A fault detection method and device of a plunger pump and an electronic device

Abstract

The application discloses a fault detection method and device of a plunger pump and electronic equipment. The method comprises the following steps: obtaining running state data of the plunger pump, wherein the running state data comprises vibration state data and rotation state data; performing feature extraction based on the running state data to obtain state features of the plunger pump; constructing an abnormality judgment index according to the running state data and / or the state features, and determining whether to generate an abnormality early warning according to the abnormality judgment index; in response to the generation of the abnormality early warning, generating a fault detection result according to the state features and a pre-trained fault detection model; and in response to the non-generation of the abnormality early warning, generating a health state detection result according to the state features, a pre-trained performance degradation model and historical health state data. The abnormality early warning is performed in advance based on the running state data of the plunger pump, and further fault detection or health state detection is performed, so that the detection of faults that are difficult to trigger parameter changes such as oil temperature, oil pressure and liquid level can be realized.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of plunger pumps, and in particular to a plunger pump fault detection method and device and electronic equipment. BACKGROUND

[0002] A mine reciprocating plunger pump is a core device in a liquid supply system of a coal mining face, which can provide hydraulic power for hydraulic support operation, and can be used for spraying dust and cooling of a power transmission system of a coal mining machine, and safe and stable operation of the mine reciprocating plunger pump plays an important role in efficient coal mining.

[0003] By installing a sensor on the mine reciprocating plunger pump to collect equipment state information, and then realizing fault detection through data analysis, occurrence of malignant faults is prevented, and economic losses are reduced.

[0004] In related technologies, plunger pump fault detection mainly relies on slowly changing state data such as oil temperature, oil pressure, and liquid level, and is realized in combination with a threshold alarm method. However, such a detection method can only preliminarily judge some performance degradation faults, and cannot timely detect faults that cannot timely trigger changes in oil temperature, oil pressure, and liquid level when faults occur. SUMMARY

[0005] The present application aims to at least partially solve one of the technical problems in the related art.

[0006] To this end, a first purpose of the present application is to provide a plunger pump fault detection method for solving the problem of detecting faults that cannot timely trigger changes in oil temperature, oil pressure, and liquid level in the related art.

[0007] To achieve the above purpose, a first aspect of an embodiment of the present application provides a plunger pump fault detection method, which includes: acquiring running state data of the plunger pump, the running state data including vibration state data and rotation state data; performing feature extraction based on the running state data to obtain state features of the plunger pump; constructing an abnormality judgment index according to the running state data and / or the state features, and determining whether to generate an abnormality warning according to the abnormality judgment index; in response to generating the abnormality warning, generating a fault detection result according to the state features and a pre-trained fault detection model; and in response to not generating the abnormality warning, generating a health state detection result according to the state features, a pre-trained performance degradation model, and historical health state data.

[0008] In addition, the plunger pump fault detection method according to the above embodiment of the present application can have the following additional technical features:

[0009] According to one of the embodiments of the present application, the vibration state data comprises a pump valve vibration signal of the plunger pump and a crankcase vibration signal of the plunger pump, and the rotation state data comprises a crankshaft keying signal of the plunger pump.

[0010] According to one of the embodiments of the present application, the feature extraction based on the operation state data to obtain the state feature of the plunger pump comprises: performing whole-period signal interception on the pump valve vibration signal and the crankcase vibration signal respectively based on the crankshaft keying signal to obtain whole-period pump valve vibration signal and whole-period crankcase vibration signal; performing resampling on the whole-period pump valve vibration signal and the whole-period crankcase vibration signal respectively based on order ratio tracking algorithm to obtain corresponding candidate vibration signals; and performing feature extraction on the candidate vibration signals to obtain the state feature.

[0011] According to one of the embodiments of the present application, the state feature at least comprises one of the following: time domain feature, frequency domain feature, angle domain feature and time-frequency domain feature.

[0012] According to one of the embodiments of the present application, the construction of the abnormality judgment index based on the operation state data and / or the state feature and the determination of whether to generate the abnormality early warning based on the abnormality judgment index comprises: performing signal reconstruction on the candidate vibration signal obtained based on the operation state data to obtain signal reconstruction error, taking the signal reconstruction error as the abnormality judgment index; and / or performing feature selection and dimension reduction fusion processing on the state feature to obtain abnormality feature, taking the value of the abnormality feature as the abnormality judgment index; comparing the abnormality judgment index with an abnormality early warning threshold corresponding to the abnormality judgment index currently, and determining whether to generate the abnormality early warning according to the comparison result.

[0013] According to one of the embodiments of the present application, the abnormality early warning threshold is an initial threshold or a self-adaptive updated threshold.

[0014] According to one of the embodiments of the present application, the generation of the fault detection result based on the state feature and a pre-trained fault detection model in response to the generation of the abnormality early warning comprises: performing feature selection and dimension reduction fusion processing on the state feature to obtain fault feature of the plunger pump; and generating the fault detection result based on the fault feature and the fault detection model, wherein the fault detection model is a classification model.

