Degradation trend prediction method, system and equipment of compressor and medium

Abstract

The invention provides a compressor degradation trend prediction method, system and device and a medium, and the method comprises the following steps: collecting real-time data and historical data of the operation state of a compressor, carrying out the data cleaning, and taking the data as a training sample set; performing feature extraction and selection on the training sample set to obtain a target feature vector; training the classification model according to the training sample set and the target feature vector to obtain a target prediction model, and predicting the degradation trend of the compressor based on the target prediction model; according to the method, the real-time data in the operation process of the reciprocating compressor is collected, the machine learning algorithm is used for processing the data, the state trend prediction model of the reciprocating compressor is established, the future health state of the compressor can be effectively predicted through the prediction model, the state trend of the reciprocating compressor can be effectively predicted, and the reliability of the reciprocating compressor is improved. And the reliability and the stability of the reciprocating compressor are improved.

Description

Technical Field

[0001] The present invention belongs to the technical field of compressors, and in particular relates to a method, system, equipment and medium for predicting the degradation trend of a compressor. Background Art

[0002] Reciprocating compressors have the advantages of high thermal efficiency and a wide applicable flow and pressure range. They are widely used in the petrochemical industry, natural gas compression stations, etc., and are indispensable key reciprocating compressors in their process production processes. Because they have many moving parts, harsh working conditions of moving parts and many wearing parts, the probability of unexpected failures is high during long-term operation. Unplanned shutdown of reciprocating compressors due to failures often causes high economic losses and even causes serious safety accidents. Therefore, making accurate judgments and predictions on reciprocating compressor failures to avoid unplanned shutdowns of reciprocating compressors due to failures is of great significance to the development of the reciprocating compressor industry.

[0003] In recent years, with the tremendous progress of artificial intelligence technology, many scholars in the field of reciprocating compressors have also applied artificial intelligence technology to the performance analysis and fault diagnosis of reciprocating compressors. However, the current common artificial intelligence diagnostic methods have problems such as high technical cost, insufficient diagnostic accuracy and difficulty in implementation. For example, fault diagnosis based on PV diagrams requires pressure measurement through holes in the cylinder, which will destroy the integrity and airtightness of the cylinder. In addition, the exhaust pressure of large reciprocating compressors is high, and temporary drilling of holes in the cylinder body is often not allowed. Summary of the invention

[0004] In view of the problems existing in the prior art, the present invention provides a method, system, device and medium for predicting the degradation trend of a compressor. The method is simple and reliable, can ensure the normal operation of the reciprocating compressor, and predict the occurrence of failures in advance.

[0005] The present invention is achieved through the following technical solutions:

[0006] A method for predicting a deterioration trend of a compressor comprises the following steps:

[0007] Collect real-time and historical data of compressor operation status, and use them as training sample sets after data cleaning;

[0008] Extract and select features from the training sample set to obtain the target feature vector;

[0009] The classification model is trained according to the training sample set and the target feature vector to obtain a target prediction model, and the degradation trend of the compressor is predicted based on the target prediction model.

[0010] Furthermore, the real-time data and historical data collected about the operating status of the compressor include: vibration information, displacement information, pressure information, flow information and temperature information.

[0011] Furthermore, the data cleaning includes: cleaning downtime data and removing abnormal values.

[0012] Furthermore, the process of extracting and selecting features from the training sample set to obtain the target feature vector is as follows:

[0013] By extracting features from the cleaned data, the original historical data is converted into feature vectors for modeling using machine learning algorithms;

[0014] The feature selection method is used to screen the feature vector to obtain the target feature vector.

[0015] Furthermore, the feature selection method adopts autocorrelation analysis and partial autocorrelation analysis.

[0016] Furthermore, the process of training the classification model according to the training sample set and the target feature vector to obtain the target prediction model is:

[0017] Inputting the target feature vector into the differential autoregressive moving average algorithm model and the long short-term memory neural network model for training respectively to obtain a first prediction model and a second prediction model;

[0018] The accuracy of the first prediction model and the second prediction model is verified, and the prediction model with higher accuracy is used as the target prediction model.

[0019] Furthermore, the process of predicting the degradation trend of the compressor based on the target prediction model is as follows:

[0020] Acquire historical data of the operating status of the target reciprocating compressor, and input the historical data into the target prediction model;

[0021] Predict the operating status at time t+1, predict the operating status at time t+2, and predict the operating status at time t+n, and the prediction results will be used to form time series data output.

