Instrument fault detection method, device and electronic equipment

Abstract

This disclosure relates to a method, apparatus, and electronic device for instrument fault detection, belonging to the field of instrument detection technology. It enables remote fault detection of different types of instruments through different fault detection modes. The instrument fault detection method includes: acquiring time-series data of each instrument in a natural gas purification plant within a preset time period; for any two instruments, determining the correlation coefficient between the two instruments based on their corresponding time-series data, and performing fault detection on the two instruments based on the relationship between the correlation coefficient and a preset interval, until fault detection of all instruments in the natural gas purification plant is completed, thereby obtaining the fault detection results of the instruments in the natural gas purification plant.

Description

Technical Field

[0001] This disclosure relates to the field of instrument testing technology, specifically to an instrument fault detection method, apparatus, and electronic equipment. Background Technology

[0002] In recent years, with the increasing demand for natural gas and the rise in natural gas purification plants, ensuring the safe and stable operation of instruments is a key aspect of the stable operation of purification units. However, traditional instrument maintenance methods suffer from problems such as low inspection efficiency, low fault analysis efficiency, and the inability of process operators to monitor instrument status online. Summary of the Invention

[0003] To address the shortcomings of existing technologies, this disclosure provides a method, apparatus, and electronic device for instrument fault detection.

[0004] To achieve the above objectives, in a first aspect, this disclosure provides an instrument fault detection method, the method comprising:

[0005] Acquire time-series data of each instrument in the natural gas purification plant within a preset time period;

[0006] For any two instruments, the correlation coefficient between the two instruments is determined based on the time series data corresponding to the two instruments. Then, based on the relationship between the correlation coefficient and the preset interval, fault detection is performed on the two instruments until the fault detection of all instruments in the natural gas purification plant is completed, and the fault detection results of the instruments in the natural gas purification plant are obtained.

[0007] Secondly, this disclosure provides an instrument fault detection device, the device comprising:

[0008] The acquisition module is configured to acquire time-series data of each instrument in the natural gas purification plant within a preset time period;

[0009] The detection module is configured to determine the correlation coefficient between any two instruments based on the time-series data corresponding to the two instruments, and to perform fault detection on the two instruments based on the relationship between the correlation coefficient and a preset interval, until the fault detection of all instruments in the natural gas purification plant is completed, and the fault detection results of the instruments in the natural gas purification plant are obtained.

[0010] Thirdly, this disclosure provides an electronic device, including:

[0011] A memory on which computer programs are stored;

[0012] A processor for executing the computer program in the memory to implement the method described in the first aspect.

[0013] Fourthly, this disclosure provides a non-transitory computer-readable storage medium having a computer program stored thereon that, when executed by a processor, implements the method described in the first aspect.

[0014] Fifthly, this disclosure provides a non-transitory computer-readable storage medium having a computer program stored thereon that, when executed by a processor, implements the method described in the first aspect.

[0015] By employing the aforementioned technical solution, and acquiring time-series data of each instrument in the natural gas purification plant within a preset time period, and by obtaining rich information and historical data of various instrument devices in the natural gas purification plant in real time, accurate detection of multiple instrument fault modes can be achieved even in the absence of fault sample markers. For any two instruments, the correlation coefficient between the two instruments is determined based on their corresponding time-series data. Fault detection is then performed on the two instruments based on the relationship between the correlation coefficient and a preset interval, until fault detection is completed for all instruments in the natural gas purification plant, yielding the fault detection results. Based on the relationship between the correlation coefficient between two instruments and the preset interval, remote instrument fault detection is achieved through different fault detection modes, enabling fault detection of different types of instruments and demonstrating strong adaptability.

[0016] Other features and advantages of this disclosure will be described in detail in the following detailed description section. Attached Figure Description

[0017] The accompanying drawings are provided to further illustrate the present disclosure and form part of the specification. They are used together with the following detailed description to explain the present disclosure, but do not constitute a limitation thereof. In the drawings:

[0018] Figure 1 This is a flowchart illustrating an instrument fault detection method according to an exemplary embodiment of the present disclosure.