[0015] According to one embodiment of the present application, in response to the abnormal early warning not being generated, the health state detection result is generated according to the state feature, the pre-trained performance degradation model and historical health state data, including: performing feature selection and dimension reduction fusion processing on the state feature to obtain health fusion features of the plunger pump; and generating the health state detection result according to the health fusion features, the performance degradation model and the historical health state data.

[0016] To achieve the above object, the second aspect of the present application provides a plunger pump fault detection device, which comprises: a data acquisition module configured to acquire running state data of the plunger pump, the running state data comprising vibration state data and rotation state data; a preprocessing module configured to perform feature extraction based on the running state data to obtain state features of the plunger pump; an abnormal early warning module configured to construct an abnormality judgment index according to the running state data and / or the state features, and determine whether to generate an abnormal early warning according to the abnormality judgment index; a fault detection module configured to, in response to the abnormal early warning being generated, generate a fault detection result according to the state features and a pre-trained fault detection model; and a health state detection module configured to, in response to the abnormal early warning not being generated, generate a health state detection result according to the state features, a pre-trained performance degradation model and historical health state data.

[0017] In addition, the plunger pump fault detection device according to the above embodiments of the present application can have the following additional technical features:

[0018] According to one embodiment of the present application, the vibration state data comprises a pump valve vibration signal of the plunger pump and a crankcase vibration signal of the plunger pump, and the rotation state data comprises a crankshaft keying signal of the plunger pump.

[0019] According to one embodiment of the present application, the preprocessing module is further configured to perform whole-period signal interception on the pump valve vibration signal and the crankcase vibration signal respectively according to the crankshaft keying signal to obtain whole-period pump valve vibration signals and whole-period crankcase vibration signals; perform resampling on the whole-period pump valve vibration signals and the whole-period crankcase vibration signals respectively based on an order ratio tracking algorithm to obtain corresponding candidate vibration signals; and perform feature extraction on the candidate vibration signals to obtain the state features.

[0020] According to one embodiment of the present application, the state features at least include one of the following: time domain features, frequency domain features, angle domain features and time-frequency domain features.

[0021] According to an embodiment of the present application, the abnormality early warning module is further configured to: perform signal reconstruction on the candidate vibration signal obtained based on the running state data to obtain a signal reconstruction error, and take the signal reconstruction error as the abnormality judgment index; and / or perform feature selection and dimension reduction fusion processing on the state features to obtain abnormality features, and take a value of the abnormality features as the abnormality judgment index; compare the abnormality judgment index with an abnormality early warning threshold corresponding to the abnormality judgment index, and determine whether to generate the abnormality early warning according to a comparison result.

[0022] According to an embodiment of the present application, the abnormality early warning threshold is an initial threshold or an adaptively updated threshold.

[0023] According to an embodiment of the present application, the fault detection module is further configured to: perform feature selection and dimension reduction fusion processing on the state features to obtain fault features of the plunger pump; and generate the fault detection result according to the fault features and a fault detection model, the fault detection model being a classification model.

[0024] According to an embodiment of the present application, the health state detection module is further configured to: perform feature selection and dimension reduction fusion processing on the state features to obtain health fusion features of the plunger pump; and generate a health state detection result according to the health fusion features, the performance degradation model and the historical health state data.

[0025] To achieve the above object, a third aspect of the present application provides an electronic device, comprising a memory, a processor and a computer program stored in the memory and capable of running on the processor, wherein the processor executes the program to implement the fault detection method of the plunger pump according to any one of the first aspect of the present application.

[0026] The embodiment of the present application provides a fault detection method of a plunger pump, operation state data of the plunger pump is acquired, the operation state data includes vibration state data and rotation state data; feature extraction is performed based on the operation state data, and state features of the plunger pump are obtained; an abnormality judgment index is constructed according to the operation state data and / or the state features, and whether to generate an abnormality early warning is determined according to the abnormality judgment index; in response to generating the abnormality early warning, a fault detection result is generated according to the state features and a pre-trained fault detection model; and in response to not generating the abnormality early warning, a health state detection result is generated according to the state features, a pre-trained performance degradation model and historical health state data. The embodiment of the present application performs abnormality early warning on the plunger pump based on the vibration state data and the rotation state data of the plunger pump, further performs fault detection or health state detection according to the abnormality early warning result, analyzes the working state of parts of the plunger pump, and thus timely detection of faults that are difficult to timely trigger parameter changes such as oil temperature, oil pressure and liquid level is realized. BRIEF DESCRIPTION OF DRAWINGS

[0027] Figure 1 FIG. 1 is a flowchart of a fault detection method of a plunger pump according to an embodiment of the present application.

[0028] Figure 2 FIG. 2 is a flowchart of a fault detection method of a plunger pump according to another embodiment of the present application.

[0029] Figure 3 FIG. 3 is a flowchart of a fault detection method of a plunger pump according to another embodiment of the present application.

[0030] Figure 4 FIG. 4 is a flowchart of a fault detection method of a plunger pump according to another embodiment of the present application.

[0031] Figure 5 FIG. 5 is a flowchart of a fault detection method of a plunger pump according to another embodiment of the present application.

[0032] Figure 6 FIG. 6 is a system block diagram of an application system of a fault detection method of a plunger pump according to an embodiment of the present application.

