Fault warning method, device and plunger pump equipment system of plunger pump equipment

Abstract

The present application provides a fault warning method, device and plunger pump equipment system for a plunger pump device. The method includes: obtaining target data of a plurality of plunger pump devices; determining corresponding training data based on each target data to obtain a plurality of training data; training the plurality of training data, and determining to divide the plurality of plunger pump devices into a plurality of clusters based on the training results; determining a target cluster and a warning plunger pump device according to a plurality of historical fault event data. By training the training data corresponding to the target data of a plurality of plunger pump devices, dividing the plurality of plunger pump devices into a plurality of clusters based on the training results, and then determining the target cluster and the warning plunger pump device according to the plurality of historical fault event data, the present application can give warnings to a large number of devices, solves the problem in the prior art that warnings cannot be given to a large number of plunger pump devices simultaneously, and has good universality.

Description

[0001] This disclosure claims priority to the PCT patent application filed with the International Bureau of the World Intellectual Property Organization on December 23, 2021, with international application number PCT / CN2021 / 140973 and application title "Fault Warning Method, Device and Plunger Pump Equipment System for Plunger Pump Equipment", the entire content of which is incorporated herein by reference. Technical Field

[0002] The present application relates to the technical field of fault warning for plunger pump equipment. Specifically, it relates to a fault warning method for plunger pump equipment, a fault warning device, a computer-readable storage medium, a processor, and a plunger pump equipment system. Background Art

[0003] With the full spread of industrial digitization, there are more and more intelligent application scenarios for plunger pump equipment. Scenarios such as life prediction, health diagnosis, and fault detection of plunger pump equipment have also been greatly derived and developed. There are also various device fault diagnosis methods including various deep neural networks. However, the overall detection effect is not ideal and there are mainly the following problems: on the one hand, the detection object is only for a single device, and it is impossible to uniformly detect multiple devices in a large working site; on the other hand, the model is severely affected by working conditions and has poor universality.

[0004] Therefore, in the prior art, it is impossible to simultaneously give warnings for a large number of plunger pump equipment, and there is an urgent need for a method that can simultaneously give warnings for a large number of plunger pump equipment.

[0005] The above information disclosed in the background art section is only used to enhance the understanding of the background art of the technology described in this article. Therefore, the background art may contain certain information that is not prior art known to those skilled in the art in their own country. Summary of the Invention

[0006] The main purpose of the present application is to provide a fault warning method for plunger pump equipment, a fault warning device, a computer-readable storage medium, a processor, and a plunger pump equipment system to solve the problem in the prior art that it is impossible to simultaneously give warnings for a large number of plunger pump equipment.

[0007] According to one aspect of an embodiment of the present invention, a method for fault warning of a plunger pump device is provided, including: obtaining target data of a plurality of plunger pump devices, where the target data includes historical fault event data of the plunger pump device, file data of the plunger pump device, operation data of the plunger pump device, and maintenance data of the plunger pump device; based on each of the target data, determining corresponding training data to obtain a plurality of the training data, and each of the training data includes index values corresponding to a plurality of different characteristic indexes; training the plurality of training data, and based on the training result, determining to divide the plurality of plunger pump devices into a plurality of clusters; according to the plurality of historical fault event data, determining a target cluster and a warning plunger pump device, where the target cluster is the cluster including the plunger pump devices that have had faults, and the warning plunger pump device is at least the plunger pump device that has not had the fault in the target cluster.

[0008] Optionally, training the plurality of training data, and based on the training result, determining to divide the plurality of plunger pump devices into a plurality of clusters, includes: based on the K-means clustering algorithm, dividing the plurality of training data into G clustering data groups according to the grouping method corresponding to the best clustering effect, each of the clustering data groups corresponding to a plunger pump device type, each of the G groups corresponding to a best clustering center, and the G best clustering centers constituting a best clustering center data set; according to the index value and the best clustering center data set, calculating a corresponding best two-dimensional membership matrix, and determining the group corresponding to the maximum membership of the target plunger pump device according to the best two-dimensional membership matrix, and dividing the plurality of plunger pump devices into a plurality of clusters, where the plunger pump devices in the predetermined cluster have the maximum membership corresponding to the plunger pump device type corresponding to the predetermined cluster, and the predetermined cluster is any one of the clusters.

[0009] Optionally, based on the K-means clustering algorithm, multiple pieces of the training data are divided into G clusters of data according to the grouping method with the best clustering effect, including: a grouping step, based on the K-means clustering algorithm, dividing multiple pieces of the training data into K initial data groups and calculating K initial cluster centers, and the K initial cluster centers form an initial cluster center data set; a membership matrix calculation step, calculating a two-dimensional membership matrix according to the index value and the initial cluster center data set; a value function calculation step, calculating a value function based on the index value and the initial cluster center data set; in the case where the value function is less than a threshold, stopping the calculation and determining the current K initial data groups as K clusters of data groups, and G = K; in the case where the value function is greater than or equal to the threshold, updating the initial cluster centers and sequentially repeating the grouping step, the membership matrix calculation step, and the value function calculation step until the value function is less than the threshold and stopping, and determining the current updated multiple initial data groups as G clusters of data groups.

[0010] Optionally, based on multiple pieces of the historical fault event data, a target cluster group and a warning plunger pump device are determined, including: according to multiple pieces of the historical fault event data, determining all fault types that have occurred in all the plunger pump devices and generating enumeration type labels corresponding to different fault types, and labeling the corresponding fault type labels for the plunger pump devices that have had faults of the corresponding fault types; determining the cluster group including the plunger pump devices labeled with the fault type labels as the target cluster group; determining the fault type with the most occurrences in the target cluster group as the typical fault type; determining the plunger pump devices in the target cluster group that have not had faults of the typical fault type as the warning plunger pump devices.

[0011] Optionally, based on each piece of the target data, corresponding training data is determined to obtain multiple pieces of the training data, including: based on multiple pieces of the target data, determining multiple characteristic indicators, and the multiple characteristic indicators include archive type characteristic indicators of the plunger pump device, operation type characteristic indicators of the plunger pump device, maintenance type characteristic indicators of the plunger pump device, and fault type characteristic indicators of the plunger pump device; based on multiple pieces of the target data, determining the index values corresponding to the characteristic indicators to obtain each piece of the training data.

[0012] Optionally, based on multiple pieces of the target data, the index values corresponding to the characteristic indicators are determined to obtain each piece of the training data, including: preprocessing each piece of the target data to obtain numericalized data; converting the numericalized data into data under the same dimension to obtain unified dimension data; performing dimensionality reduction processing on the unified dimension data to obtain each piece of the training data.

[0013] Optionally, preprocess the multiple pieces of target data to obtain numerical data, including: performing word segmentation processing and vectorization processing on the unstructured text data in the target data in sequence; performing discretization processing on the continuous numerical fields in the target data; and performing one-hot encoding processing on the enumerated string data.