[0019] Figure 2 This is another flowchart illustrating an instrument fault detection method according to exemplary embodiments of the present disclosure.

[0020] Figure 3 This is a flowchart illustrating a dynamic time warping detection mode according to an exemplary embodiment of the present disclosure.

[0021] Figure 4 This is a flowchart illustrating a waveform index detection mode and a coefficient of variation detection mode according to exemplary embodiments of this disclosure.

[0022] Figure 5 This is a block diagram illustrating an instrument fault detection device according to exemplary embodiments of the present disclosure.

[0023] Figure 6 This is a block diagram illustrating an electronic device according to exemplary embodiments of the present disclosure. Detailed Implementation

[0024] The specific embodiments of this disclosure will be described in detail below with reference to the accompanying drawings. It should be understood that the specific embodiments described herein are for illustration and explanation only and are not intended to limit this disclosure.

[0025] It should be noted that all actions involving the acquisition of signals, information, or data in this disclosure are carried out in compliance with the relevant data protection laws and policies of the country where the location is situated, and with authorization from the owner of the relevant device.

[0026] As mentioned in the background section, traditional instrument maintenance methods suffer from low inspection efficiency, low fault analysis efficiency, and the inability of process operators to monitor instrument status online. Statistical analysis of a large amount of real industrial data reveals that the variance of instrument measurement signal variations can effectively characterize the stable state of instrument performance.

[0027] In related technologies, differential time series based on real-time instrument signals are processed by filtering and removing singular points. Then, variance estimation is performed on the preprocessed data to extract the fluctuation status information of the instrument. Then, based on the variance statistics classification threshold learned from historical fault data, the current instrument status information is obtained by comparison, thereby realizing the corresponding fault detection.

[0028] The inventors found that the above method does not have multiple detection modes, and for complex systems and situations with multiple factors, variance as a single indicator may not be sufficient to fully reveal the complexity of the fault.

[0029] In view of this, the present disclosure provides an instrument fault detection method, device and electronic equipment, which can realize remote fault detection of different types of instruments through different fault detection modes.

[0030] Figure 1 This is a flowchart illustrating an instrument fault detection method according to an exemplary embodiment of this disclosure. Figure 1 As shown, the instrument fault detection method may include the following steps:

[0031] In step S101, time-series data of each instrument in the natural gas purification plant within a preset time period are obtained.

[0032] The preset time period can be preset according to the detection accuracy of the instrument, and this disclosure does not limit it.

[0033] In step S102, for any two instruments, the correlation coefficient between the two instruments is determined based on the time series data corresponding to the two instruments, and the two instruments are fault detected based on the relationship between the correlation coefficient and the preset interval, until the fault detection of all instruments in the natural gas purification plant is completed, and the fault detection results of the instruments in the natural gas purification plant are obtained.

[0034] In this embodiment, the preset interval can be preset according to the accuracy requirements of instrument fault diagnosis. This disclosure takes the preset interval as [1, 0.7]. Accordingly, based on the relationship between the correlation coefficient and the preset interval, the corresponding detection mode is determined from multiple preset detection modes to perform fault detection on the two instruments until the fault detection of all instruments is completed. The preset detection modes include, but are not limited to, dynamic time warping detection mode, waveform index detection mode, and coefficient of variation detection mode.

[0035] In the above technical solution, by acquiring the time-series data of each instrument in the natural gas purification plant within a preset time period, and by acquiring rich information and historical data of various instrument equipment in the natural gas purification plant in real time, accurate detection of multiple instrument fault modes can be achieved even in the absence of fault sample markers. For any two instruments, the correlation coefficient between the two instruments is determined based on their corresponding time-series data. Based on the relationship between the correlation coefficient and a preset interval, fault detection is performed on the two instruments until fault detection of all instruments in the natural gas purification plant is completed, yielding the fault detection results. Based on the relationship between the correlation coefficient between two instruments and the preset interval, remote instrument fault detection is achieved through different fault detection modes, enabling fault detection of different types of instruments and demonstrating strong adaptability.