[0033] Figure 7 FIG. 7 is a schematic diagram of an overall flow of a fault detection method of a plunger pump according to an embodiment of the present application.

[0034] Figure 8 FIG. 8 is a structural schematic diagram of a fault detection device of a plunger pump according to an embodiment of the present application.

[0035] Figure 9 FIG. 9 is a structural schematic diagram of an electronic device according to an embodiment of the present application. DETAILED DESCRIPTION

[0036] For a better understanding of the above technical solutions, the exemplary embodiments of the present application will be described in more detail below with reference to the accompanying drawings. Although the exemplary embodiments of the present application are shown in the accompanying drawings, it should be understood that the present application can be implemented in various forms and should not be limited by the embodiments set forth herein. On the contrary, these embodiments are provided so that the present disclosure can be more thoroughly understood and the scope of the present application can be accurately conveyed to those skilled in the art.

[0037] A fault detection method, device and electronic equipment of a plunger pump are described below with reference to the accompanying drawings.

[0038] Figure 1 is a flowchart of a fault detection method of a plunger pump disclosed in an embodiment of the present application.

[0039] As shown in Figure 1 , the fault detection method of the plunger pump proposed in the embodiment of the present application specifically includes the following steps:

[0040] S101, obtaining running state data of the plunger pump, the running state data including vibration state data and rotation state data.

[0041] The execution subject of the fault detection method of the plunger pump of the embodiment of the present application can be the fault detection device of the plunger pump provided in the embodiment of the present application, and the fault detection device of the plunger pump can be a hardware device with data information processing capability and / or necessary software required to drive the hardware device to work.

[0042] In the embodiment of the present application, the running state data of the plunger pump can include vibration state data and rotation state data, wherein the vibration state data can be a pump valve vibration signal of the plunger pump and a crankcase vibration signal of the plunger pump, and the rotation state data can be a crankshaft key signal of the plunger pump.

[0043] In some embodiments, the vertical vibration acceleration of the pump valve can be monitored by a vibration sensor installed on the plunger pump discharge cavity blocking plate to obtain the pump valve vibration signal; the crankshaft key signal can be monitored by three signal sensors installed on the crankshaft and a rotation speed sensor installed on the crankcase end cover, wherein the signal sensor is used to obtain the rotation angle of the crankshaft to increase the signal fineness of the crankshaft key signal, and in addition, the signal sensor can also be replaced by a toothed disc; and the vibration acceleration of the upper surface and the side surface of the crankcase can be detected by vibration sensors installed on the upper surface and the side surface of the crankcase body to obtain the crankcase vibration signal; in addition, the above pump valve vibration signal, crankshaft key signal and crankcase key signal can be collected by a data collector installed on the plunger pump train, and these signals are sent to a database as the running state data of the plunger pump for storage, so as to obtain the required data when performing fault detection.

[0044] S102, feature extraction is performed based on the operation state data to obtain a state feature of the plunger pump.

[0045] In the embodiments of the present application, feature extraction is performed on the operation state data of the plunger pump to obtain a state feature of the plunger pump, which can specifically include a state feature of a zero component in the plunger pump.

[0046] S103, an abnormality judgment index is constructed according to the operation state data and / or the state feature, and it is determined whether to generate an abnormality warning according to the abnormality judgment index.

[0047] In the embodiments of the present application, the abnormality judgment index corresponding to the plunger pump can be constructed according to the operation state data of the plunger pump, or the abnormality judgment index corresponding to the plunger pump can be constructed according to the state feature of the plunger pump. Abnormality warning judgment is performed on the plunger pump according to these abnormality judgment indexes to determine whether to generate an abnormality warning.

[0048] S104, in response to generating an abnormality warning, a fault detection result is generated according to the state feature and a pre-trained fault detection model.

[0049] In the case of generating an abnormality warning, fault detection is performed on the plunger pump according to the state feature of the plunger pump and a pre-trained fault detection model, and the model outputs a fault detection result.

[0050] S105, in response to not generating an abnormality warning, a health state detection result is generated according to the state feature, a pre-trained performance degradation model and historical health state data.

[0051] In the case of not generating an abnormality warning, the health state detection result can be further obtained according to the state feature of the plunger pump and a pre-trained performance degradation model and historical health state data of the plunger pump in different health states, wherein the health state detection result can include a health state level.

[0052] In summary, the fault detection method of the plunger pump in the embodiment of the application obtains the running state data of the plunger pump, the running state data including vibration state data and rotation state data; based on the running state data, feature extraction is performed to obtain the state feature of the plunger pump; according to the running state data and / or the state feature, an abnormality judgment index is constructed, and it is determined whether to generate an abnormality warning according to the abnormality judgment index; in response to generating the abnormality warning, a fault detection result is generated according to the state feature and a pre-trained fault detection model; in response to not generating the abnormality warning, a health state detection result is generated according to the state feature, a pre-trained performance degradation model and historical health state data. The embodiment of the application pre-warns the plunger pump based on the vibration state data and the rotation state data of the plunger pump, and further performs fault detection or health state detection according to whether the abnormality warning is generated, so as to analyze the working state of the components of the plunger pump, thereby realizing timely detection of faults that are difficult to timely trigger changes in parameters such as oil temperature, oil pressure and liquid level.

