnormality detection models, the abnormality detection model corresponding to the working condition to which the first field operation data belongs can be directly created and trained.
The embodiment of the application also provides a compressor abnormality detection device, which can include, but is not limited to, a processor and a computer-readable storage medium, wherein the computer-readable storage medium stores instructions, and when the instructions are executed by the processor, the compressor abnormality detection method can be realized.
The embodiment of the application also provides a computer readable storage medium, on which a computer program is stored, and the computer program realizes the compressor abnormality detection method when being executed by a processor.
Compared with the related art, the embodiment of the application can comprise the following steps: acquiring first field operation data in the operation process of equipment to be detected, and identifying the working condition of the first field operation data; when an abnormality detection model corresponding to the working condition of the first field operation data exists, the first field operation data can be input into the abnormality detection model corresponding to the working condition to perform abnormality detection, and a detection result is output; when the detection result accords with the preset condition, the first field operation data is added into a training data set, and the abnormality detection model under the working condition of the input first field operation data can be automatically updated, so that the updated abnormality detection model can be used for abnormality detection in the subsequent abnormality detection; the preset conditions include: the abnormal value x contained in the detection result is larger than or equal to an abnormal alarm threshold t and smaller than or equal to N times of t; wherein N is a positive integer greater than 1. According to the embodiment, the diagnosis model has higher accuracy under the full working condition without adding additional sensors or additional fault experiments, the training data quality can be improved, and the operation amount is reduced.
Additional features and advantages of the application will be set forth in the description which follows, and in part will be obvious from the description, or may be learned by practice of the application. Other advantages of the present application may be realized and attained by the structure particularly pointed out in the written description and drawings.
Drawings
The accompanying drawings are included to provide an understanding of the technical aspects of the present application, and are incorporated in and constitute a part of this specification, illustrate the technical aspects of the present application and together with the examples of the present application, and not constitute a limitation of the technical aspects of the present application.
FIG. 1 is a flow chart of a method for detecting compressor anomalies according to an embodiment of the present application;
fig. 2 is a block diagram showing the constitution of a compressor abnormality detection apparatus according to an embodiment of the present application.
Detailed Description
The present application describes a number of embodiments, but the description is illustrative and not limiting and it will be apparent to those of ordinary skill in the art that many more embodiments and implementations are possible within the scope of the embodiments described herein. Although many possible combinations of features are shown in the drawings and discussed in the detailed description, many other combinations of the disclosed features are possible. Any feature or element of any embodiment may be used in combination with or in place of any other feature or element of any other embodiment unless specifically limited.
The present application includes and contemplates combinations of features and elements known to those of ordinary skill in the art. The embodiments, features and elements of the present disclosure may also be combined with any conventional features or elements to form a unique inventive arrangement as defined in the claims. Any feature or element of any embodiment may also be combined with features or elements from other inventive arrangements to form another unique inventive arrangement as defined in the claims. Thus, it should be understood that any of the features shown and/or discussed in this application may be implemented alone or in any suitable combination. Accordingly, the embodiments are not to be restricted except in light of the attached claims and their equivalents. Further, various modifications and changes may be made within the scope of the appended claims.
Furthermore, in describing representative embodiments, the specification may have presented the method and/or process as a particular sequence of steps. However, to the extent that the method or process does not rely on the particular order of steps set forth herein, the method or process should not be limited to the particular sequence of steps described. Other sequences of steps are possible as will be appreciated by those of ordinary skill in the art. Accordingly, the particular order of the steps set forth in the specification should not be construed as limitations on the claims. Furthermore, the claims directed to the method and/or process should not be limited to the performance of their steps in the order written, and one skilled in the art can readily appreciate that the sequences may be varied and still remain within the spirit and scope of the embodiments of the present application.
The embodiment of the application provides a method for detecting compressor abnormality, as shown in fig. 1, the method may include steps S101-S103:
s101, acquiring first field operation data in the operation process of equipment to be detected, and identifying the working condition of the first field operation data;
s102, under the condition that an abnormality detection model corresponding to the working condition of the first field operation data exists, inputting the first field operation data into the abnormality detection model corresponding to the working condition to detect abnormality, and outputting a detection result;
s103, when the detection result accords with a preset condition, adding the first field operation data into a training data set, and automatically updating an abnormality detection model corresponding to the working condition to which the input first field operation data belongs, so that the updated abnormality detection model can be used for abnormality detection in the subsequent abnormality detection; the preset conditions may include: the abnormal value x contained in the detection result is larger than or equal to an abnormal alarm threshold t and smaller than or equal to N times of t; n is a positive integer greater than 1; wherein N is a positive integer greater than 1.