[0036] To enable those skilled in the art to better understand the instrument fault detection method provided in this disclosure, the following detailed examples illustrate each step of the method.

[0037] In one feasible embodiment, the two instruments include a first instrument and a second instrument, and determining the correlation coefficient between the two instruments based on the time-series data corresponding to the two instruments includes:

[0038] The covariance is determined based on the first time-series data corresponding to the first instrument and the second time-series data corresponding to the second instrument;

[0039] A first variance is determined based on the first time series data, a second variance is determined based on the second time series data, and the square root of the product of the first variance and the second variance is determined.

[0040] Dividing the covariance by the square root yields the correlation coefficient between the first instrument and the second instrument.

[0041] In this embodiment, the first time-series data corresponding to the first instrument and the second time-series data corresponding to the second instrument are substituted into the following calculation formula to obtain the correlation coefficient r(X,Y) between the first instrument and the second instrument:

[0042]

[0043] Where X represents the first time series data corresponding to the first instrument, Y represents the second time series data corresponding to the second instrument, Cov(X,Y) represents the covariance of X and Y, Var[X] represents the variance of X, and Var[Y] represents the variance of Y.

[0044] In a feasible embodiment, the step of detecting faults in the two instruments based on the relationship between the correlation coefficient and a preset interval includes:

[0045] If the correlation coefficient is equal to 1, it is determined that neither of the two instruments has malfunctioned;

[0046] When the correlation coefficient is less than 1 and greater than the preset correlation coefficient threshold, the two instruments are subjected to fault detection according to the dynamic time warping detection mode.

[0047] If the correlation coefficient is less than the preset correlation coefficient threshold, fault detection is performed on the two instruments according to the waveform index detection mode and the coefficient of variation detection mode.

[0048] In this embodiment, as Figure 2 As shown, the correlation coefficient r between each instrument is calculated; when r = 1, no faults occur in any instrument; when 0.7 < r < 1, fault detection is performed on the two instruments according to the dynamic time warping detection mode; when r < 0.7, fault detection is performed on the two instruments according to the waveform index detection mode and the coefficient of variation detection mode; the detection results are output.

[0049] In the above technical solution, different fault detection modes are selected based on the relationship between the correlation coefficient and the preset interval to realize remote detection of instrument faults. By analyzing the correlation information between different instruments, fault detection of different types of instruments can be realized, thereby improving the accuracy of fault detection of instruments in natural gas purification plants.

[0050] In a feasible embodiment, the fault detection of the two instruments according to the dynamic time warping detection mode includes:

[0051] Determine the dynamic time warping distance between the data points of the time series data corresponding to the two instruments at each time moment to obtain the dynamic time warping distance sequence;

[0052] Determine the first waveform index of the dynamic time-warped distance sequence;

[0053] If the first waveform indicator is greater than the preset waveform indicator threshold, the two instruments are fault detected according to the waveform indicator detection mode.

[0054] If the first waveform index is less than the preset waveform index threshold, the two instruments are fault detected according to the waveform index detection mode and the coefficient of variation detection mode.

[0055] The preset waveform indicators can be preset according to the instrument model or according to the instrument's detection accuracy; this disclosure does not limit this.

[0056] In this embodiment, as Figure 3 As shown, the time series data of the instrument pair (two instruments) are imported, and the first waveform index of the dynamic time warp distance sequence is calculated based on the time series data. When the first waveform index is higher than the preset waveform index threshold, the two instruments are tested for faults through the waveform index detection mode. When the first waveform index is lower than the preset waveform index threshold, the two instruments are tested for faults according to the waveform index detection mode and the coefficient of variation detection mode. The test results of the instrument pair are summarized and output.

[0057] In one feasible embodiment, the fault detection of the two instruments according to the waveform index detection mode includes:

[0058] Determine the second waveform index corresponding to the first time series data, and the third waveform index corresponding to the second time series data;

[0059] When the second waveform indicator is greater than the third waveform indicator, it is determined that the first instrument has malfunctioned.