[0014] Optionally, perform dimensionality reduction processing on the unified dimension data to obtain each piece of training data, including: by means of the PCA algorithm, extracting a part of the multiple index values of each piece of unified dimension data as target index values to obtain the training data.

[0015] Optionally, after determining the target cluster and the warning plunger pump device, the method further includes: generating and outputting a warning list corresponding to the warning plunger pump device.

[0016] According to another aspect of the embodiments of the present invention, there is also provided a fault warning device for a plunger pump device, including an acquisition unit, a first determination unit, a training unit, and a second determination unit. Among them, the acquisition unit is used to acquire target data of multiple plunger pump devices, and the target data includes historical fault event data of the plunger pump device, file data of the plunger pump device, operation data of the plunger pump device, and maintenance data of the plunger pump device; the first determination unit is used to determine corresponding training data based on each piece of target data to obtain multiple pieces of training data, and each piece of training data includes index values corresponding to multiple different characteristic indicators; the training unit is used to train the multiple pieces of training data and determine to divide the multiple plunger pump devices into multiple clusters based on the training results; the second determination unit determines the target cluster and the warning plunger pump device according to the multiple pieces of historical fault event data, the target cluster is the cluster including the plunger pump devices that have had faults, and the warning plunger pump device is at least the plunger pump device that has not had the fault in the target cluster.

[0017] According to another aspect of the embodiments of the present invention, there is also provided a computer-readable storage medium, and the computer-readable storage medium includes a stored program, wherein the program executes any one of the methods.

[0018] According to another aspect of the embodiments of the present invention, there is also provided a processor, and the processor is used to run a program, wherein when the program runs, it executes any one of the methods.

[0019] According to another aspect of the embodiments of the present invention, a plunger pump device system is further provided, including a plunger pump device, one or more processors, a memory, and one or more programs, wherein the one or more programs are stored in the memory and configured to be executed by the one or more processors, and the one or more programs include those for executing any one of the above methods.

[0020] In the embodiments of the present invention, the above-mentioned method for fault warning of a plunger pump device includes: First, obtain historical fault event data of multiple plunger pump devices, archive data of the plunger pump device, operation data of the plunger pump device, and maintenance data of the plunger pump device as target data; Then, based on each of the above target data, determine corresponding training data including index values corresponding to multiple different characteristic indexes, and obtain multiple pieces of the above training data; After that, train multiple pieces of the above training data, and based on the training results, determine to divide multiple above-mentioned plunger pump devices into multiple cluster groups; Finally, according to multiple pieces of the above historical fault event data, determine a target cluster group and a warning plunger pump device, where the target cluster group is the above-mentioned cluster group including the plunger pump devices that have had faults, and the warning plunger pump device is at least the above-mentioned plunger pump device that has not had the above-mentioned faults in the above target cluster group. In this application, by training the training data corresponding to the target data of multiple plunger pump devices, the amount of target data is large, and the training data can be well trained. Based on the training results, multiple plunger pump devices are divided into multiple cluster groups. Then, according to multiple pieces of the above historical fault event data, a target cluster group and a warning plunger pump device are determined. According to the collaborative filtering theory, the probability of plunger pump devices with highly similar attributes having the same type of faults is extremely high. Therefore, a large number of devices can be warned, and the universality is good. Description of the Drawings

[0021] The specification drawings forming a part of this application are used to provide a further understanding of this application. The schematic embodiments of this application and their descriptions are used to explain this application and do not constitute an improper limitation to this application. In the drawings:

[0022] Figure 1 The flowchart of the method for fault warning of a plunger pump device according to the embodiments of this application is shown;

[0023] Figure 2 The flow logic diagram of the method for fault warning of a plunger pump device according to the embodiments of this application is shown;

[0024] Figure 3 The schematic diagram of the device for fault warning of a plunger pump device according to the embodiments of this application is shown. Detailed Embodiments

[0025] It should be noted that, without conflict, the embodiments in the present application and the features in the embodiments may be combined with each other. The following will describe the present application in detail with reference to the drawings and in combination with the embodiments.

[0026] In order to enable those skilled in the art to better understand the solution of the present application, the following will clearly and completely describe the technical solutions in the embodiments of the present application with reference to the accompanying drawings in the embodiments of the present application. Obviously, the described embodiments are only a part of the embodiments of the present application, rather than all the embodiments. Based on the embodiments in the present application, all other embodiments obtained by those of ordinary skill in the art without creative efforts shall fall within the protection scope of the present application.

[0027] It should be noted that the terms "first", "second", etc. in the specification and claims of the present application and the above-mentioned drawings are used to distinguish similar objects, and do not necessarily need to be used to describe a specific order or sequence. It should be understood that such used data can be interchanged under appropriate circumstances, so as to implement the embodiments of the present application described herein. In addition, the terms "comprising" and "having" and any variations thereof are intended to cover non-exclusive inclusion. For example, a process, method, system, product or device including a series of steps or units does not necessarily have to be limited to those steps or units clearly listed, but may include other steps or units not clearly listed or inherent to these processes, methods, products or devices.

[0028] It should be understood that when an element (such as a layer, film, region, or substrate) is described as being "on" another element, the element can be directly on the other element, or there can also be an intermediate element. Moreover, in the specification and claims, when an element is described as "connected" to another element, the element can be "directly connected" to the other element, or "connected" to the other element through a third element.

[0029] As described in the background art, it is impossible to simultaneously give early warnings for a large number of plunger pump devices in the prior art. To solve the above problems, in a typical implementation manner of the present application, a fault warning method, a fault warning device, a computer-readable storage medium, a processor, and a plunger pump device system for the plunger pump device are provided.

[0030] According to an embodiment of the present application, a fault warning method for a plunger pump device is provided.

[0031] Figure 1 is a flowchart of the fault warning method for the plunger pump device according to an embodiment of the present application. As Figure 1 shown, the method includes the following steps:

[0032] Step S101, obtain the target data of multiple plunger pump devices, where the above-mentioned target data includes the historical fault event data of the plunger pump devices, the archive data of the plunger pump devices, the operation data of the plunger pump devices, and the maintenance data of the plunger pump devices;

[0033] Step S102, based on each of the above-mentioned target data, determine the corresponding training data to obtain multiple pieces of the above-mentioned training data, and each of the above-mentioned training data includes the index values corresponding to multiple different feature indexes;

[0034] Step S103, train multiple pieces of the above-mentioned training data, and based on the training results, determine that multiple above-mentioned plunger pump devices are divided into multiple clusters;

[0035] Step S104, according to multiple pieces of the above-mentioned historical fault event data, determine the target cluster and the warning plunger pump devices. The above-mentioned target cluster is the cluster that includes the above-mentioned plunger pump devices that have had faults, and the above-mentioned warning plunger pump devices are at least the above-mentioned plunger pump devices that have not had the above-mentioned faults in the above-mentioned target cluster.