In exemplary embodiments of the present application, the device to be inspected may include, but is not limited to, a magnetic levitation compressor. The following describes embodiments of the present application using a magnetic levitation compressor as an example.
In the exemplary embodiment of the present application, it is known through analysis of the current abnormality detection method of the magnetic suspension compressor that, in the current common abnormality detection method, the fault detection using vibration signals and the like clearly requires additional devices such as vibration sensors, which necessarily increases the cost of fault early warning of the magnetic suspension compressor, if no additional devices such as sensors are used, a detection model can be used, however, a development period of a sufficiently accurate physical model is long, and a large amount of training data is required, and the abnormality detection scheme also has great difficulty.
In order to solve the above problems, in an exemplary embodiment of the present application, the present application embodiment performs training of an anomaly detection model based on various sensors carried by a current magnetic levitation compressor, and implements automatic update of the anomaly detection model by using field operation data of the magnetic levitation compressor, thereby obtaining an anomaly detection model with high accuracy.
In an exemplary embodiment of the present application, for example, in order not to add additional sensors, abnormal state identification of the magnetic levitation compressor may be implemented using collected data of various sensors included in the magnetic levitation compressor set; in order to avoid the need of additional fault experiments, the training of the abnormal detection model can be realized only by the normal operation data of the magnetic suspension compressor; and through setting up the anomaly detection model under different operating modes to and make the anomaly detection model automatic update under different operating modes, can all have higher detection accuracy under different operating modes, and the automatic update of anomaly detection model is realized based on real-time operation data in the anomaly detection model use, thereby reduced the use of experimental training data.
In an exemplary embodiment of the present application, in order to implement the embodiment of the present application, an anomaly detection model under different working conditions may be first obtained.
In an exemplary embodiment of the present application, the method may further include obtaining an anomaly detection model corresponding to a working condition to which the first field operation data belongs according to the following manner:
calling an anomaly detection model corresponding to the working condition of the first field operation data which is created in advance and trained; or alternatively, the process may be performed,
and directly creating and training an abnormality detection model corresponding to the working condition to which the first field operation data belong.
In an exemplary embodiment of the present application, the anomaly detection model may be created and initially trained in advance, and the anomaly detection model obtained after the initial training is stored, so that the stored anomaly detection model is directly called when the embodiment of the present application is executed.
In an exemplary embodiment of the present application, in order to implement comprehensive fault detection for the magnetic levitation compressor, an anomaly detection model under different working conditions may be pre-established.
In the exemplary embodiment of the present application, in other embodiments, when the abnormality detection model under a certain working condition is found to be not stored in advance when the embodiment of the present application is executed, the abnormality detection model under the working condition may also be directly created and trained.
In an exemplary embodiment of the present application, creating and training anomaly detection models corresponding to different conditions may include:
establishing anomaly detection neural network models corresponding to different working conditions;
acquiring second field operation data of equipment to be detected in the operation process, identifying the working condition of the second field operation data, and taking the acquired second field operation data as training data corresponding to an abnormality detection neural network model under the identified working condition;
for each anomaly detection neural network model under each condition, the following training steps may be performed: the training data corresponding to the working condition is input into the anomaly detection neural network model under the corresponding working condition and the anomaly detection neural model is trained, so that the trained anomaly detection neural model can be used as the anomaly detection model corresponding to the working condition.
In an exemplary embodiment of the present application, when creating the anomaly detection model, an anomaly detection neural network model capable of implementing a principal component analysis algorithm may be first established based on a preset principal component analysis algorithm (i.e., PCA algorithm), i.e., the anomaly detection neural network model may include a mathematical model of the principal component analysis algorithm.