[0053] On the basis of the above-mentioned embodiments, as shown in Figure 2 The step of "performing feature extraction based on the running state data to obtain the state feature of the plunger pump" in the above-mentioned step S102 can include the following steps:

[0054] S201, according to the crankshaft key phase signal, the pump valve vibration signal and the crankcase vibration signal are respectively subjected to whole-cycle signal interception to obtain the whole-cycle pump valve vibration signal and the whole-cycle crankcase vibration signal.

[0055] In some embodiments, based on the continuously synchronously collected plunger pump vibration signals (such as the pump valve vibration signal and the crankcase vibration signal) and the crankshaft key phase signal, periodic signal interception is performed: according to the rising edge of the periodic pulse signal of the key phase signal, the vibration signals of each measuring point corresponding to 360 degrees or a multiple of 360 degrees of crank rotation are intercepted, so as to obtain the whole-cycle vibration signals (such as the whole-cycle pump valve vibration signal and the whole-cycle crankcase vibration signal) corresponding to each measuring point.

[0056] S202, based on the order ratio tracking algorithm, the whole-cycle pump valve vibration signal and the whole-cycle crankcase vibration signal are respectively resampled to obtain the corresponding candidate vibration signals.

[0057] In some embodiments, the angular domain resampling of the vibration signal is performed based on the order ratio tracking algorithm, constant angular interval sampling is used to replace constant time interval sampling, the sampling frequency of the vibration signal can be synchronized with the crankshaft rotation speed, and the whole-cycle sample length of the vibration signal is adaptively adjusted.

[0058] Wherein, the crankshaft cumulative rotation angle can be described as the following formula:

[0059] θ(t) = a0 + a1t + a2t2

[0060] According to the three signal sensors in the key phase signal monitoring time, the corresponding crankshaft rotation angles at three times can be obtained, as shown in the following equation group:

[0061]

[0062] In the formula, Δφ is the crankshaft rotation angle change value obtained according to the crankshaft key phase signal. According to the equation group, the coefficients a0, a1 and a2 can be calculated, and the expression of the arbitrary angle θ and the time t is obtained, that is, the time is expressed as a function of the cumulative rotation angle, as shown in the following formula:

[0063]

[0064] S203, feature extraction is performed on the candidate vibration signal to obtain a state feature.

[0065] In the embodiments of the present application, the whole-period pump valve vibration signal and the whole-period crankcase vibration signal obtained after resampling (that is, the above-mentioned candidate vibration signal) are subjected to feature extraction. For example, time domain, frequency domain, angular domain and time-frequency domain features are extracted from these whole-period vibration signals to obtain state features including time domain features, frequency domain features, angular domain features and time-frequency domain features. Of course, one or more of the time domain features, frequency domain features, angular domain features and time-frequency domain features can be extracted according to the needs, and the present application does not make any limitation.

[0066] On the basis of the above-mentioned embodiments, as shown in Figure 3 The step S103 of "constructing an abnormality judgment index according to the running state data and / or the state feature, and determining whether to generate an abnormality early warning according to the abnormality judgment index" includes the following steps:

[0067] S301, the candidate vibration signal obtained based on the running state data is subjected to signal reconstruction to obtain a signal reconstruction error, and the signal reconstruction error is taken as an abnormality judgment index.

[0068] In the embodiments of the present application, the signal reconstruction error of the whole-period pump valve vibration signal and the whole-period crankcase vibration signal obtained after resampling is used to construct the abnormality judgment index representing the zero components in the plunger pump.

[0069] S302, the state feature is subjected to feature selection and dimension reduction fusion processing to obtain an abnormality feature, and the value of the abnormality feature is taken as an abnormality judgment index.

[0070] In some embodiments, the abnormality judgment index representing the zero components in the plunger pump can also be constructed according to the obtained state feature, combined with the feature selection and dimension reduction processing technology.

[0071] S303, compare the abnormality judgment index with the abnormality warning threshold corresponding to the abnormality judgment index currently, and determine whether to generate an abnormality warning according to a comparison result.

[0072] In the embodiment of the application, the abnormality judgment index of each component calculated in real time is compared with the abnormality warning threshold corresponding to the index currently, and an abnormality warning is generated according to a comparison result.

[0073] The abnormality warning threshold is an initial threshold or an adaptive update threshold, wherein the initial threshold can be an average value of the abnormality evaluation state index calculated by the abnormality warning model in combination with the state data of the plunger pump at the factory test, and the adaptive update threshold can be an average value of the abnormality evaluation state index calculated by the abnormality warning model in combination with the state data of a whole day running six days ago.

[0074] On the basis of the above-mentioned embodiment, as shown in Figure 4 The step S104 of “generating a fault detection result according to the state feature and the pre-trained fault detection model in response to generating an abnormality warning” can include the following steps:

[0075] S401, performing feature selection and dimension reduction fusion processing on the state feature to obtain a fault feature of the plunger pump.