[0036] The above-mentioned fault warning method for plunger pump devices includes: First, obtain the historical fault event data of multiple plunger pump devices, the archive data of the plunger pump devices, the operation data of the plunger pump devices, and the maintenance data of the plunger pump devices as the target data; then, based on each of the above-mentioned target data, determine the corresponding training data including the index values corresponding to multiple different feature indexes to obtain multiple pieces of the above-mentioned training data; after that, train multiple pieces of the above-mentioned training data, and based on the training results, determine that multiple above-mentioned plunger pump devices are divided into multiple clusters; finally, according to multiple pieces of the above-mentioned historical fault event data, determine the target cluster and the warning plunger pump devices. The above-mentioned target cluster is the cluster that includes the above-mentioned plunger pump devices that have had faults, and the above-mentioned warning plunger pump devices are at least the above-mentioned plunger pump devices that have not had the above-mentioned faults in the above-mentioned target cluster. In this application, by training the training data corresponding to the target data of multiple plunger pump devices, based on the training results, multiple plunger pump devices are divided into multiple clusters, and then according to multiple pieces of the above-mentioned historical fault event data, the target cluster and the warning plunger pump devices are determined. According to the collaborative filtering theory, the probability that plunger pump devices with highly similar attributes have the same type of faults is extremely high. Therefore, early warnings can be given to a large number of devices, solving the problem in the prior art that early warnings cannot be given to a large number of plunger pump devices at the same time; moreover, the above-mentioned target data includes data at each stage of the plunger pump, so the universality is relatively good.

[0037] It should be noted that the steps shown in the flowchart of the accompanying drawings can be executed in a computer system such as a set of computer executable instructions, and although the logical order is shown in the flowchart, in some cases, the steps shown or described can be executed in a different order than here.

[0038] In an embodiment of the present application, multiple pieces of the above-mentioned training data are trained, and based on the training results, it is determined to divide multiple above-mentioned plunger pump devices into multiple clusters, including: based on the K-means clustering algorithm, dividing multiple above-mentioned training data into G clustering data groups according to the grouping method corresponding to the best clustering effect. Each of the above-mentioned clustering data groups corresponds to a plunger pump device type. Each of the G groups corresponds to a best clustering center, and the G above-mentioned best clustering centers form a best clustering center data set; according to the above-mentioned index value and the above-mentioned best clustering center data set, calculate the corresponding best two-dimensional membership degree matrix, and determine the group corresponding to the maximum membership degree of the target plunger pump device according to the above-mentioned best two-dimensional membership degree matrix, and divide multiple above-mentioned plunger pump devices into multiple clusters. The plunger pump devices in the above-mentioned predetermined cluster have the maximum membership degree corresponding to the plunger pump device type corresponding to the above-mentioned predetermined cluster, and the above-mentioned predetermined cluster is any one of the above-mentioned clusters. By using the K-means clustering algorithm to group multiple above-mentioned training data according to the grouping corresponding to the best clustering effect, and each group of data corresponds to a plunger pump device type, multiple above-mentioned plunger pump devices can be accurately divided into multiple clusters, so that the early warning plunger pump devices can be determined more accurately subsequently.

[0039] In another embodiment of the present application, based on the K-means clustering algorithm, multiple above-mentioned training data are divided into G clustering data groups according to the grouping method of the best clustering effect, including: as Figure 2 shown, the grouping step: based on the K-means clustering algorithm, divide multiple above-mentioned training data into K initial data groups, and calculate K initial clustering centers. The K above-mentioned initial clustering centers form an initial clustering center data set; the membership degree matrix calculation step: calculate a two-dimensional membership degree matrix according to the above-mentioned index value and the above-mentioned initial clustering center data set; the value function calculation step: calculate a value function based on the above-mentioned index value and the above-mentioned initial clustering center data set; in the case where the above-mentioned value function is less than the threshold, stop the calculation, and determine the current K above-mentioned initial data groups as K above-mentioned clustering data groups, and G = K; in the case where the above-mentioned value function is greater than or equal to the above-mentioned threshold, update the above-mentioned initial clustering center, and sequentially repeat the above-mentioned grouping step, the above-mentioned membership degree matrix calculation step, and the above-mentioned value function calculation step until the above-mentioned value function is less than the above-mentioned threshold and stop, and determine the current updated multiple above-mentioned initial data groups as G above-mentioned clustering data groups. The principle of the above-mentioned K-means clustering algorithm is simple, easy to operate and has a very high execution efficiency, and can accurately divide the data into G clustering data groups, thus further ensuring that the early warning plunger pump devices can be determined more accurately subsequently.

[0040] In a specific embodiment of the present application, based on the K-means clustering algorithm, dividing multiple pieces of the above training data into G clustering data groups according to the grouping method with the best clustering effect mainly includes the following steps: By means of the principle of the K-means clustering algorithm, dividing multiple pieces of the above training data into K initial data groups, calculating the mean value of each group as the initial clustering center, that is, generating the initial clustering center data set C = {C1, C2, ΛC k}, where C i is the initial clustering center, i is an integer and 1 ≤ i ≤ K. Based on the feature index and the initial clustering center data set C = {C1, C2, ΛC k}, calculating the two-dimensional membership matrix U; Based on the feature index and the initial clustering center data set C = {C1, C2, ΛC k}, calculating the value function J; Setting a threshold β, according to the value function J, if J < β, then stop the algorithm; if J ≥ β, then update the clustering center and repeat the above grouping steps, the above membership matrix calculation steps, and the above value function calculation steps in sequence until J < β stops. Based on the above clustering process, finally obtain the membership matrix U under the best clustering effect, and determine its clustering result according to the proportion probability of each device in each category in U, and finally obtain G clustering data groups.

[0041] In another embodiment of the present application, as Figure 2 shown, according to multiple pieces of the above historical fault event data, determining the target cluster group and the warning plunger pump device includes: According to multiple pieces of the above historical fault event data, determining all the fault types that have occurred in all the above plunger pump devices, and generating enumeration type labels corresponding to different fault types, and labeling the above plunger pump devices that have had faults of the corresponding fault types with the above fault type labels; Determining the above cluster group including the above plunger pump devices labeled with the above fault type labels as the above target cluster group; Determining the above fault type with the most occurrences in the above target cluster group as the typical fault type; Determining the above plunger pump devices in the above target cluster group that have not had faults of the above typical fault type as the above warning plunger pump devices. The method for determining the target cluster group and the warning plunger pump device takes the fault type with the most occurrences in the target cluster group as the typical fault type, ensuring that the possibility of the determined warning plunger pump having a fault is very high, so the accuracy of the determined warning plunger pump can be further improved.