The PCA (Principal Component Analysis ) algorithm is a commonly used method of data analysis. PCA transforms raw data into a set of linearly independent representations of each dimension through linear transformation, and can be used for extracting main characteristic components of data and is commonly used for dimension reduction of high-dimension data. The abnormal detection of PCA applied to the magnetic suspension compressor can decompose the data space formed by the field operation data of the magnetic suspension compressor into two subspaces: a main component subspace and a residual subspace; the mathematical model of the PCA algorithm can be expressed as:
A＝A1+A2；
A1＝PP T A；
A2＝(I-PP T )A；
wherein, A is the field operation data (for example, first field operation data and second field operation data) of the magnetic suspension compressor, A1 is the projection of X in the main component space, and A2 is the projection of A in the residual subspace; wherein, PP T And I-PP T Respectively a projection matrix, wherein I is a unit matrix, P is a load matrix, and the load matrix is obtained by a series of field operation data of the magnetic suspension compressor under the normal working condition; the main component space is mainly the normal value of the measured data, and the residual subspace is fault and measured noise.
In an exemplary embodiment of the present application, the load matrix P in the anomaly detection neural network model (e.g., the mathematical model of the PCA algorithm) corresponding to different operating conditions is different.
In an exemplary embodiment of the present application, the principal component analysis algorithm may be used to detect whether one or more field operational data input into the anomaly detection model is anomaly data.
In an exemplary embodiment of the present application, prior to performing the principal component analysis algorithm described, may include, but is not limited to: feature screening (i.e., feature extraction), normalization, and the like.
In the exemplary embodiment of the present application, since the anomaly detection model to be created is an anomaly detection model under different working conditions, the anomaly detection neural network model may be created as a different anomaly detection neural network model according to different working conditions, or an average load matrix P1 may be calculated according to an average value of a series of field operation data of the magnetic suspension compressor under a normal working condition, so that a unified anomaly detection neural network model is created according to the mathematical model of the PCA algorithm and the average load matrix P1, and the anomaly detection model under different working conditions is obtained by training the unified anomaly detection neural network model through training data under different working conditions.
In an exemplary embodiment of the present application, after the anomaly detection neural network model is created, training data may be acquired to train the anomaly detection neural network model to obtain an anomaly detection model for the magnetic levitation compressor.
In an exemplary embodiment of the present application, second field operation data of the magnetic suspension compressor in the operation process may be obtained, and the second field operation data is used as training data. In order to obtain the anomaly detection model under different working conditions, the obtained second field operation data can be identified through a preset identification algorithm, so that the working condition corresponding to the second field operation data obtained at present is identified, and the created anomaly detection neural network model is trained as training data under the working condition.
In an exemplary embodiment of the present application, when the device to be detected is a magnetic suspension compressor, the collected second field operation data is operation data obtained from a sensor when the magnetic suspension compressor operates normally; the second field operational data may include, but is not limited to, any one or more of the following:
the method comprises the following steps of feeding back the opening degree of a guide vane, feeding back the pressure of secondary exhaust, feeding back the frequency of a frequency converter, sucking pressure, AZ current, FY current, FX current, RY current, RX current, FY displacement, FX displacement, RY displacement, RX displacement and AZ displacement.
In an exemplary embodiment of the present application, before the training data is input into the anomaly detection neural network model under the corresponding working condition for training, the method may further include:
detecting whether the quantity of training data under corresponding working conditions reaches a preset quantity threshold value or not after new training data are added;
when the number of the training data under the corresponding working conditions reaches the number threshold, the training data under the corresponding working conditions can be input into the anomaly detection neural network model under the corresponding working conditions for training;
and when the quantity of the training data under the corresponding working condition does not reach the quantity threshold, continuing to wait for the addition of new training data under the working condition.
In the exemplary embodiment of the present application, in order to ensure the training effect, the training data is required to reach a certain amount, in addition, in order to save data, shorten the collection time of the training data, improve the training efficiency, and the amount of the training data cannot be too large, so that the amount threshold value can be preset, and when the second field operation data under a certain collected working condition reaches the amount threshold value, the model training can be started to be performed.
In an exemplary embodiment of the present application, the data threshold may be defined by itself according to different working conditions and different application scenarios, and the detailed numerical value of the data threshold is not limited herein.
In an exemplary embodiment of the present application, when it is detected that the training data in the training data set under a certain working condition reaches a preset number threshold, data processing may be performed on the training data, for example, including but not limited to: marking data, namely marking abnormal data in training data; and performing model training by using the processed training data.