[0076] In the embodiment of the application, the fault detection is started after the abnormality warning occurs, and the fault feature representing the components in the plunger pump is constructed based on the state feature, in combination with the feature selection and the dimension reduction fusion technology of the feature.

[0077] S402, generating a fault detection result according to the fault feature and the fault detection model, and the fault detection model is a classification model.

[0078] On the basis of the above-mentioned embodiment, the real-time calculated fault feature is taken as the model input based on the pre-trained fault detection model of each component, and whether the fault occurs in the plunger pump component and what type of fault occurs are determined according to the output of the model. If the fault occurs, a fault alarm is issued, and if the fault does not occur, the current running state of the plunger pump is maintained.

[0079] On the basis of the above-mentioned embodiment, as shown in Figure 5 The step S105 of “generating a health state detection result according to the state feature, the pre-trained performance degradation model and the historical health state data in response to not generating an abnormality warning” includes the following steps:

[0080] S501, performing feature selection and dimension reduction fusion processing on the state feature to obtain a health fusion feature of the plunger pump.

[0081] In the case where no abnormal early warning occurs, health state detection is started, and based on state features, combined with feature selection and feature dimension reduction fusion technology, health fusion features representing components in the plunger pump are constructed.

[0082] In S502, health state detection results are generated according to the health fusion features, the performance degradation model and the historical health state data.

[0083] In some embodiments, the health state of the plunger pump is detected by using the health fusion features, combined with the performance degradation model of the components and the trend change law of the failure signs, and further combined with the analytic hierarchy process to obtain the health state grade of the plunger pump. The trend change law of the failure signs can be obtained based on historical data (i.e. historical health state data) of the plunger pump under different health states.

[0084] In some embodiments, the running state of the plunger pump, the running state data, the state features, the health detection results, the failure early warning and the failure detection results, etc. can also be displayed on the display device, which facilitates the plunger pump running management and maintenance decision-making of relevant personnel. When the plunger pump fails, further information related to the failure is output, including the failure position, the failure type and the failure elimination method.

[0085] In some embodiments, parameter configuration can also be performed in advance before failure detection, wherein the parameters to be configured can be technical parameters of the plunger pump (including flow, pressure, motor speed, gear ratio of the reduction box, movement sequence of the plunger in the liquid suction and discharge process, etc.), state monitoring parameters (including sensor arrangement information, sampling frequency, sampling point number, etc.), and parameters of the failure early warning model and the failure detection model (including model number, method principle of the model, structure of the model, data used for model training, framework used for the model, etc.).

[0086] In some embodiments, the running state data obtained online, the extracted state features and the data analysis results (such as failure early warning, failure detection and health detection results) can also be stored in the database, for example: for the vibration signals and keyphasers obtained in real time, data intensive storage is adopted, and the data coverage time interval is set to 7 days; for the state features and the data analysis results, they can be stored once every 0.1s, and no data coverage strategy is set.

[0087] In summary, the fault detection method of the plunger pump in the embodiment of the application obtains the running state data of the plunger pump, the running state data including vibration state data and rotation state data; based on the running state data, feature extraction is performed to obtain the state feature of the plunger pump; according to the running state data and / or the state feature, an abnormality judgment index is constructed, and it is determined whether to generate an abnormality early warning according to the abnormality judgment index; in response to generating the abnormality early warning, a fault detection result is generated according to the state feature and a pre-trained fault detection model; in response to not generating the abnormality early warning, a health state detection result is generated according to the state feature, a pre-trained performance degradation model and historical health state data. The embodiment of the application obtains more comprehensive plunger pump state information by increasing multiple sensor measurement points, preforms abnormality early warning on the plunger pump based on the vibration state data and the rotation state data of the plunger pump, and further performs fault detection or health state detection according to whether the abnormality early warning is generated, so as to analyze the working state of the components of the plunger pump. The application can be applied to abnormality early warning and fault detection of pump valve core or valve seat wear, spring fracture and other faults, can detect the component faults in the crankcase, and can analyze the health state of the components in the normal working state, so as to realize timely detection of faults that are difficult to timely trigger changes in oil temperature, oil pressure, liquid level and other parameters.

[0088] To clearly describe the fault detection method of the plunger pump in the embodiment of the application, the embodiment of the application is described in detail in combination with Figure 6 and Figure 7 The fault detection method of the plunger pump in the embodiment of the application can be applied to the fault detection system as shown in Figure 6 Figure 6 ​As shown, the fault detection system comprises a state monitoring device, a system parameter setting module, a data storage module, a data preprocessing module, an abnormality early warning module, a health detection module, a fault detection module and an interface display module. In implementation, the state monitoring device, such as a vibration sensor, a key phase sensor (or a signal sensor) and a key phase sensor, is installed in each component of the plunger pump. For example, the vibration sensor is installed on the plunger pump liquid end to obtain the pump valve vibration signal, the key phase sensor is installed on the crankshaft of the plunger pump power end to obtain the crankshaft rotation angle, the key phase sensor installed on the crankcase body obtains the high-precision crankshaft key phase signal, and the vibration sensor installed on the surface of the crankcase body obtains the crankcase vibration signal. The data collector stores the collected crankcase vibration signal, pump valve vibration signal and crankshaft key phase signal into the data storage module. The data preprocessing module obtains the running state data from the data storage module and extracts the state features according to the running state data. The abnormality early warning module determines whether to generate an abnormality early warning according to the state features. If the abnormality early warning is generated, the fault detection module is started. The fault detection module generates the fault detection result according to the state features obtained by the data preprocessing module. If the abnormality early warning is not generated, the health detection module is started. The health detection module generates the health state detection result according to the state features obtained by the data preprocessing module. During the process, the state features, the abnormality early warning, the fault detection result and the health state detection result can be displayed on the interface display module. In addition, the interface display module and the data collector can be set with relevant parameters according to the system parameter setting module, and the data preprocessing module, the abnormality early warning module, the health detection module and the fault detection module can be set with relevant parameters according to the data preprocessing module. Figure 6 The other modules not shown in connection with the system parameter setting module are set with relevant parameters.