[0022] A compressor degradation trend prediction system, comprising:

[0023] The acquisition module is used to collect real-time data and historical data of the compressor operation status, and after data cleaning, use it as a training sample set;

[0024] A processing module is used to extract and select features from a training sample set to obtain a target feature vector;

[0025] The output module is used to train the classification model according to the training sample set and the target feature vector to obtain the target prediction model, and predict the degradation trend of the compressor based on the target prediction model.

[0026] A computer device comprises a memory, a processor and a computer program stored in the memory and executable on the processor. When the processor executes the computer program, the steps of a method for predicting a degradation trend of a compressor are implemented.

[0027] A computer-readable storage medium stores a computer program, which, when executed by a processor, implements the steps of a method for predicting a degradation trend of a compressor.

[0028] Compared with the prior art, the present invention has the following beneficial technical effects:

[0029] The present invention provides a method, system, device and medium for predicting the degradation trend of a compressor, comprising the following steps: collecting real-time data and historical data of the operating state of the compressor, performing data cleaning, and using the data as a training sample set; performing feature extraction and selection on the training sample set to obtain a target feature vector; training a classification model according to the training sample set and the target feature vector to obtain a target prediction model, and predicting the degradation trend of the compressor based on the target prediction model; the present application collects real-time data during the operation of a reciprocating compressor, processes the data using a machine learning algorithm, and establishes a reciprocating compressor state trend prediction model, through which the future health state of the compressor can be effectively predicted, the state trend of the reciprocating compressor can be effectively predicted, and the reliability and stability of the reciprocating compressor can be improved; the prediction model can also realize active detection of the operating state of the reciprocating compressor, and issue an early warning when it is detected that the operating state of the reciprocating compressor is a degradation trend, thereby realizing comprehensive monitoring and maintenance of the reciprocating compressor, improving the operating efficiency and reliability of the equipment, thereby reducing the cost of manual maintenance and inspection, improving the use efficiency of the reciprocating compressor, and extending the service life of the reciprocating compressor. BRIEF DESCRIPTION OF THE DRAWINGS

[0030] Figure 1 This is a flow chart of a method for predicting a deterioration trend of a compressor according to an embodiment of the present invention;

[0031] Figure 2 It is a flowchart of a method for predicting a deterioration trend of a compressor according to an embodiment of the present invention;

[0032] Figure 3 It is a schematic diagram of the box plot method in the embodiment of the invention. DETAILED DESCRIPTION

[0033] The present invention is further described in detail below in conjunction with specific embodiments, which are intended to explain the present invention rather than to limit it.

[0034] In order to enable those skilled in the art to better understand the scheme of the present invention, the technical scheme in the embodiments of the present invention will be clearly and completely described below in conjunction with the drawings in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work should fall within the scope of protection of the present invention.

[0035] It should be noted that the terms "first", "second", etc. in the specification and claims of the present invention and the above-mentioned drawings are used to distinguish similar objects, and are not necessarily used to describe a specific order or sequence. It should be understood that the data used in this way can be interchanged where appropriate, so that the embodiments of the present invention described herein can be implemented in an order other than those illustrated or described herein. In addition, the terms "including" and "having" and any variations thereof are intended to cover non-exclusive inclusions, for example, a process, method, system, product or device that includes a series of steps or units is not necessarily limited to those steps or units that are clearly listed, but may include other steps or units that are not clearly listed or inherent to these processes, methods, products or devices.

[0036] The present invention provides a method for predicting the deterioration trend of a compressor. Figure 1 As shown, the following steps are included:

[0037] Collect real-time and historical data of compressor operation status, and use them as training sample sets after data cleaning;

[0038] Extract and select features from the training sample set to obtain the target feature vector;

[0039] The classification model is trained according to the training sample set and the target feature vector to obtain a target prediction model, and the degradation trend of the compressor is predicted based on the target prediction model.

[0040] Preferably, the real-time data and historical data collected about the operating status of the compressor include: vibration information, displacement information, pressure information, flow information and temperature information.

[0041] Preferably, the data cleaning includes: cleaning downtime data and removing outliers.

[0042] Preferably, the process of extracting and selecting features from the training sample set to obtain the target feature vector is:

[0043] By extracting features from the cleaned data, the original historical data is converted into feature vectors for modeling using machine learning algorithms;

[0044] The feature selection method is used to screen the feature vector to obtain the target feature vector.

[0045] Preferably, the feature selection method uses autocorrelation analysis and partial autocorrelation analysis.

[0046] Preferably, the process of training the classification model according to the training sample set and the target feature vector to obtain the target prediction model is:

[0047] Inputting the target feature vector into the differential autoregressive moving average algorithm model and the long short-term memory neural network model for training respectively to obtain a first prediction model and a second prediction model;

[0048] The accuracy of the first prediction model and the second prediction model is verified, and the prediction model with higher accuracy is used as the target prediction model.