In an exemplary embodiment of the present application, if it is detected that the number of training data in the training data set under a certain working condition does not reach the preset number threshold, the method is in a waiting state, and new second field operation data under the working condition is waited for to be added into the training data set until the number of training data reaches the data threshold.
In an exemplary embodiment of the present application, after the training data is input into the anomaly detection neural network model, a preset principal component analysis algorithm may be used in the anomaly detection neural network model to perform data analysis on the training data, thereby implementing detection of the anomaly data and implementing training of the anomaly detection neural network model to obtain the anomaly detection model.
In the exemplary embodiment of the present application, the trained anomaly detection model may also be checked by using pre-prepared check data, so as to verify the detection performance of the anomaly detection model, and when the verification passes, the anomaly detection model may be used as a qualified anomaly detection model and may be put into daily anomaly detection application. When the verification fails, training the abnormality detection model can be continued until the abnormality detection model becomes a qualified abnormality detection model, and the abnormality detection model is put into daily abnormality detection application.
In the exemplary embodiment of the present application, an initial abnormality detection model is obtained through the above scheme, and based on the initial abnormality detection model, the magnetic suspension compressor may be subjected to real-time abnormality detection, and the initial abnormality detection model may be updated in real time according to the detection result.
In an exemplary embodiment of the present application, the initial anomaly detection model is an anomaly detection model for a plurality of different conditions.
In an exemplary embodiment of the present application, the method may further include:
when the abnormality detection model corresponding to the working condition to which the first field operation data belongs does not exist in all the abnormality detection models, the abnormality detection model corresponding to the working condition to which the first field operation data belongs can be directly created and trained.
In the exemplary embodiment of the present application, since the working conditions that may be considered when the initial anomaly detection model is established are not comprehensive enough, in the actual detection process, if the anomaly detection model under a certain working condition is not found after the first field operation data under the working condition is obtained, the anomaly detection model under the working condition can be directly created, so that the working condition type of the anomaly detection model is continuously improved in practice, and the anomaly detection model library is enriched.
In the exemplary embodiment of the present application, in a daily anomaly detection process, real-time operation data (i.e., the first field operation data) of the magnetic suspension compressor may be obtained, a working condition corresponding to the first field operation data is identified, after the working condition is determined, the first field operation data is input to an anomaly detection model corresponding to the working condition, and data analysis is performed through the anomaly detection model corresponding to the working condition to implement anomaly detection, and a detection result is output.
In an exemplary embodiment of the present application, when the device to be detected is a magnetic suspension compressor, the first field operation data collected in real time is operation data obtained from a sensor when the magnetic suspension compressor is operating normally; the first field operational data may include, but is not limited to, any one or more of the following:
the method comprises the following steps of feeding back the opening degree of a guide vane, feeding back the pressure of secondary exhaust, feeding back the frequency of a frequency converter, sucking pressure, AZ current, FY current, FX current, RY current, RX current, FY displacement, FX displacement, RY displacement, RX displacement and AZ displacement.
In an exemplary embodiment of the present application, the inputting the first field operation data into an anomaly detection model under the working condition to perform anomaly detection may include:
after feature extraction and normalization are performed on the first field operation data in the abnormality detection model, a preset principal component analysis algorithm is adopted to perform data analysis on the first field operation data, so that abnormality detection is realized.
In an exemplary embodiment of the present application, whether the first field operation data input to the anomaly detection model has the anomaly data may be detected by the principal component analysis algorithm.
In an exemplary embodiment of the present application, a preset operation may be performed on all the input first field operation data in the principal component analysis algorithm, and an operation value may be obtained, where the operation value may be used as an outlier (may be referred to as an SPE value) x to represent whether or not there is any outlier in all the first field operation data, and the outlier x may be included in the detection result and output, so as to determine whether or not there is any outlier currently according to the output outlier.
In an exemplary embodiment of the present application, x= iia2 ii2=a2t (I-PPT) a2+.ltoreq.δ, δ being the confidence limit of x, can be obtained by statistical analysis.