[0089] Figure 7 The overall flowchart of the plunger pump fault detection method disclosed in an embodiment of the present application can include the following steps:

[0090] S701, obtaining the crankcase vibration signal, the crankshaft key phase signal and the pump valve vibration signal.

[0091] S702, according to the crankshaft key phase signal, the pump valve vibration signal and the crankcase vibration signal are respectively subjected to whole cycle signal interception to obtain the whole cycle pump valve vibration signal and the whole cycle crankcase vibration signal.

[0092] S703, based on the order ratio tracking algorithm, the whole cycle pump valve vibration signal and the whole cycle crankcase vibration signal are respectively subjected to resampling to obtain the corresponding candidate vibration signal.

[0093] S704, the candidate vibration signal is subjected to feature extraction to obtain the state feature.

[0094] S705, perform signal reconstruction on the candidate vibration signal obtained based on the operating status data to obtain the signal reconstruction error, and use the signal reconstruction error as the anomaly judgment index; and / or perform feature selection and dimensionality reduction fusion processing on the state features to obtain abnormal features, and use the value of the abnormal features as the anomaly judgment index.

[0095] S706, compare the anomaly judgment index with the anomaly warning threshold corresponding to the current anomaly judgment index, and determine whether to generate the anomaly warning based on the comparison result. If yes, proceed to steps S707 and S708; if no, proceed to step S712.

[0096] S707 displays an anomaly warning.

[0097] S708, in response to generating an anomaly warning, generates fault detection results based on state characteristics and a pre-trained fault detection model.

[0098] S709. Determine whether a fault has occurred based on the fault detection results. If yes, proceed to step S710; otherwise, proceed to step S711.

[0099] S710 issued a fault alarm.

[0100] S711, maintains the current operating status of the plunger pump.

[0101] S712, in response to the lack of anomaly warning, generates health status detection results based on state characteristics, a pre-trained performance degradation model, and historical health status data.

[0102] S713 displays the health status detection results.

[0103] Figure 8 This is a schematic diagram of the structure of a plunger pump fault detection device disclosed in one embodiment of this application.

[0104] like Figure 8 As shown, the fault detection device 800 for the plunger pump includes: a data acquisition module 801, a preprocessing module 802, an anomaly early warning module 803, a fault detection module 804, and a health status detection module 805. Among them,

[0105] The data acquisition module 801 is used to acquire the operating status data of the plunger pump, including vibration status data and rotation status data.

[0106] The preprocessing module 802 is used to extract features based on the operating status data to obtain the state features of the plunger pump.

[0107] The anomaly early warning module 803 is configured to construct an anomaly judgment index according to the running state data and / or the state feature, and determine whether to generate an anomaly early warning according to the anomaly judgment index.

[0108] The fault detection module 804 is configured to, in response to the generation of the anomaly early warning, generate a fault detection result according to the state feature and a pre-trained fault detection model.

[0109] The health state detection module 805 is configured to, in response to the non-generation of the anomaly early warning, generate a health state detection result according to the state feature, a pre-trained performance degradation model and historical health state data.

[0110] According to an embodiment of the present application, the vibration state data includes a pump valve vibration signal of the plunger pump and a crankcase vibration signal of the plunger pump, and the rotation state data includes a crank keying signal of the plunger pump.

[0111] According to an embodiment of the present application, the preprocessing module 802 is further configured to, according to the crank keying signal, respectively perform whole-cycle signal interception on the pump valve vibration signal and the crankcase vibration signal to obtain a whole-cycle pump valve vibration signal and a whole-cycle crankcase vibration signal; perform resampling on the whole-cycle pump valve vibration signal and the whole-cycle crankcase vibration signal based on an order ratio tracking algorithm to obtain corresponding candidate vibration signals; and perform feature extraction on the candidate vibration signals to obtain the state feature.

[0112] According to an embodiment of the present application, the state feature at least includes one of a time domain feature, a frequency domain feature, an angle domain feature and a time-frequency domain feature.

[0113] According to an embodiment of the present application, the anomaly early warning module 803 is further configured to: perform signal reconstruction on the candidate vibration signal obtained based on the running state data to obtain a signal reconstruction error, take the signal reconstruction error as an anomaly judgment index; and / or perform feature selection and dimension reduction fusion processing on the state feature to obtain an anomaly feature, take a value of the anomaly feature as an anomaly judgment index; compare the anomaly judgment index with an anomaly early warning threshold corresponding to the anomaly judgment index, and determine whether to generate an anomaly early warning according to a comparison result.