[0060] When the second waveform indicator is less than the third waveform indicator, it is determined that the second instrument has malfunctioned.

[0061] In this embodiment, as Figure 3 As shown, the waveform indicators corresponding to each instrument are calculated. If the waveform indicator of instrument 1 is high, it is determined that instrument 1 is faulty; if the waveform indicator of instrument 2 is high, it is determined that instrument 2 is faulty.

[0062] In a feasible embodiment, determining the dynamic time warping distance between data points of the time-series data corresponding to the two instruments at each time moment to obtain the dynamic time warping distance sequence includes:

[0063] For each time point, determine the first distance between the data points of the first time series data and the second time series data at the first time point, and the time weight corresponding to the first distance;

[0064] Based on the first distance and the time weight, and the shortest first distance between the first time series data and the second time series data at the second time, the dynamic time warping distance between the first instrument and the second instrument at the first time is calculated, where the second time is the time before the first time.

[0065] The dynamic time-normalized distances corresponding to each moment of the first instrument and the second instrument within the preset time period are arranged in chronological order to obtain a dynamic time-normalized distance sequence.

[0066] In this embodiment, the timing data D of the first instrument is... X Timing data D of the second instrument Y Substituting into the following formula, we obtain the dynamic time-warped distance between data points:

[0067] M(i,j)=(D X [i]-D Y [j]) 2 ,

[0068] Where M(i,j) represents the time series data D X The i-th value and the time series data D Y The distance between the j-th values, D X [i] represents the time-series data D of the first instrument. X The i-th value, D Y [j] Characterizes the time-series data D of the second instrument. Y The j-th value.

[0069] Time series data D X The i-th value and time series data D Y Substituting the j-th value into the following formula, we obtain the time weight W of M(i,j). ij :

[0070]

[0071] Where n represents the time series data D X The number of data points, m, represents the time series data D. Y The number of data points is then used to calculate the shortest distance of the dynamically time-normalized distance sequence by weighting it according to time weights.

[0072] R(i,j)=W(i,j)*D(i,j)+min[R(i-1,j),R(i,j-1),R(i-1,j-1)],

[0073] disc = R(n, m).

[0074] Wherein, min[R(i-1,j),R(i,j-1),R(i-1,j-1)] represents the shortest dynamic time warp distance at the previous time step.

[0075] Arrange all dynamic time-normalized distances in chronological order to obtain a dynamic time-normalized distance sequence.

[0076] In a feasible embodiment, the fault detection of the two instruments based on the waveform index detection mode and the coefficient of variation detection mode includes:

[0077] For either of the two instruments, determine the fourth waveform index and coefficient of variation based on the time-series data of the instrument;

[0078] If the fourth waveform indicator is greater than the preset waveform indicator threshold, or if the coefficient of variation is greater than the preset coefficient of variation threshold, it is determined that the instrument has malfunctioned.

[0079] If the fourth waveform indicator is less than the preset waveform indicator threshold, or the coefficient of variation is less than the preset coefficient of variation threshold, it is determined that the instrument has not malfunctioned.

[0080] The preset coefficient of variation threshold can be preset according to the instrument model or according to the detection accuracy requirements of the instrument. This disclosure does not limit this.

[0081] In this embodiment, as Figure 4 As shown, import the timing data of the instrument pair.

[0082] For each instrument, the waveform index is calculated. For instrument 1, if the waveform index is higher than the preset waveform index threshold, it is determined that instrument 1 has failed; if the waveform index is lower than the preset waveform index threshold, it is determined that instrument 1 has not failed. For instrument 2, if the waveform index is higher than the preset waveform index threshold, it is determined that instrument 2 has failed; if the waveform index is lower than the preset waveform index threshold, it is determined that instrument 2 has not failed.