In the exemplary embodiment of the application, in order to solve the problem that the traditional data driving method seriously depends on training data, so that the abnormality detection model can be automatically updated by utilizing daily operation data, the abnormality detection model has higher accuracy under all working conditions, and the embodiment of the application provides an automatic updating strategy of the abnormality detection model.
In an exemplary embodiment of the present application, the automatic update timing of the abnormality detection model may be determined according to the detection result. For example, the determination may be made based on the magnitude of the abnormal value x in the abnormality detection result.
In the exemplary embodiment of the present application, since the abnormality detection model needs to be continuously updated because the abnormality detection model is generally unstable in the early stage and is susceptible to the influence of factors such as temperature variation, external disturbance, etc., performance fluctuation affects the detection result, when determining the timing of updating the abnormality detection model, if it is determined that the timing of updating has arrived when detecting that the abnormality detection model has caused the detection performance fluctuation due to the disturbance of the adverse factors, updating of the abnormality detection model may be performed.
In the exemplary embodiment of the present application, when the abnormal value is sufficiently small, it may be indicated that the input data of the abnormality detection model is not abnormal, when the abnormal value is sufficiently large, it may be indicated that the input data of the abnormality detection model is necessarily abnormal, and when the abnormal value is between certain values, it is indicated that the input data is checked to be abnormal, but the detection result may be caused by interference of adverse factors, and at this time, the abnormality detection model may be updated.
In an exemplary embodiment of the present application, the magnitude of the outlier x may be determined by an outlier alarm threshold t, e.g., when x < t, it may be determined that the outlier x is small enough, where the input data of the anomaly detection model is not outlier (i.e., the existing operational data is substantially similar to the data in the training data set, with little difference), when t+.x+.nt, it may be determined that the anomaly detection model needs to be updated, and when x > Nt, it may be determined that the outlier x is large enough (i.e., the existing operational data is very different from the data in the training data set), where the input data of the anomaly detection model may be determined to be outlier.
In the exemplary embodiment of the application, whether the first operation data is added into the training data set is determined by judging the magnitude of the abnormal value x, so that the purpose of accurately screening the training data set is achieved, the quality of the training data can be improved, and compared with the case that any operation data is added into the training data set, a large amount of invalid operation data is adopted for training, and the operation amount and the workload are reduced.
In an exemplary embodiment of the present application, the method may further include: after each abnormality detection of the abnormality detection model, outputting the abnormality alarm threshold t and the multiple N; the detection result comprises the abnormal alarm threshold t and the multiple N.
In the exemplary embodiment of the present application, the anomaly alarm threshold t and the multiple N are not constant values and may be continuously optimized as the anomaly detection model is updated.
In an exemplary embodiment of the present application, the automatically updating the anomaly detection model corresponding to the working condition of the input first field operation data may include:
training the abnormality detection model again by adopting training data added with first field operation data, so as to automatically update the abnormality detection model, and taking the retrained abnormality detection model as an updated abnormality detection model corresponding to the working condition.
In an exemplary embodiment of the present application, embodiments of the present application include at least the following advantages:
1. based on mathematical statistics, the data acquired by the sensors such as temperature, pressure, current, displacement and the like of the unit are utilized, and no additional sensor is needed, so that the anomaly detection cost is reduced;
2. only the normal operation data of the magnetic suspension compressor unit is required to be input, and no additional fault experiment is required; saving manpower and material resources.
3. The model self-updating strategy is provided, the problem that the traditional data driving method seriously depends on training data is solved, the abnormal detection model can be automatically updated by utilizing daily operation data, and the abnormal detection model has higher accuracy under all working conditions.
The embodiment of the present application further provides a compressor abnormality detection device 1, as shown in fig. 2, which may include, but is not limited to, a processor 11 and a computer readable storage medium 12, where instructions are stored in the computer readable storage medium 12, and when the instructions are executed by the processor 11, the compressor abnormality detection method may be implemented.
In the exemplary embodiments of the present application, any embodiment of the foregoing method for detecting abnormal compressor is applicable to the embodiment of the apparatus, and will not be described herein in detail.
The embodiment of the application also provides a computer readable storage medium, on which a computer program is stored, and the computer program realizes the compressor abnormality detection method when being executed by a processor.