[0114] According to an embodiment of the present application, the anomaly early warning threshold is an initial threshold or a self-adaptive updated threshold.

[0115] According to an embodiment of the present application, the fault detection module 804 is further configured to: perform feature selection and dimension reduction fusion processing on the state feature to obtain a fault feature of the plunger pump; and generate the fault detection result according to the fault feature and a fault detection model, the fault detection model being a classification model.

[0116] According to one embodiment of the present application, the health state detection module 805 is further configured to: perform feature selection and dimension reduction fusion processing on the state features to obtain health fusion features of the plunger pump; and generate a health state detection result according to the health fusion features, the performance degradation model, and historical health state data.

[0117] It should be noted that the above explanation of the plunger pump fault detection method embodiment is also applicable to the plunger pump fault detection device of this embodiment, which will not be repeated here.

[0118] In summary, according to the plunger pump fault detection device provided in the embodiments of the present application, the running state data of the plunger pump is obtained, including vibration state data and rotation state data; feature extraction is performed based on the running state data to obtain state features of the plunger pump; an abnormality judgment index is constructed according to the running state data and / or the state features, and it is determined whether to generate an abnormality warning according to the abnormality judgment index; in response to generating the abnormality warning, a fault detection result is generated according to the state features and a pre-trained fault detection model; and in response to not generating the abnormality warning, a health state detection result is generated according to the state features, a pre-trained performance degradation model, and historical health state data. The embodiments of the present application obtain more comprehensive plunger pump state information by increasing multiple sensor measurement points, pre-raise an abnormality warning based on the vibration state data and the rotation state data of the plunger pump, and further perform fault detection or health state detection according to whether an abnormality warning is generated to analyze the working state of the components of the plunger pump. The present application can be applied to abnormality warning and fault detection of pump valve core or valve seat wear, spring breakage, and the like, can detect component faults in the crankcase, and can analyze the health state of components in a normal working state, achieving timely detection of faults that cannot trigger changes in oil temperature, oil pressure, liquid level, and the like in time.

[0119] To implement the above-mentioned embodiments, the present application further provides an electronic device 900, as shown in Figure 9 The electronic device 900 includes a memory 901, a processor 902, and a computer program stored in the memory 903 and executable on the processor 903. When the processor executes the program, the above-mentioned plunger pump fault detection method is implemented.

[0120] In the description of the present application, it should be understood that the terms "center", "longitudinal", "transverse", "length", "width", "thickness", "upper", "lower", "front", "back", "left", "right", "vertical", "horizontal", "top", "bottom", "inner", "outer", "clockwise", "counterclockwise", "axial", "radial", "circumferential" and the like indicate the orientation or positional relationship shown in the drawings, which are only for the convenience of describing the present application and simplifying the description, and do not indicate or imply that the device or element referred to must have a particular orientation, be constructed and operated in a particular orientation, and therefore cannot be understood as a limitation of the present application.

[0121] In addition, the terms "first", "second" are only for descriptive purposes and cannot be understood as indicating or implying relative importance or implicitly indicating the number of the technical features indicated. Therefore, the features defined with "first", "second" can explicitly or implicitly include one or more of the features. In the description of the present application, the meaning of "a plurality of" is two or more, unless otherwise explicitly specified and limited.

[0122] In the present application, unless otherwise explicitly specified and limited, the terms "mounting", "connecting", "connecting", "fixing" and the like should be understood broadly, for example, it can be fixed connection, or detachable connection, or integral; it can be mechanical connection, or electrical connection; it can be direct connection, or indirect connection through intermediate medium; it can be the internal communication of two elements or the interaction relationship between two elements. For those skilled in the art, the specific meaning of the above terms in the present application can be understood according to the specific circumstances.

[0123] In the present application, unless otherwise explicitly specified and limited, the first feature "on" or "under" the second feature can be direct contact between the first and second features, or indirect contact between the first and second features through intermediate medium. Moreover, the first feature "above", "over" and "on" the second feature can be directly above or obliquely above the first feature, or only indicate that the horizontal height of the first feature is higher than that of the second feature. The first feature "below", "under" and "under" the second feature can be directly below or obliquely below the first feature, or only indicate that the horizontal height of the first feature is less than that of the second feature.

[0124] In the description of the specification, the description of the terms "one embodiment", "some embodiments", "an example", "a specific example", or "some examples" etc. means that the specific features, structures, materials or characteristics described in connection with the embodiment or example are contained in at least one embodiment or example of the present application. In the specification, the illustrative description of the above terms does not necessarily refer to the same embodiment or example. Moreover, the specific features, structures, materials or characteristics described can be combined in any appropriate manner in any one or more embodiments or examples. In addition, the person skilled in the art can combine and combine the different embodiments or examples described in the specification and the features of the different embodiments or examples without contradiction.