[0083] The system calculates the coefficient of variation for each instrument. For instrument 1, if the coefficient of variation is higher than the preset coefficient of variation threshold, it determines that instrument 1 has malfunctioned; if the coefficient of variation is lower than the preset coefficient of variation threshold, it determines that instrument 1 has not malfunctioned. For instrument 2, if the coefficient of variation is higher than the preset coefficient of variation threshold, it determines that instrument 2 has malfunctioned; if the coefficient of variation is lower than the preset coefficient of variation threshold, it determines that instrument 2 has not malfunctioned.

[0084] In one feasible embodiment, determining the fourth waveform index and coefficient of variation based on the time-series data of the instrument includes:

[0085] Determine the root mean square, standard deviation, and mean of the time-series data of the instrument;

[0086] The fourth waveform index is determined based on the root mean square and the mean value;

[0087] The coefficient of variation is determined based on the standard deviation and the mean.

[0088] In this embodiment, the time-series data corresponding to the instrument are substituted into the following calculation formula to obtain the waveform index S of the instrument:

[0089] S = X rms / |μ|,

[0090] Among them, X rms The root mean square of time series data X is represented by μ, and the mean of time series data X is represented by μ.

[0091] Substitute the time-series data corresponding to the instrument into the following calculation formula to obtain the coefficient of variation (CV) of the instrument:

[0092] CV = σ / |μ|,

[0093] Here, σ represents the standard deviation of the time series data X.

[0094] The instrument fault detection method disclosed herein achieves remote instrument fault detection by calculating correlation coefficients, time-weighted dynamic time warping, waveform indicators, and coefficients of variation, and based on threshold indicators, utilizing different fault detection modes. It has the capability to detect faults in different types of instruments, adapting to changes in production process conditions; the detection process is clear and highly interpretable, facilitating parameter adjustment based on performance during application; it does not rely on fitting a large number of historical fault samples, requiring less work to construct the fault detection model; it has high sensitivity and a low false alarm rate.

[0095] The instrument fault detection method provided in this disclosure is described below with a complete embodiment:

[0096] Taking a natural gas purification plant of a certain company as an example, the field instruments in the natural gas purification plant transmit signals via 4-20mA and HART protocol. The collected instrument data is stored in the InfluxDB time series database. The test fault samples are shown in Table 1:

[0097] Table 1

[0098]

[0099]

[0100] Based on the 11 instrument fault samples in Table 1, the same number of non-fault samples were added to form a test sample set. The output values ​​of different algorithms and the instrument fault detection results are shown in Table 2, where numbers 1 to 11 are fault samples and numbers 12 to 22 are non-fault samples.

[0101] Table 2

[0102]

[0103]

[0104] The confusion matrix can summarize the classification performance of the classifier relative to certain test data. The confusion matrix generated by the result is shown in Table 3. It can be seen that the accuracy of the detection result is 1.0 and the recall rate is 0.91.

[0105]

[0106] Based on the same inventive concept, this disclosure also provides an instrument fault detection device, such as... Figure 5 As shown, the instrument fault detection device includes:

[0107] The acquisition module 501 is configured to acquire time-series data of each instrument in the natural gas purification plant within a preset time period;

[0108] The detection module 502 is configured to determine the correlation coefficient between any two instruments based on the time-series data corresponding to the two instruments, and to perform fault detection on the two instruments based on the relationship between the correlation coefficient and a preset interval, until the fault detection of all instruments in the natural gas purification plant is completed, and the fault detection results of the instruments in the natural gas purification plant are obtained.

[0109] In the above technical solution, by acquiring the time-series data of each instrument in the natural gas purification plant within a preset time period, and by acquiring rich information and historical data of various instrument equipment in the natural gas purification plant in real time, accurate detection of multiple instrument fault modes can be achieved even in the absence of fault sample markers. For any two instruments, the correlation coefficient between the two instruments is determined based on their corresponding time-series data. Based on the relationship between the correlation coefficient and a preset interval, fault detection is performed on the two instruments until fault detection of all instruments in the natural gas purification plant is completed, yielding the fault detection results. Based on the relationship between the correlation coefficient between two instruments and the preset interval, remote instrument fault detection is achieved through different fault detection modes, enabling fault detection of different types of instruments and demonstrating strong adaptability.