[0049] Preferably, the process of predicting the degradation trend of the compressor based on the target prediction model is:

[0050] Acquire historical data of the operating status of the target reciprocating compressor, and input the historical data into the target prediction model;

[0051] Predict the operating status at time t+1, predict the operating status at time t+2, and predict the operating status at time t+n, and output the prediction results as time series data;

[0052] It should be noted that those skilled in the art can plot the obtained time series data into a curve, and thus can obtain a more intuitive degradation trend curve of the compressor.

[0053] Example:

[0054] The process of this embodiment is as follows Figure 2 As shown:

[0055] First, a training sample set is obtained. The training sample set includes real-time data and historical data of the reciprocating compressor. The data includes vibration information, displacement information, pressure information, flow information, temperature information, etc. of the reciprocating compressor.

[0056] Furthermore, the training sample set is preprocessed to remove data in the training sample set whose operating status is not normal, so as to ensure the accuracy and reliability of the data. Specifically, firstly, based on the start and stop status of the reciprocating compressor or the operating frequency of the reciprocating compressor, the data when the reciprocating compressor is shut down in the training sample set is cleaned; secondly, the training sample set is cut based on the maintenance record of the reciprocating compressor to ensure that each segment of data is the natural development trend of the continuous operation of the reciprocating compressor without repair or maintenance; finally, the Laida criterion, box plot method or 4D test method can be used to remove outliers in the training sample set.

[0057] It should be noted that the Laida criterion described in this embodiment means that a group of test data is first assumed to contain only random errors, and the standard deviation is obtained by calculating and processing it, and an interval is determined according to a certain probability. It is considered that any error exceeding this interval is not a random error but a gross error, and the data containing this error should be eliminated.

[0058] It should be noted that if Figure 3 As shown, the box plot described in this embodiment is also called a box-whisker plot, a box plot or a box-whisker plot. It is a statistical chart that uses five statistical quantities in the data: minimum value, upper quartile, median, lower quartile and maximum value to describe the data; it can intuitively display the outliers of the data, the degree of dispersion of the distribution and the symmetry of the data.

[0059] The median refers to the value in the middle position after the data is arranged in ascending order. If the sequence is an even number, it is the average of the two middle numbers.

[0060] The lower quartile Q1 is the number located at the 25% position of the data sequence.

[0061] Upper quartile Q3: The number located at the 75% position of the data series.

[0062] Interquartile range IQR: that is, IQR = Q3-Q1.

[0063] Lower edge = Q1–1.5*IQR.

[0064] Upper edge = Q3 + 1.5*IQR.

[0065] Furthermore, the process of extracting and selecting features from data is as follows: by extracting features from data, the historical data in the training sample set can be converted into feature vectors with better expression capabilities, so that the machine learning algorithm can model them;

[0066] Then, the extracted features are screened using a feature selection method. In the embodiment of the present application, feature selection can use autocorrelation analysis and partial autocorrelation analysis.

[0067] The core of the autocorrelation analysis is the autocorrelation coefficient measurement, which refers to the degree of correlation between the same event in two different periods. In other words, it measures the impact of one's past behavior on one's present, which can be interpreted as a function of the similarity between two observations and the time difference between them.

[0068] The partial autocorrelation is a summary of the relationship between the observations in the time series and the observations in the previous time step with the relationship between the intervening observations removed. The autocorrelations between the observations in the previous time step and the observations include direct correlations and indirect correlations. These indirect correlations are linear functions of the correlations of the observations, including the observations at the intervening time step, and the partial autocorrelation function attempts to eliminate these indirect correlations.

[0069] Further, after feature selection, the classification model is trained according to the training sample set and the selected features to obtain a target prediction model, and the health status of the reciprocating compressor is analyzed and evaluated through the target prediction model.

[0070] In the embodiments of the present application, ARIMA (autoregressive differential moving average algorithm) and LSTM (long short-term memory neural network algorithm) are used for time series prediction.

[0071] The ARIMA is a differential autoregressive moving average algorithm model, and its basic idea is to transform a non-stationary time series into a stationary time series, and then establish a model by regressing the dependent variable only on its lagged value and the present value and lagged value of the random error term; the ARIMA model has three adjustable parameters p, d, and q, where d is the order of autoregression, which is generally first or second order, and the second order is taken by default; p and q are automatically optimized based on the Bayesian Information Criterion (BIC).