In the exemplary embodiments of the present application, any embodiment of the foregoing method for detecting a compressor abnormality is applicable to an embodiment of the computer-readable storage medium, and will not be described herein in detail.
Those of ordinary skill in the art will appreciate that all or some of the steps, systems, functional modules/units in the apparatus, and methods disclosed above may be implemented as software, firmware, hardware, and suitable combinations thereof. In a hardware implementation, the division between the functional modules/units mentioned in the above description does not necessarily correspond to the division of physical components; for example, one physical component may have multiple functions, or one function or step may be performed cooperatively by several physical components. Some or all of the components may be implemented as software executed by a processor, such as a digital signal processor or microprocessor, or as hardware, or as an integrated circuit, such as an application specific integrated circuit. Such software may be distributed on computer readable media, which may include computer storage media (or non-transitory media) and communication media (or transitory media). The term computer storage media includes both volatile and nonvolatile, removable and non-removable media implemented in any method or technology for storage of information such as computer readable instructions, data structures, program modules or other data, as known to those skilled in the art. Computer storage media includes, but is not limited to, RAM, ROM, EEPROM, flash memory or other memory technology, CD-ROM, digital Versatile Disks (DVD) or other optical disk storage, magnetic cassettes, magnetic tape, magnetic disk storage or other magnetic storage devices, or any other medium which can be used to store the desired information and which can be accessed by a computer. Furthermore, as is well known to those of ordinary skill in the art, communication media typically embodies computer readable instructions, data structures, program modules or other data in a modulated data signal such as a carrier wave or other transport mechanism and includes any information delivery media.
Claims (10)
1. A compressor anomaly detection method, the method comprising:
acquiring first field operation data of equipment to be detected, and identifying a working condition to which the first field operation data belongs;
under the condition that an abnormality detection model corresponding to the working condition of the first field operation data exists, the first field operation data is input into the abnormality detection model corresponding to the working condition of the first field operation data to carry out abnormality detection, and a detection result is output;
under the condition that the detection result meets the preset condition, adding the first field operation data into a training data set, and updating an abnormal detection model corresponding to the working condition of the first field operation data; the preset conditions include: the abnormal value x contained in the detection result is larger than or equal to an abnormal alarm threshold t and smaller than or equal to N times of t; n is a positive integer greater than 1.
2. The compressor anomaly detection method of claim 1, wherein updating the anomaly detection model corresponding to the condition to which the first field operation data pertains comprises:
training data added with the first field operation data is adopted to retrain the abnormality detection model corresponding to the working condition, and the retrained abnormality detection model is used as an updated abnormality detection model corresponding to the working condition.
3. The compressor anomaly detection method according to claim 2, wherein the detection result includes the anomaly alarm threshold t and a multiple N.
4. The method for detecting an abnormality of a compressor according to claim 1, wherein said inputting the first field operation data into an abnormality detection model corresponding to a condition to which the first field operation data belongs to performs abnormality detection, includes:
and after feature extraction and normalization are carried out on the first field operation data in the abnormality detection model, carrying out data analysis on the first field operation data by adopting a preset principal component analysis algorithm so as to realize abnormality detection.
5. The method for detecting abnormal conditions of a compressor according to claim 1, further comprising obtaining an abnormality detection model corresponding to a working condition to which the first field operation data belongs according to the following manner:
and invoking an anomaly detection model corresponding to the working condition of the first field operation data which is created and trained in advance, or creating and training the anomaly detection model corresponding to the working condition of the first field operation data.
6. The compressor anomaly detection method of claim 5, wherein creating and training anomaly detection models corresponding to different conditions comprises:
establishing an anomaly detection neural network model under different working conditions;
acquiring second field operation data of equipment to be detected, identifying working conditions to which the second field operation data belong, and taking the second field operation data as training data corresponding to the identified working conditions;
for the anomaly detection neural network model under each working condition, the following training steps are respectively carried out: and inputting training data corresponding to the working condition into an anomaly detection neural network model corresponding to the working condition, training the anomaly detection neural model, and taking the trained anomaly detection neural model as the anomaly detection model corresponding to the working condition.