[0110] Furthermore, when the two instruments include a first instrument and a second instrument, the detection module 502 is configured to determine the covariance based on the first time-series data corresponding to the first instrument and the second time-series data corresponding to the second instrument.

[0111] A first variance is determined based on the first time series data, a second variance is determined based on the second time series data, and the square root of the product of the first variance and the second variance is determined.

[0112] Dividing the covariance by the square root yields the correlation coefficient between the first instrument and the second instrument.

[0113] Furthermore, the detection module 502 is configured to determine that neither of the two instruments has malfunctioned when the correlation coefficient is equal to 1;

[0114] When the correlation coefficient is less than 1 and greater than the preset correlation coefficient threshold, the two instruments are subjected to fault detection according to the dynamic time warping detection mode.

[0115] If the correlation coefficient is less than the preset correlation coefficient threshold, fault detection is performed on the two instruments according to the waveform index detection mode and the coefficient of variation detection mode.

[0116] Furthermore, the detection module 502 is configured to determine the dynamic time warping distance between the data points of the time-series data corresponding to the two instruments at each time moment, and obtain a dynamic time warping distance sequence.

[0117] Determine the first waveform index of the dynamic time-warped distance sequence;

[0118] If the first waveform indicator is greater than the preset waveform indicator threshold, the two instruments are fault detected according to the waveform indicator detection mode.

[0119] If the first waveform index is less than the preset waveform index threshold, the two instruments are fault detected according to the waveform index detection mode and the coefficient of variation detection mode.

[0120] Furthermore, the detection module 502 is configured to determine the second waveform index corresponding to the first time series data and the third waveform index corresponding to the second time series data.

[0121] When the second waveform indicator is greater than the third waveform indicator, it is determined that the first instrument has malfunctioned.

[0122] When the second waveform indicator is less than the third waveform indicator, it is determined that the second instrument has malfunctioned.

[0123] Furthermore, the detection module 502 is configured to determine, for each time moment, a first distance between the data points of the first time-series data and the second time-series data at the first time moment, and the time weight corresponding to the first distance;

[0124] Based on the first distance and the time weight, and the shortest first distance between the first time series data and the second time series data at the second time, the dynamic time warping distance between the first instrument and the second instrument at the first time is calculated, where the second time is the time before the first time.

[0125] The dynamic time-normalized distances corresponding to each moment of the first instrument and the second instrument within the preset time period are arranged in chronological order to obtain a dynamic time-normalized distance sequence.

[0126] Furthermore, the detection module 502 is configured to determine a fourth waveform index and a coefficient of variation for either of the two instruments based on the time-series data of the instrument.

[0127] If the fourth waveform indicator is greater than the preset waveform indicator threshold, or if the coefficient of variation is greater than the preset coefficient of variation threshold, it is determined that the instrument has malfunctioned.

[0128] If the fourth waveform indicator is less than the preset waveform indicator threshold, or the coefficient of variation is less than the preset coefficient of variation threshold, it is determined that the instrument has not malfunctioned.

[0129] Furthermore, the detection module 502 is configured to determine the root mean square, standard deviation, and mean of the time-series data of the instrument;

[0130] The fourth waveform index is determined based on the root mean square and the mean value;

[0131] The coefficient of variation is determined based on the standard deviation and the mean.

[0132] Regarding the instrument fault detection device in the above embodiments, the specific way in which each module performs its operation has been described in detail in the embodiments related to the method, and will not be elaborated here.

[0133] Based on the same inventive concept, this disclosure also provides an electronic device, comprising:

[0134] A memory on which computer programs are stored;

[0135] A processor is used to execute the computer program in the memory to implement the steps of the instrument fault detection method described above.

[0136] Figure 6This is a block diagram illustrating an electronic device 600 according to an exemplary embodiment. For example... Figure 6 As shown, the electronic device 600 may include a processor 601 and a memory 602. The electronic device 600 may also include one or more of a multimedia component 603, an input / output (I / O) interface 604, and a communication component 605.