[0042] In a specific embodiment of the present application, according to multiple pieces of the above historical fault event data, determining all the fault types that have occurred in all the above plunger pump devices, and generating enumeration type labels corresponding to different fault types, obtaining the fault type label set F = {F1, F2, ΛF k}, where F is the fault type label set, f nIt is a label for a certain type of fault. And the above plunger pump equipment that has experienced faults of the corresponding fault type is labeled with the corresponding above-mentioned fault type label. Among them, according to the multiple above-mentioned historical fault event data, the common fault types of the plunger pump are shown in the following table:

[0043] Fault type label Fault type name <![CDATA[f1]]> Reducer gear fault <![CDATA[f2]]> Reducer bearing fault <![CDATA[f3]]> Crankcase fault <![CDATA[f4]]> Crosshead fault <![CDATA[f5]]> Hydraulic end pump valve fault <![CDATA[f6]]> Hydraulic end packing fault <![CDATA[f7]]> Hydraulic end plunger fault

[0044] In another embodiment of the present application, based on the above-mentioned target data, corresponding training data is determined, and multiple above-mentioned training data is obtained, including: based on the multiple above-mentioned target data, multiple feature indicators are determined, and the multiple above-mentioned feature indicators include file type feature indicators of the plunger pump equipment, operation type feature indicators of the plunger pump equipment, maintenance type feature indicators of the plunger pump equipment, and fault type feature indicators of the plunger pump equipment; based on the multiple above-mentioned target data, the above-mentioned index values corresponding to the above-mentioned feature indicators are determined to obtain the above-mentioned training data. This method obtains more comprehensive training data, which is beneficial to improving the accuracy of subsequent fault warning.

[0045] In another specific embodiment of the present application, based on the multiple above-mentioned target data, multiple plunger pump equipment feature indicators are determined, as shown in the following table:

[0046]

[0047]

[0048] Based on the above three types of data indicators, a plunger pump feature indicator system X = {X1, X2, ΛX i} is constructed, where X i represents the i-th plunger pump feature indicator.

[0049] In another embodiment of the present application, as Figure 2 shown, based on the multiple above-mentioned target data, the above-mentioned index values corresponding to the above-mentioned feature indicators are determined to obtain the above-mentioned training data, including: preprocessing the above-mentioned target data to obtain numerical data; converting the above-mentioned numerical data into data under the same dimension to obtain unified dimension data; performing dimensionality reduction processing on the above-mentioned unified dimension data to obtain the above-mentioned training data. Processing the target data makes the obtained training data have less storage space, shorter calculation time, no noise, high model performance, fewer data dimensions, and is convenient for subsequent processing.

[0050] Specifically, the above-mentioned numerical data can be converted to the same dimension through min-max standardization to reduce the interference caused by the dimension to the data.

[0051] In another embodiment of the present application, preprocessing is performed on a plurality of the above-mentioned target data to obtain numerical data, including: performing word segmentation processing and vectorization processing on the unstructured text data in the above-mentioned target data in sequence; performing discretization processing on the continuous numerical fields in the above-mentioned target data; and performing dummy coding processing on the enumerated string data.

[0052] In a specific embodiment of the present application, the above-mentioned unstructured text data mainly includes data containing unstructured text fields such as equipment maintenance reasons and equipment maintenance contents. For such data, it is necessary to perform word segmentation on the text fields through the jieba algorithm, cut the long text into several word forms, and regard punctuation marks, spaces, common words, modal particles, etc. as stop words and remove them from the word segmentation results to obtain effective vocabulary combinations. Then, the words after word segmentation are formed into a term list, and then a corresponding vector is constructed for each term list. The dimension of the vector is the same as the dimension of the term list, and the value of the vector is the number of times each term in the term list appears in the text, that is, a bag-of-words model is constructed, and then it is transformed into a weight vector based on the TF-IDF algorithm.

[0053] Specifically, the discretization processing of the continuous numerical fields in the above-mentioned target data can be equal-frequency discretization processing or equal-width discretization processing, which can reduce the time and space overhead of the algorithm and improve the clustering ability and anti-noise ability of the system for samples; for the enumerated string data, dummy coding processing is performed, and the enumerated string data values are transformed into a discrete data type format combined with 0 and 1 by means of one-hot encoding.

[0054] In still another embodiment of the present application, dimensionality reduction processing is performed on the above-mentioned unified dimension data to obtain each of the above-mentioned training data, including: through the PCA algorithm, extracting a part of the above-mentioned index values of each of the above-mentioned unified dimension data as target index values to obtain the above-mentioned training data. The PCA algorithm is a complex multi-parameter problem that can be transformed into a problem with only a few parameters through hierarchical classification. Mathematically, by projecting a high-dimensional vector into a low-dimensional space, the key information of the data is retained and reflected in a few variables in the low-dimensional space, greatly reducing the subsequent data processing difficulty.

[0055] In another specific embodiment of the present application, through the PCA algorithm, the principal components of each device attribute characteristic index in the plunger pump characteristic index system X = {X1, X2, ΛX i} are extracted to construct a new device attribute index system X' = {X'1, X'2, ΛX' i}, where X'1 represents the principal component index generated by the linear combination of X1, X2, ΛX i and j < i.

[0056] In another embodiment of the present application, after determining the target cluster and the warning plunger pump device, the above method further includes: generating and outputting a warning list corresponding to the above warning plunger pump device, as Figure 2 shown. Generating and outputting a warning list facilitates maintenance personnel to conduct inspection and troubleshooting according to the devices in the warning list and specific potential fault hazards before a fault occurs, and solve potential fault hazards.

[0057] The embodiment of the present application also provides a fault warning device for a plunger pump device. It should be noted that the fault warning device for the plunger pump device in the embodiment of the present application can be used to execute the fault warning method for the plunger pump device provided in the embodiment of the present application. The following introduces the fault warning device for the plunger pump device provided in the embodiment of the present application.

[0058] Figure 3 is a schematic diagram of a fault warning device for a plunger pump device according to an embodiment of the present application. As Figure 3 shown, the device includes:

[0059] An acquisition unit 10, configured to acquire target data of a plurality of plunger pump devices, where the target data includes historical fault event data of the plunger pump device, archive data of the plunger pump device, operation data of the plunger pump device, and maintenance data of the plunger pump device;

[0060] A first determination unit 20, configured to determine corresponding training data based on each of the above target data to obtain a plurality of the above training data, and each of the above training data includes index values corresponding to a plurality of different characteristic indexes;

[0061] A training unit 30, configured to train a plurality of the above training data and determine to divide the plurality of the above plunger pump devices into a plurality of clusters based on the training result;

[0062] A second determination unit 40, configured to determine a target cluster and a warning plunger pump device according to a plurality of the above historical fault event data, where the target cluster is the above cluster including the plunger pump devices that have had faults, and the warning plunger pump device is at least the plunger pump device that has not had the above faults in the above target cluster.