[0072] The LSTM, i.e., long short-term memory neural network, is a time recursive neural network suitable for processing and predicting important events with relatively long intervals and delays in time series. The LSTM adjustable parameters include: window_size, batch_size, num_features, num_units, optimizer, learning_rate, steps, etc.; customized debugging is required for each reciprocating compressor and cannot be extracted and given; for reciprocating compressors that cannot be customized and debugged, this embodiment does not use the algorithm.

[0073] Specifically, the process of training using the ARIMA model in this embodiment is as follows:

[0074] First, determine the ARIMA model parameters (p, d, q) and LSTM model parameters;

[0075] Secondly, the historical data set is divided into a training sample set, a test sample set and a verification sample set. The ARIMA model and the LSTM model are trained according to the training sample set and the above target features to obtain a first prediction model and a second prediction model. The first prediction model and the second prediction model are tested and verified according to the test sample set and the verification sample set to obtain a first accuracy rate of the first prediction model and a second accuracy rate of the second prediction model.

[0076] The first accuracy rate is compared with the second accuracy rate, and the prediction model with higher accuracy rate is used as the target prediction model.

[0077] Furthermore, the future operating state of the reciprocating compressor is predicted according to the target prediction model, and the specific process is: obtaining the historical data of the operating state of the reciprocating compressor at time t, inputting the historical data of the operating state at time t into the target prediction model of the reciprocating compressor, predicting the operating state at time t+1, inputting the historical data of the operating state at time t and the predicted operating state at time t+1 into the target prediction model to obtain the operating state at time t+2, and so on, inputting the predicted operating state at time t+1 to the predicted operating state at time t+n-1 into the target prediction model to obtain the operating state at time t+n, and composing the prediction results into time series data output, and the prediction results can be displayed on the screen in the form of a time series curve.

[0078] It should be noted that for reciprocating compressors that may become in a poor state within a period of time, early warning, fault time and fault type should be given in a timely manner.

[0079] In summary, by inputting information such as the current operating status and historical fault records of the reciprocating compressor, the health change trend of the reciprocating compressor in the future can be predicted, and risk assessment and preventive maintenance can be performed, which can effectively reduce manual maintenance and inspection costs, improve equipment utilization efficiency, and extend the service life of the reciprocating compressor.

[0080] The present invention provides a compressor degradation trend prediction system, comprising:

[0081] The acquisition module is used to collect real-time data and historical data of the compressor operation status, and after data cleaning, use it as a training sample set;

[0082] A processing module is used to extract and select features from a training sample set to obtain a target feature vector;

[0083] The output module is used to train the classification model according to the training sample set and the target feature vector to obtain the target prediction model, and predict the degradation trend of the compressor based on the target prediction model.

[0084] In another embodiment of the present invention, a computer device is provided, the computer device comprising a processor and a memory, the memory being used to store a computer program, the computer program comprising program instructions, and the processor being used to execute the program instructions stored in the computer storage medium. The processor may be a central processing unit (CPU), or may be other general-purpose processors, digital signal processors (DSP), application-specific integrated circuits (ASIC), field-programmable gate arrays (FPGA) or other programmable logic devices, discrete gates or transistor logic devices, discrete hardware components, etc., which are the computing core and control core of the terminal, and are suitable for implementing one or more instructions, and are specifically suitable for loading and executing one or more instructions in a computer storage medium to implement a corresponding method flow or corresponding function; the processor described in the embodiment of the present invention can be used for the operation of a method for predicting a degradation trend of a compressor.

[0085] In another embodiment of the present invention, the present invention further provides a storage medium, specifically a computer-readable storage medium (Memory), which is a memory device in a computer device for storing programs and data. It is understandable that the computer-readable storage medium here can include both built-in storage media in the computer device and, of course, extended storage media supported by the computer device. The computer-readable storage medium provides a storage space, which stores the operating system of the terminal. In addition, one or more instructions suitable for being loaded and executed by the processor are also stored in the storage space, and these instructions can be one or more computer programs (including program codes). It should be noted that the computer-readable storage medium here can be a high-speed RAM memory or a non-volatile memory, such as at least one disk memory. The processor can load and execute one or more instructions stored in the computer-readable storage medium to implement the corresponding steps of the above-mentioned embodiment of a method for predicting the degradation trend of a compressor.

[0086] Those skilled in the art will appreciate that embodiments of the present invention may be provided as methods, systems, or computer program products. Therefore, the present invention may take the form of a complete hardware embodiment, a complete software embodiment, or an embodiment combining software and hardware. Moreover, the present invention may take the form of a computer program product implemented on one or more computer-usable storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code.