[0137] The processor 601 controls the overall operation of the electronic device 600 to complete all or part of the steps in the instrument fault detection method described above. The memory 602 stores various types of data to support the operation of the electronic device 600. This data may include, for example, instructions for any application or method operating on the electronic device 600, and application-related data such as timing data, correlation coefficients, fault detection results, etc. The memory 602 can be implemented by any type of volatile or non-volatile storage device or a combination thereof, such as Static Random Access Memory (SRAM), Electrically Erasable Programmable Read-Only Memory (EEPROM), Erasable Programmable Read-Only Memory (EPROM), Programmable Read-Only Memory (PROM), Read-Only Memory (ROM), magnetic storage, flash memory, magnetic disk, or optical disk. The multimedia component 603 may include a screen and audio components. The screen may be, for example, a touchscreen, and the audio component is used to output and / or input audio signals. For example, the audio component may include a microphone for receiving external audio signals. The received audio signals may be further stored in memory 602 or transmitted via communication component 605. The audio component also includes at least one speaker for outputting audio signals. I / O interface 604 provides an interface between processor 601 and other interface modules, such as a keyboard, mouse, buttons, etc. These buttons may be virtual or physical buttons. Communication component 605 is used for wired or wireless communication between the electronic device 600 and other devices. Wireless communication, such as Wi-Fi, Bluetooth, Near Field Communication (NFC), 2G, 3G, 4G, NB-IoT, eMTC, or other 5G technologies, or combinations thereof, is not limited here. Therefore, the corresponding communication component 605 may include: a Wi-Fi module, a Bluetooth module, an NFC module, etc.

[0138] In an exemplary embodiment, the electronic device 600 may be implemented by one or more application-specific integrated circuits (ASICs), digital signal processors (DSPs), digital signal processing devices (DSPDs), programmable logic devices (PLDs), field-programmable gate arrays (FPGAs), controllers, microcontrollers, microprocessors, or other electronic components to perform the instrument fault detection method described above.

[0139] In another exemplary embodiment, a computer-readable storage medium including program instructions is also provided, which, when executed by a processor, implement the steps of the instrument fault detection method described above. For example, the computer-readable storage medium may be the memory 602 including the program instructions described above, which may be executed by the processor 601 of the electronic device 600 to complete the instrument fault detection method described above.

[0140] In another exemplary embodiment, a computer program product is also provided, which includes a computer program executable by a programmable device, the computer program having a code portion for performing the instrument fault detection method described above when executed by the programmable device.

[0141] The preferred embodiments of this disclosure have been described in detail above with reference to the accompanying drawings. However, this disclosure is not limited to the specific details of the above embodiments. Within the scope of the technical concept of this disclosure, various simple modifications can be made to the technical solutions of this disclosure, and these simple modifications all fall within the protection scope of this disclosure.

[0142] It should also be noted that the various specific technical features described in the above embodiments can be combined in any suitable manner without contradiction. To avoid unnecessary repetition, this disclosure will not describe the various possible combinations separately.

[0143] Furthermore, various different embodiments of this disclosure can be combined in any way, as long as they do not violate the spirit of this disclosure, they should also be regarded as the content disclosed in this disclosure.

Claims

1. A method for detecting instrument faults, characterized in that, The method includes: Acquire time-series data of each instrument in the natural gas purification plant within a preset time period; For any two instruments, the correlation coefficient between the two instruments is determined based on the time series data corresponding to the two instruments. Then, based on the relationship between the correlation coefficient and the preset interval, fault detection is performed on the two instruments until the fault detection of all instruments in the natural gas purification plant is completed, and the fault detection results of the instruments in the natural gas purification plant are obtained.

2. The instrument fault detection method according to claim 1, characterized in that, The two instruments include a first instrument and a second instrument. Determining the correlation coefficient between the two instruments based on their corresponding time-series data includes: The covariance is determined based on the first time-series data corresponding to the first instrument and the second time-series data corresponding to the second instrument; A first variance is determined based on the first time series data, a second variance is determined based on the second time series data, and the square root of the product of the first variance and the second variance is determined. Dividing the covariance by the square root yields the correlation coefficient between the first instrument and the second instrument.