[0063] The fault warning device for the above-mentioned plunger pump equipment obtains, through the acquisition unit 10, historical fault event data, archive data, operation data, and maintenance data of multiple plunger pump equipment as target data; then, through the first determination unit 20, based on each of the above-mentioned target data, training data corresponding to multiple different characteristic indicators and their corresponding index values are determined to obtain multiple sets of the above-mentioned training data; afterwards, through the training unit 30, multiple sets of the above-mentioned training data are trained, and based on the training results, multiple plunger pump equipment are divided into multiple cluster groups; finally, through the second determination unit 40, according to multiple sets of the above-mentioned historical fault event data, a target cluster group and warning plunger pump equipment are determined. The above-mentioned target cluster group is the cluster group including the plunger pump equipment that has had a fault, and the above-mentioned warning plunger pump equipment is at least the plunger pump equipment that has not had the above-mentioned fault in the above-mentioned target cluster group. In this application, by training the training data corresponding to the target data of multiple plunger pump equipment, multiple plunger pump equipment are divided into multiple cluster groups based on the training results, and then according to multiple sets of the above-mentioned historical fault event data, a target cluster group and warning plunger pump equipment are determined. According to the collaborative filtering theory, the probability of plunger pump equipment with highly similar attributes having the same type of fault is extremely high. Therefore, early warnings can be given to a large number of equipment, solving the problem in the prior art that early warnings cannot be given to a large number of plunger pump equipment simultaneously; moreover, the above-mentioned target data includes data at each stage of the plunger pump, so the universality is relatively good.

[0064] In an embodiment of the present application, the above-mentioned training unit includes a processing subunit and a first determination subunit. Among them, the above-mentioned processing subunit is used to divide multiple sets of the above-mentioned training data into G clustering data groups according to the grouping method corresponding to the best clustering effect based on the K-means clustering algorithm. Each of the above-mentioned clustering data groups corresponds to a type of plunger pump equipment. Each of the G groups corresponds to a best clustering center, and the G above-mentioned best clustering centers form a best clustering center data set; the above-mentioned first determination subunit is used to calculate the corresponding best two-dimensional membership matrix according to the above-mentioned index values and the above-mentioned best clustering center data set, and determine the group corresponding to the maximum membership of the target plunger pump equipment according to the above-mentioned best two-dimensional membership matrix, and divide multiple plunger pump equipment into multiple cluster groups. The plunger pump equipment in the above-mentioned predetermined cluster group has the largest membership corresponding to the type of plunger pump equipment corresponding to the above-mentioned predetermined cluster group. The above-mentioned predetermined cluster group is any one of the above-mentioned cluster groups. By using the K-means clustering algorithm to group multiple sets of the above-mentioned training data according to the grouping method corresponding to the best clustering effect, and each group of data corresponds to a type of plunger pump equipment, multiple plunger pump equipment can be accurately divided into multiple cluster groups, so that the warning plunger pump equipment can be determined more accurately subsequently.

[0065] In another embodiment of the present application, the above processing subunit includes a grouping module, a first calculation module, a second calculation module, a first determination module, and a second determination module. Among them, the grouping module is used to perform the grouping step. Based on the K-means clustering algorithm, multiple pieces of the above training data are divided into K initial data groups, and K initial cluster centers are calculated. The K initial cluster centers form an initial cluster center data set. The first calculation module is used to perform the membership matrix calculation step. According to the above index values and the initial cluster center data set, a two-dimensional membership matrix is calculated. The second calculation module is used to perform the value function calculation step. Based on the above index values and the initial cluster center data set, a value function is calculated. The first determination module is used to stop the calculation when the value function is less than the threshold, and determine the current K initial data groups as the K cluster data groups, and G = K. The second determination module is used to update the initial cluster center when the value function is greater than or equal to the threshold, and sequentially repeat the above grouping step, the membership matrix calculation step, and the value function calculation step until the value function is less than the threshold and stops, and determine the current updated multiple initial data groups as G cluster data groups. The principle of the K-means clustering algorithm is simple, easy to operate, and has a very high execution efficiency. It can accurately divide the data into G cluster data groups, thus further ensuring that the early warning plunger pump equipment can be determined more accurately subsequently.

[0066] In a specific embodiment of the present application, based on the K-means clustering algorithm, dividing multiple pieces of the above training data into G cluster data groups according to the grouping method with the best clustering effect mainly includes the following steps: With the help of the principle of the K-means clustering algorithm, multiple pieces of the above training data are divided into K initial data groups, and the mean value of each group is calculated as the initial cluster center, that is, an initial cluster center data set C = {C1, C2, ΛC k} is generated, where C i is the initial cluster center, i is an integer and 1 ≤ i ≤ K. Based on the characteristic index and the initial cluster center data set C = {C1, C2, ΛC k}, a two-dimensional membership matrix U is calculated; based on the characteristic index and the initial cluster center data set C = {C1, C2, ΛC k}, a value function J is calculated; a threshold β is set. According to the value function J, if J < β, the algorithm stops; if J ≥ β, the cluster center is updated and the above grouping step, the membership matrix calculation step, and the value function calculation step are sequentially repeated until J < β stops. Based on the above clustering process, the membership matrix U under the best clustering effect is finally obtained. According to the probability of each device in each category in U, its clustering result is determined, and finally G cluster data groups are obtained.

[0067] In another embodiment of the present application, the second determination unit includes a second determination subunit, a third determination subunit, a fourth determination subunit, and a fifth determination subunit. Among them, the second determination subunit is used to determine all the fault types that have occurred in all the plunger pump devices according to the multiple historical fault event data, generate enumeration tags corresponding to different fault types, and label the corresponding fault type tags for the plunger pump devices that have had faults of the corresponding fault types; the third determination subunit is used to determine the cluster group including the plunger pump devices labeled with the fault type tags as the target cluster group; the fourth determination subunit is used to determine the fault type with the most occurrences in the target cluster group as the typical fault type; the fifth determination subunit is used to determine the plunger pump devices in the target cluster group that have not had faults of the typical fault type as the early warning plunger pump devices. The method for determining the target cluster group and the early warning plunger pump devices takes the fault type with the most occurrences in the target cluster group as the typical fault type, ensuring that the probability that the determined early warning plunger pump may have a fault is very high. Therefore, the accuracy of the determined early warning plunger pump can be further improved.