[0125] Although the embodiments of the present application have been shown and described above, it is understood that the above-described embodiments are exemplary and are not to be construed as limiting the present application, and the person skilled in the art can make changes, modifications, replacements and variations to the above-described embodiments within the scope of the present application.

Claims

1. A method of fault detection for a plunger pump, characterized by, The method comprises: acquiring operation state data of the plunger pump, the operation state data comprising vibration state data and rotation state data, wherein the vibration state data comprises a pump valve vibration signal of the plunger pump and a crankcase vibration signal of the plunger pump, and the rotation state data comprises a crankshaft keying signal of the plunger pump; based on the operation state data, performing feature extraction to obtain state features of the plunger pump, wherein the state features comprise at least one of the following: time domain features, frequency domain features, angular domain features, and time-frequency domain features; constructing an abnormality judgment index according to the operation state data and / or the state features, and determining whether to generate an abnormality warning according to the abnormality judgment index; in response to generating the abnormality warning, generating a fault detection result according to the state features and a pre-trained fault detection model; in response to not generating the abnormality warning, generating a health state detection result according to the state features, a pre-trained performance degradation model, and historical health state data; wherein the constructing of the abnormality judgment index according to the operation state data and / or the state features comprises: performing whole-cycle signal interception on the pump valve vibration signal and the crankcase vibration signal respectively according to the crankshaft keying signal, to obtain whole-cycle pump valve vibration signals and whole-cycle crankcase vibration signals; based on an order ratio tracking algorithm, resampling the whole-cycle pump valve vibration signals and the whole-cycle crankcase vibration signals respectively to obtain corresponding candidate vibration signals; performing signal reconstruction on the candidate vibration signals to obtain signal reconstruction errors, and taking the signal reconstruction errors as the abnormality judgment index; and / or performing feature selection and dimension reduction fusion processing on the state features to obtain abnormal features, and taking values of the abnormal features as the abnormality judgment index.

2. The fault detection method according to claim 1, characterized in that, The method further comprises: performing feature extraction on the candidate vibration signals to obtain the state features.

3. The fault detection method of claim 1, wherein, The determining of whether to generate the abnormality warning according to the abnormality judgment index comprises: comparing the abnormality judgment index with an abnormality warning threshold currently corresponding to the abnormality judgment index, and determining whether to generate the abnormality warning according to a comparison result.

4. The fault detection method of claim 1, wherein, The abnormality warning threshold is an initial threshold or an adaptively updated threshold.

5. The fault detection method of claim 1, wherein, The generating of the fault detection result according to the state features and the pre-trained fault detection model in response to generating the abnormality warning comprises: performing feature selection and dimension reduction fusion processing on the state features to obtain fault features of the plunger pump; generating the fault detection result according to the fault features and the fault detection model, the fault detection model being a classification model.

6. The fault detection method of claim 1, wherein, The generating of the health state detection result according to the state features, the pre-trained performance degradation model, and the historical health state data in response to not generating the abnormality warning comprises: performing feature selection and dimension reduction fusion processing on the state features to obtain health fusion features of the plunger pump; generating the health state detection result according to the health fusion features, the performance degradation model, and the historical health state data.

7. A fault detection device for a plunger pump, characterized by The method comprises: The data acquisition module is configured to acquire operation state data of the plunger pump, the operation state data including vibration state data and rotation state data, wherein the vibration state data includes a pump valve vibration signal of the plunger pump and a crankcase vibration signal of the plunger pump, and the rotation state data includes a crankshaft keying signal of the plunger pump. The preprocessing module is configured to perform feature extraction based on the operation state data to obtain state features of the plunger pump, wherein the state features include at least one of the following: time domain features, frequency domain features, angular domain features, and time-frequency domain features. The abnormality early warning module is configured to construct an abnormality judgment index according to the operation state data and / or the state features, and determine whether to generate an abnormality early warning according to the abnormality judgment index. The fault detection module is configured to generate a fault detection result according to the state features and a pre-trained fault detection model in response to the generation of the abnormality early warning. The health state detection module is configured to generate a health state detection result according to the state features, a pre-trained performance degradation model, and historical health state data in response to the non-generation of the abnormality early warning. The abnormality early warning module is further configured to: perform whole-cycle signal interception on the pump valve vibration signal and the crankcase vibration signal respectively based on the crankshaft keying signal to obtain whole-cycle pump valve vibration signals and whole-cycle crankcase vibration signals; perform resampling on the whole-cycle pump valve vibration signals and the whole-cycle crankcase vibration signals respectively based on an order ratio tracking algorithm to obtain corresponding candidate vibration signals; perform signal reconstruction on the candidate vibration signals to obtain signal reconstruction errors, and use the signal reconstruction errors as the abnormality judgment index; and / or perform feature selection and dimension reduction fusion processing on the state features to obtain abnormal features, and use values of the abnormal features as the abnormality judgment index.

8. An electronic device, comprising: The computer program is stored in the memory and executable on the processor, and when the processor executes the program, the plunger pump fault detection method according to any one of claims 1-6 is implemented. The computer program is stored in the memory and executable on the processor, and when the processor executes the program, the plunger pump fault detection method according to any one of claims 1-6 is implemented.

Images