[0087] The present invention is described with reference to flowcharts and / or block diagrams of methods, devices (systems), and computer program products according to embodiments of the present invention. It should be understood that each process and / or block in the flowchart and / or block diagram, as well as the combination of processes and / or blocks in the flowchart and / or block diagram, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, a special-purpose computer, an embedded processor, or other programmable data processing device to produce a machine, so that the instructions executed by the processor of the computer or other programmable data processing device generate instructions for implementing the processes in the flowchart and / or block diagram. Figure 1 A process or multiple processes and / or boxes Figure 1 A device that provides the functions specified in a block or multiple blocks.

[0088] These computer program instructions may also be stored in a computer-readable memory capable of directing a computer or other programmable data processing device to operate in a specific manner, so that the instructions stored in the computer-readable memory produce an article of manufacture comprising an instruction device, which implements the process Figure 1 A process or multiple processes and / or boxes Figure 1 A function specified in one or more boxes.

[0089] These computer program instructions can also be loaded onto a computer or other programmable data processing device so that a series of operating steps are executed on the computer or other programmable device to produce a computer-implemented process, thereby providing instructions for implementing the process. Figure 1 A process or multiple processes and / or boxes Figure 1 The steps for the functions specified in one or more boxes.

[0090] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention, rather than to limit it. Although the present invention has been described in detail with reference to the aforementioned embodiments, those skilled in the art should understand that they can still modify the technical solutions described in the aforementioned embodiments, or replace some or all of the technical features therein with equivalents. However, these modifications or replacements do not cause the essence of the corresponding technical solutions to deviate from the scope of the technical solutions of the embodiments of the present invention.

Claims

1. A method for predicting the deterioration trend of a compressor, characterized in that: The following steps are involved: Collect real-time and historical data of compressor operation status, and use them as training sample sets after data cleaning; Extract and select features from the training sample set to obtain the target feature vector; The classification model is trained according to the training sample set and the target feature vector to obtain a target prediction model, and the degradation trend of the compressor is predicted based on the target prediction model.

2. The method for predicting the deterioration trend of a compressor according to claim 1, characterized in that: The real-time data and historical data collected about the operation status of the compressor include: vibration information, displacement information, pressure information, flow information and temperature information.

3. The method for predicting the deterioration trend of a compressor according to claim 1, characterized in that: The data cleaning includes: cleaning downtime data and removing abnormal values.

4. The method for predicting the deterioration trend of a compressor according to claim 1, characterized in that: The process of extracting and selecting features from the training sample set to obtain the target feature vector is as follows: By extracting features from the cleaned data, the original historical data is converted into feature vectors for modeling using machine learning algorithms; The feature selection method is used to screen the feature vector to obtain the target feature vector.

5. A method for predicting the deterioration trend of a compressor according to claim 4, characterized in that: The feature selection method adopts autocorrelation analysis and partial autocorrelation analysis.

6. A method for predicting the deterioration trend of a compressor according to claim 1, characterized in that: The process of training the classification model according to the training sample set and the target feature vector to obtain the target prediction model is as follows: Inputting the target feature vector into the differential autoregressive moving average algorithm model and the long short-term memory neural network model for training respectively to obtain a first prediction model and a second prediction model; The accuracy of the first prediction model and the second prediction model is verified, and the prediction model with higher accuracy is used as the target prediction model.

7. A method for predicting the deterioration trend of a compressor according to claim 1, characterized in that: The process of predicting the deterioration trend of the compressor based on the target prediction model is as follows: Acquire historical data of the operating status of the target reciprocating compressor, and input the historical data into the target prediction model; Predict the operating status at time t+1, predict the operating status at time t+2, and predict the operating status at time t+n, and the prediction results will be used to form time series data output.

8. A compressor degradation trend prediction system, characterized in that: A method for predicting a deterioration trend of a compressor according to any one of claims 1 to 7, comprising: The acquisition module is used to collect real-time data and historical data of the compressor operation status, and use the data as a training sample set after data cleaning; A processing module is used to extract and select features from a training sample set to obtain a target feature vector; The output module is used to train the classification model according to the training sample set and the target feature vector to obtain the target prediction model, and predict the degradation trend of the compressor based on the target prediction model.

9. A computer device comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, characterized in that: When the processor executes the computer program, the steps of the method for predicting the degradation trend of a compressor as described in any one of claims 1 to 7 are implemented.

10. A computer-readable storage medium storing a computer program, characterized in that: When the computer program is executed by a processor, the steps of the method for predicting a degradation trend of a compressor as described in any one of claims 1 to 7 are implemented.

Images