[0068] In a specific embodiment of the present application, according to the multiple historical fault event data, all the fault types that have occurred in all the plunger pump devices are determined, and enumeration tags corresponding to different fault types are generated to obtain a fault type tag set F = {F1, F2, ΛF k}, where F is the fault type tag set and f n is a certain type of fault type tag. And the corresponding fault type tags are labeled for the plunger pump devices that have had faults of the corresponding fault types. Among them, according to the multiple historical fault event data, the common fault types of the plunger pump are shown in the following table:

[0069]

[0070]

[0071] In yet another embodiment of the present application, the first determination unit includes a sixth determination subunit and a seventh determination subunit. Among them, the sixth determination subunit is used to determine multiple characteristic indicators based on the multiple target data. The multiple characteristic indicators include the file type characteristic indicators of the plunger pump device, the operation type characteristic indicators of the plunger pump device, the maintenance type characteristic indicators of the plunger pump device, and the fault type characteristic indicators of the plunger pump device; the seventh determination subunit is used to determine the corresponding indicator values of each of the characteristic indicators based on the multiple target data to obtain each of the training data. This method obtains more comprehensive training data, which is beneficial to improving the accuracy of subsequent fault early warning.

[0072] In yet another specific embodiment of the present application, based on the above-mentioned multiple target data, multiple characteristic indexes of the piston pump device are determined as shown in the following table:

[0073]

[0074]

[0075] Based on the above three types of data indexes, a piston pump characteristic index system X = {X1, X2, ΛX i} is constructed, where X i represents the i-th type of piston pump characteristic index.

[0076] In another embodiment of the present application, the above-mentioned seventh determination subunit includes a preprocessing module, a conversion module, and a dimensionality reduction module. Among them, the preprocessing module is used to preprocess each of the above-mentioned target data to obtain numerical data; the conversion module is used to convert the above-mentioned numerical data into data under the same dimension to obtain unified dimension data; the dimensionality reduction module is used to perform dimensionality reduction processing on the above-mentioned unified dimension data to obtain each of the above-mentioned training data. Processing the target data can make the obtained training data have less storage space, shorter calculation time, no noise, high model performance, and fewer data dimensions, which is convenient for subsequent processing.

[0077] Specifically, the above-mentioned numerical data can be converted to the same dimension through min-max normalization to reduce the interference caused by the dimension to the data.

[0078] In yet another embodiment of the present application, the preprocessing module includes a first processing sub-module, a second processing sub-module, and a third processing sub-module. Among them, the first processing sub-module is used to perform word segmentation processing and vectorization processing on the unstructured text data in the above-mentioned target data in sequence; the second processing sub-module is used to perform discretization processing on the continuous numerical fields in the above-mentioned target data; the third processing sub-module is used to perform dummy coding processing on the enumerated string data.

[0079] In a specific embodiment of the present application, the above-mentioned unstructured text data mainly includes data containing unstructured text fields such as equipment maintenance reasons and equipment maintenance contents. For such data, the jieba algorithm needs to be used to segment the text fields, cut the long text into several word forms, and regard punctuation marks, spaces, common words, modal particles, etc. as stop words and remove them from the word segmentation results to obtain an effective vocabulary combination. Then, the words after word segmentation are combined into a list of terms, and then a corresponding vector is constructed for each list of terms. The dimension of the vector is the same as the dimension of the list of terms, and the value of the vector is the number of times each term in the list of terms appears in the text, that is, a bag-of-words model is constructed, and then it is converted into a weight vector based on the TF-IDF algorithm.

[0080] Specifically, the discretization process for the continuous numerical fields in the above target data can be equal-frequency discretization or equal-width discretization, which can reduce the time and space overhead of the algorithm and improve the clustering ability and anti-noise ability of the system for samples; for the enumerated string data, dummy coding is performed, and the enumerated string data values are converted into a discrete data type format of a combination of 0 and 1 with the help of one-hot encoding.

[0081] In another embodiment of the present application, the above-mentioned dimensionality reduction module includes an extraction sub-module, and the extraction sub-module is used to extract a part of the above-mentioned index values of each of the above-mentioned unified dimension data as target index values through the PCA algorithm to obtain the above-mentioned training data. The PCA algorithm is a complex multi-parameter problem that can be transformed into a problem with only a few parameters through hierarchical classification. Mathematically, by projecting a high-dimensional vector into a low-dimensional space, the key information of the data is retained and reflected in a few variables in the low-dimensional space, greatly reducing the processing difficulty of subsequent data.

[0082] In another specific embodiment of the present application, the dimensionality reduction module extracts the principal components of each device attribute characteristic index in the plunger pump characteristic index system X = {X1, X2, ΛX i} through the PCA algorithm to construct a new device attribute index system X' = {X'1, X'2, ΛX' i}, where X'1 represents the principal component index generated by the linear combination of X1, X2, ΛX i and j < i.

[0083] In another embodiment of the present application, the above-mentioned device further includes an output unit, and the output unit is used to generate and output a warning list corresponding to the above-mentioned warning plunger pump device after determining the target cluster group and warning the plunger pump device. Generating and outputting a warning list facilitates maintenance personnel to conduct patrol inspections and troubleshoot according to the equipment and specific potential faults in the warning list before a failure occurs, and solve potential faults.

[0084] The above-mentioned fault warning device for plunger pump equipment includes a processor and a memory. The above-mentioned acquisition unit, first determination unit, training unit, and second determination unit are all stored in the memory as program units, and the processor executes the above-mentioned program units stored in the memory to implement corresponding functions.

[0085] The processor contains a kernel, and the kernel retrieves the corresponding program units from the memory. One or more kernels can be set, and by adjusting the kernel parameters, the problem in the prior art that it is impossible to simultaneously warn a large number of plunger pump devices can be solved.

[0086] The memory may include non - permanent memory in the form of computer - readable media, such as random access memory (RAM) and / or non - volatile memory, such as read - only memory (ROM) or flash RAM, and the memory includes at least one storage chip.

[0087] An embodiment of the present invention provides a computer - readable storage medium, on which a program is stored, and when the program is executed by a processor, the fault warning method of the above - mentioned plunger pump device is implemented.

[0088] An embodiment of the present invention provides a processor, which is used to run a program. When the program runs, the fault warning method of the above - mentioned plunger pump device is executed.

[0089] An embodiment of the present invention provides a device, which includes a processor, a memory, and a program stored on the memory and executable on the processor. When the processor executes the program, at least the following steps are implemented:

[0090] Step S101: Obtain the target data of multiple plunger pump devices. The target data includes the historical fault event data, the file data, the operation data, and the maintenance data of the plunger pump device.

[0091] Step S102: Based on each of the above - mentioned target data, determine the corresponding training data to obtain multiple pieces of the above - mentioned training data. Each of the above - mentioned training data includes the index values corresponding to multiple different characteristic indexes.

[0092] Step S103: Train multiple pieces of the above - mentioned training data, and based on the training results, determine to divide multiple plunger pump devices into multiple cluster groups.

[0093] Step S104: According to multiple pieces of the above - mentioned historical fault event data, determine the target cluster group and the warning plunger pump devices. The target cluster group is the cluster group including the plunger pump devices that have had faults, and the warning plunger pump devices are at least the plunger pump devices that have not had the above - mentioned faults in the above - mentioned target cluster group.