3. The instrument fault detection method according to claim 2, characterized in that, The step of detecting faults in the two instruments based on the relationship between the correlation coefficient and the preset interval includes: If the correlation coefficient is equal to 1, it is determined that neither of the two instruments has malfunctioned; When the correlation coefficient is less than 1 and greater than the preset correlation coefficient threshold, the two instruments are subjected to fault detection according to the dynamic time warping detection mode. If the correlation coefficient is less than the preset correlation coefficient threshold, fault detection is performed on the two instruments according to the waveform index detection mode and the coefficient of variation detection mode.

4. The instrument fault detection method according to claim 3, characterized in that, The fault detection of the two instruments according to the dynamic time warping detection mode includes: Determine the dynamic time warping distance between the data points of the time series data corresponding to the two instruments at each time moment to obtain the dynamic time warping distance sequence; Determine the first waveform index of the dynamic time-warped distance sequence; If the first waveform indicator is greater than the preset waveform indicator threshold, the two instruments are fault detected according to the waveform indicator detection mode. If the first waveform index is less than the preset waveform index threshold, the two instruments are fault detected according to the waveform index detection mode and the coefficient of variation detection mode.

5. The instrument fault detection method according to claim 4, characterized in that, The fault detection of the two instruments according to the waveform index detection mode includes: Determine the second waveform index corresponding to the first time series data, and the third waveform index corresponding to the second time series data; When the second waveform indicator is greater than the third waveform indicator, it is determined that the first instrument has malfunctioned. When the second waveform indicator is less than the third waveform indicator, it is determined that the second instrument has malfunctioned.

6. The instrument fault detection method according to claim 4, characterized in that, The step of determining the dynamic time warping distance between the data points of the time-series data corresponding to the two instruments at each time moment, and obtaining the dynamic time warping distance sequence, includes: For each time point, determine the first distance between the data points of the first time series data and the second time series data at the first time point, and the time weight corresponding to the first distance; Based on the first distance and the time weight, and the shortest first distance between the first time series data and the second time series data at the second time, the dynamic time warping distance between the first instrument and the second instrument at the first time is calculated, where the second time is the time before the first time. The dynamic time-normalized distances corresponding to each moment of the first instrument and the second instrument within the preset time period are arranged in chronological order to obtain a dynamic time-normalized distance sequence.

7. The instrument fault detection method according to claim 4, characterized in that, The fault detection of the two instruments based on the waveform index detection mode and the coefficient of variation detection mode includes: For either of the two instruments, determine the fourth waveform index and coefficient of variation based on the time-series data of the instrument; If the fourth waveform indicator is greater than the preset waveform indicator threshold, or if the coefficient of variation is greater than the preset coefficient of variation threshold, it is determined that the instrument has malfunctioned. If the fourth waveform indicator is less than the preset waveform indicator threshold, or the coefficient of variation is less than the preset coefficient of variation threshold, it is determined that the instrument has not malfunctioned.

8. The instrument fault detection method according to claim 7, characterized in that, The step of determining the fourth waveform index and coefficient of variation based on the time-series data of the instrument includes: Determine the root mean square, standard deviation, and mean of the time-series data of the instrument; The fourth waveform index is determined based on the root mean square and the mean value; The coefficient of variation is determined based on the standard deviation and the mean.

9. An instrument fault detection device, characterized in that, The device includes: The acquisition module is configured to acquire time-series data of each instrument in the natural gas purification plant within a preset time period; The detection module is configured to determine the correlation coefficient between any two instruments based on the time-series data corresponding to the two instruments, and to perform fault detection on the two instruments based on the relationship between the correlation coefficient and a preset interval, until the fault detection of all instruments in the natural gas purification plant is completed, and the fault detection results of the instruments in the natural gas purification plant are obtained.

10. An electronic device, characterized in that, include: A memory on which computer programs are stored; A processor for executing the computer program in the memory to implement the method of any one of claims 1-9.

Images