7. The compressor anomaly detection method of claim 6, wherein before inputting training data corresponding to the operating condition into the anomaly detection neural network model corresponding to the operating condition, the method further comprises:
detecting whether the number of training data under corresponding working conditions reaches a preset number threshold value or not after new training data are added each time;
when the quantity of the training data under the corresponding working conditions reaches the quantity threshold, inputting the training data under the corresponding working conditions into an anomaly detection neural network model corresponding to the corresponding working conditions;
and waiting for new training data to be added under the working condition when the quantity of the training data under the corresponding working condition does not reach the quantity threshold value.
8. The compressor anomaly detection method of claim 6, wherein the device to be detected comprises a magnetic levitation compressor; the first field operation data and the second field operation data are operation data obtained by a sensor when the magnetic suspension compressor normally operates;
the first field operation data and the second field operation data each include any one or more of:
secondary exhaust pressure, suction pressure, vane opening feedback, frequency converter feedback frequency, AZ current, FX current, FY current, RX current, RY current, FX displacement, FY displacement, RX displacement, RY displacement, and AZ displacement.
9. A compressor anomaly detection device comprising a processor and a computer readable storage medium having instructions stored therein that when executed by the processor implement the compressor anomaly detection method of any one of claims 1-8.
10. A computer-readable storage medium, on which a computer program is stored, characterized in that the computer program, when executed by a processor, implements the compressor anomaly detection method according to any one of claims 1 to 8.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202310420187.XA CN116517820A (en) | 2023-04-18 | 2023-04-18 | Compressor abnormality detection method, device and computer readable storage medium |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202310420187.XA CN116517820A (en) | 2023-04-18 | 2023-04-18 | Compressor abnormality detection method, device and computer readable storage medium |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN116517820A true CN116517820A (en) | 2023-08-01 |
Family
ID=87395230
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202310420187.XA Pending CN116517820A (en) | 2023-04-18 | 2023-04-18 | Compressor abnormality detection method, device and computer readable storage medium |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN116517820A (en) |
-
2023
- 2023-04-18 CN CN202310420187.XA patent/CN116517820A/en active Pending
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN110659173B (en) | Operation and maintenance system and method | |
| CN110321371B (en) | Log data anomaly detection method, device, terminal and medium | |
| US20200258057A1 (en) | Repair management and execution | |
| CN116046187B (en) | A unusual remote monitoring system of temperature for communication cabinet | |
| US7930122B2 (en) | Evaluating anomaly for one-class classifiers in machine condition monitoring | |
| CN111435366A (en) | Equipment fault diagnosis method and device and electronic equipment | |
| CN111695620B (en) | Method and system for detecting and correcting time sequence abnormal data of power system | |
| CN111459692B (en) | Method, apparatus and computer program product for predicting drive failure | |
| US20220004163A1 (en) | Apparatus for predicting equipment damage | |
| CN116304909A (en) | Abnormality detection model training method, fault scene positioning method and device | |
| WO2023044770A1 (en) | Dry pump downtime early warning method and apparatus, electronic device, storage medium, and program | |
| CN111310778A (en) | Detection device, detection method, and recording medium on which detection program is recorded | |
| CN111030299A (en) | Side channel-based power grid embedded terminal safety monitoring method and system | |
| CN114861522A (en) | Precision manufacturing quality monitoring method and device based on artificial intelligence meta-learning technology | |
| KR20210011822A (en) | Method of detecting abnormal log based on artificial intelligence and system implementing thereof | |
| CN113111585A (en) | Intelligent cabinet fault prediction method and system and intelligent cabinet | |
| CN116517820A (en) | Compressor abnormality detection method, device and computer readable storage medium | |
| CN113869373A (en) | Equipment abnormality detection method and device, computer equipment and storage medium | |
| CN115837862A (en) | Abnormal data detection method and device, electronic equipment and storage medium | |
| CN113535458A (en) | Abnormal false alarm processing method and device, storage medium and terminal | |
| CN114626400A (en) | Equipment fault diagnosis device, equipment fault diagnosis method, intelligent factory system and application agent | |
| KR102497586B1 (en) | Method and system for diagnosing abnormality of a plurality of apparatus based on sequential learning | |
| CN117466153B (en) | Fault detection method, device, computer equipment and readable storage medium | |
| CN117520040B (en) | Micro-service fault root cause determining method, electronic equipment and storage medium | |
| CN117478017B (en) | Control method of servo motor |
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 |