[0094] The device in this article can be a server, a PC, a PAD, a mobile phone, etc.

[0095] The present application also provides a computer program product, which is suitable for executing a program initialized with at least the following method steps when executed on a data - processing device:

[0096] Step S101: Obtain the target data of multiple plunger pump devices. The target data includes the historical fault event data, the file data, the operation data, and the maintenance data of the plunger pump device.

[0097] Step S102: Based on each of the above target data, determine the corresponding training data to obtain multiple pieces of the above training data. Each of the above training data includes the index values corresponding to multiple different characteristic indices.

[0098] Step S103: Train multiple pieces of the above training data, and based on the training results, determine to divide multiple above plunger pump devices into multiple clusters.

[0099] Step S104: According to multiple pieces of the above historical fault event data, determine the target cluster and the warning plunger pump devices. The above target cluster is the cluster that includes the above plunger pump devices that have had faults, and the above warning plunger pump devices are at least the above plunger pump devices that have not had the above faults in the above target cluster.

[0100] According to another aspect of the embodiments of the present invention, there is also provided a plunger pump device system, including a plunger pump device, one or more processors, a memory, and one or more programs. Among them, the above one or more programs are stored in the above memory and are configured to be executed by the above one or more processors. The above one or more programs include those for executing any one of the above methods.

[0101] The above plunger pump device system trains the training data corresponding to the target data of multiple plunger pump devices, divides multiple plunger pump devices into multiple clusters based on the training results, and then determines the target cluster and the warning plunger pump devices according to multiple pieces of the above historical fault event data. According to the collaborative filtering theory, the probability that plunger pump devices with highly similar attributes have the same type of faults is extremely high. Therefore, it is possible to give warnings to a large number of devices, solving the problem in the prior art that it is impossible to give warnings to a large number of plunger pump devices simultaneously. And, the above target data includes data at each stage of the plunger pump, so the universality is relatively good.

[0102] In the above embodiments of the present invention, the descriptions of each embodiment have their own focuses. For parts not detailed in a certain embodiment, reference can be made to the relevant descriptions of other embodiments.

[0103] In several embodiments provided in the present application, it should be understood that the disclosed technical content can be implemented in other ways. Among them, the device embodiments described above are only illustrative. For example, the above unit division can be a logical function division. In actual implementation, there can be other division methods. For example, multiple units or components can be combined or integrated into another system, or some features can be ignored or not executed. Another point is that the couplings or direct couplings or communication connections shown or discussed with each other can be through some interfaces. The indirect couplings or communication connections of units or modules can be electrical or other forms.

[0104] The units described above as separate components may or may not be physically separated. The components shown as units may or may not be physical units, that is, they may be located in one place or may be distributed over multiple units. Some or all of the units can be selected according to actual needs to achieve the purpose of the solution of this embodiment.

[0105] In addition, each functional unit in various embodiments of the present invention may be integrated in a processing unit, may exist separately as individual physical units, or two or more units may be integrated in one unit. The above-mentioned integrated units can be implemented in the form of hardware or in the form of software functional units.

[0106] If the above-mentioned integrated unit is implemented in the form of a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on such an understanding, the technical solution of the present invention, in essence, or the part that contributes to the prior art, or all or part of the technical solution, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions for causing a computer device (which may be a personal computer, a server, or a network device, etc.) to execute all or part of the steps of the above-mentioned methods in various embodiments of the present invention. The foregoing storage medium includes: various media such as USB flash drives, read-only memories (ROMs), random access memories (RAMs), mobile hard disks, magnetic disks, or optical discs that can store program codes.

[0107] From the above description, it can be seen that the above embodiments of the present application achieve the following technical effects:

[0108] 1) The fault warning method for the plunger pump equipment of the present application includes: First, obtain the historical fault event data of multiple plunger pump equipment, the file data of the plunger pump equipment, the operation data of the plunger pump equipment, and the maintenance data of the plunger pump equipment as target data; Then, based on each of the above target data, determine the corresponding training data including the index values corresponding to multiple different characteristic indexes, and obtain multiple pieces of the above training data; After that, train multiple pieces of the above training data, and based on the training results, determine to divide multiple plunger pump equipment into multiple cluster groups; Finally, according to multiple pieces of the above historical fault event data, determine the target cluster group and the warning plunger pump equipment. The above target cluster group is the cluster group including the above plunger pump equipment that has had a fault, and the above warning plunger pump equipment is at least the above plunger pump equipment that has not had the above fault in the above target cluster group. By training the training data corresponding to the target data of multiple plunger pump equipment, dividing multiple plunger pump equipment into multiple cluster groups based on the training results, and then determining the target cluster group and the warning plunger pump equipment according to multiple pieces of the above historical fault event data, according to the collaborative filtering theory, the probability of plunger pump equipment with highly similar attributes having the same type of fault is extremely high, so it is possible to warn a large number of equipment, solving the problem in the prior art that it is impossible to warn a large number of plunger pump equipment simultaneously; Moreover, the above target data includes data in each stage of the plunger pump, so the universality is better.

[0109] 2) The fault warning device for the plunger pump equipment of the present application obtains the historical fault event data of multiple plunger pump equipment, the file data of the plunger pump equipment, the operation data of the plunger pump equipment, and the maintenance data of the plunger pump equipment as target data through an acquisition unit; Then, through a first determination unit, based on each of the above target data, determine the corresponding training data including the index values corresponding to multiple different characteristic indexes, and obtain multiple pieces of the above training data; After that, through a training unit, train multiple pieces of the above training data, and based on the training results, determine to divide multiple plunger pump equipment into multiple cluster groups; Finally, through a second determination unit, according to multiple pieces of the above historical fault event data, determine the target cluster group and the warning plunger pump equipment. The above target cluster group is the cluster group including the above plunger pump equipment that has had a fault, and the above warning plunger pump equipment is at least the above plunger pump equipment that has not had the above fault in the above target cluster group. By training the training data corresponding to the target data of multiple plunger pump equipment, dividing multiple plunger pump equipment into multiple cluster groups based on the training results, and then determining the target cluster group and the warning plunger pump equipment according to multiple pieces of the above historical fault event data, according to the collaborative filtering theory, the probability of plunger pump equipment with highly similar attributes having the same type of fault is extremely high, so it is possible to warn a large number of equipment, solving the problem in the prior art that it is impossible to warn a large number of plunger pump equipment simultaneously; Moreover, the above target data includes data in each stage of the plunger pump, so the universality is better.

[0110] 3) For the plunger pump equipment system of the present application, by training the training data corresponding to the target data of multiple plunger pump equipment, dividing the multiple plunger pump equipment into multiple clusters based on the training results, and then determining the target cluster and the warning plunger pump equipment according to the multiple historical fault event data above. According to the collaborative filtering theory, the probability of plunger pump equipment with highly similar attributes having the same type of fault is extremely high. Therefore, a large number of equipment can be warned, solving the problem in the prior art that it is impossible to warn a large number of plunger pump equipment simultaneously; moreover, the above target data includes data at each stage of the plunger pump, so the universality is relatively good.

[0111] The above are only the preferred embodiments of the present application and are not used to limit the present application. For those skilled in the art, various changes and modifications can be made to the present application. Any modification, equivalent replacement, improvement, etc. made within the spirit and principle of the present application shall be included within the protection scope of the present application.

Claims

1. A fault warning method for a plunger pump device, characterized in that, Including: Obtain the target data of multiple plunger pump devices, where the target data includes the historical fault event data of the plunger pump devices, the file data of the plunger pump devices, the operation data of the plunger pump devices, and the maintenance data of the plunger pump devices; Based on each of the target data, determine the corresponding training data, obtaining multiple pieces of the training data, and each piece of the training data includes the index values corresponding to multiple different characteristic indexes; Train multiple pieces of the training data, and based on the training results, determine that multiple plunger pump devices are divided into multiple cluster groups; According to multiple pieces of the historical fault event data, determine the target cluster group and the warning plunger pump devices. The target cluster group is the cluster group including the plunger pump devices that have had faults, and the warning plunger pump devices are at least the plunger pump devices that have not had the faults in the target cluster group.

2. The method according to claim 1, wherein Training multiple pieces of the training data and, based on the training results, determining that multiple plunger pump devices are divided into multiple cluster groups includes: Based on the K-means clustering algorithm, divide multiple pieces of the training data into G clustering data groups according to the grouping method corresponding to the best clustering effect. Each clustering data group corresponds to a type of plunger pump device. Each of the G groups corresponds to a best clustering center, and the G best clustering centers form a best clustering center data set; According to the index values and the best clustering center data set, calculate the corresponding best two-dimensional membership degree matrix, and according to the best two-dimensional membership degree matrix, determine the group corresponding to the maximum membership degree of the target plunger pump device, and divide multiple plunger pump devices into multiple cluster groups. The plunger pump devices in a predetermined cluster group have the maximum membership degree corresponding to the type of plunger pump device corresponding to the predetermined cluster group, and the predetermined cluster group is any one of the cluster groups.

3. The method according to claim 2, wherein Based on the K-means clustering algorithm, dividing multiple pieces of the training data into G clustering data groups according to the grouping method of the best clustering effect includes: Grouping step: Based on the K-means clustering algorithm, divide multiple pieces of the training data into K initial data groups, and calculate K initial clustering centers. The K initial clustering centers form an initial clustering center data set; Membership degree matrix calculation step: According to the index values and the initial clustering center data set, calculate the two-dimensional membership degree matrix; Value function calculation step: Based on the index values and the initial clustering center data set, calculate the value function; In the case where the value function is less than the threshold, stop the calculation, and determine the current K initial data groups as K clustering data groups, and G = K; In the case where the value function is greater than or equal to the threshold, update the initial clustering center, and sequentially repeat the grouping step, the membership degree matrix calculation step, and the value function calculation step until the value function is less than the threshold and stop, and determine the current updated multiple initial data groups as G clustering data groups.

4. The method according to claim 1, wherein According to multiple pieces of the historical fault event data, determining the target cluster group and the warning plunger pump devices includes: Based on the multiple historical fault event data, determine all the fault types that have occurred in all the plunger pump devices, generate enumerated tags corresponding to different fault types, and label the corresponding fault type tags for the plunger pump devices that have experienced faults of the corresponding fault types; Determine the cluster group including the plunger pump devices labeled with the fault type tags as the target cluster group; Determine the fault type with the most occurrences in the target cluster group as the typical fault type; Determine the plunger pump devices in the target cluster group that have not experienced faults of the typical fault type as the warning plunger pump devices.

5. The method according to any one of claims 1 to 4, characterized in that, Based on each of the target data, determine the corresponding training data to obtain multiple pieces of training data, including: Based on multiple pieces of target data, determine multiple characteristic indicators, where the multiple characteristic indicators include the file - type characteristic indicators of the plunger pump device, the operation - type characteristic indicators of the plunger pump device, the maintenance - type characteristic indicators of the plunger pump device, and the fault - type characteristic indicators of the plunger pump device; Based on multiple pieces of target data, determine the indicator values corresponding to each of the characteristic indicators to obtain each piece of training data.

6. The method according to claim 5, characterized in that Based on multiple pieces of target data, determine the indicator values corresponding to each of the characteristic indicators to obtain each piece of training data, including: Pre - process each piece of target data to obtain numericalized data; Convert the numericalized data into data under the same dimension to obtain unified - dimension data; Perform dimensionality reduction processing on the unified - dimension data to obtain each piece of training data.

7. The method according to claim 6, wherein Pre - process multiple pieces of target data to obtain numericalized data, including: Perform word segmentation processing and vectorization processing on the unstructured text data in the target data in sequence; Perform discretization processing on the continuous numerical fields in the target data; Perform dummy coding processing on the enumerated string data.

8. The method according to claim 6, characterized in that, Perform dimensionality reduction processing on the unified - dimension data to obtain each piece of training data, including: Through the PCA algorithm, extract some of the multiple indicator values of each piece of unified - dimension data as the target indicator values to obtain the training data.

9. The method according to any one of claims 1 to 4, characterized in that After determining the target cluster group and the warning plunger pump devices, the method further includes: Generate and output a warning list corresponding to the warning plunger pump devices.

10. A fault warning device for a plunger pump device, characterized in that, Including: An acquisition unit for acquiring the target data of multiple plunger pump devices, where the target data includes the historical fault event data of the plunger pump device, the file data of the plunger pump device, the operation data of the plunger pump device, and the maintenance data of the plunger pump device; A first determination unit for determining the corresponding training data based on each of the target data to obtain multiple pieces of training data, and each piece of training data includes the indicator values corresponding to multiple different characteristic indicators; A training unit for training multiple pieces of training data and determining to divide the multiple plunger pump devices into multiple cluster groups based on the training results; A second determination unit determines a target cluster group and a warning plunger pump device according to the multiple pieces of historical fault event data, where the target cluster group is the cluster group including the plunger pump devices that have had faults, and the warning plunger pump device is at least the plunger pump device that has not had the fault in the target cluster group.

11. A computer-readable storage medium, characterized in that, The computer-readable storage medium includes a stored program, where the program executes the method according to any one of claims 1 to 9.

12. A processor, characterized in that, The processor is used to run a program, where when the program runs, it executes the method according to any one of claims 1 to 9.

13. A plunger pump equipment system, characterized in that, Comprising: A plunger pump device, one or more processors, a memory, and one or more programs, where the one or more programs are stored in the memory and are configured to be executed by the one or more processors, and the one or more programs include a method for executing any one of claims 1 to 9.

Images