Intelligent detection system for oil pumping unit failure

Abstract

The invention discloses an intelligent detection system for oil pump faults, and relates to the technical field of oil pump detection. In this scheme, a first fault detection unit and a second fault detection unit are established; the first fault detection unit outputs a first fault detection result, and the implementation principle is to convert the complex analysis and processing of the oil pump indicator diagram into the research of the oil pump power curve, which not only ensures the timeliness of fault detection, but also can output the detection results of a single or multiple faults; the second fault detection unit outputs a second fault detection result, and the implementation principle is to use the fault history data of the oil pump, based on the differences in the changes of the working parameters of the oil pump under different faults; the first fault detection result and the second fault detection result are processed by the fault result output unit to obtain the final fault detection result, and the fault detection results of the two fault detection units are used to verify and supplement each other, thereby improving the accuracy of the whole oil pump fault detection system for the oil pump fault detection.

Description

Technical Field

[0001] The invention relates to the technical field of oil pumping unit detection, and in particular to an intelligent detection system for oil pumping unit faults. Background Art

[0002] As an important part of many equipments in oil extraction, the oil pump plays a vital role in the entire oil extraction operation. At the same time, the oil pump equipment fully reflects the characteristics of simple operation and convenience, which can not only effectively increase the oil extraction volume, but also bring certain conveniences to the maintenance of the equipment. However, oil extraction is a high-intensity and long-term operation. During the uninterrupted operation of the oil pump, it is easy to be affected by many factors, and then different fault problems will occur. If effective diagnosis and troubleshooting work is not carried out in time, it will not only affect the normal operation of the oil pumping operation, but also greatly hinder the smooth progress of the entire oil extraction operation.

[0003] According to the current situation, the detection of common faults of oil pumps mainly relies on the rich experience of technicians to analyze the operation of the oil pumps by drawing indicator diagrams, so as to confirm whether there is a fault problem with the oil pumps. However, this method requires high professional quality of technicians, and human factors are also an important factor affecting the diagnosis results. At the same time, since the uninterrupted operation of the oil pump is easily affected by many factors, that is, a single fault may be caused by multiple reasons, it will be more difficult for technicians to analyze such faults, and it will also take a lot of time, which ultimately leads to the conclusions drawn from the analysis cannot accurately reflect the fault problem of the oil pump, and the timeliness of the fault detection of the oil pump is also insufficient. Summary of the invention

[0004] In order to solve the above-mentioned deficiencies in the prior art, the present invention provides an intelligent detection system for oil pumping unit failures, the system comprising:

[0005] Standard unit: obtain the standard power diagram of the pumping unit and process it to obtain the standard power curve diagram;

[0006] The first monitoring unit is used to obtain the power diagram of the oil pumping unit in real time and process it to obtain a first power curve diagram, and put all the first power curve diagrams into a first set;

[0007] First data unit: obtaining multiple preset fault types, each preset fault type includes several sub-faults, each sub-fault includes a corresponding power diagram, processing the power diagram included in each sub-fault to obtain a corresponding second power curve diagram, and putting all the second power curve diagrams and corresponding sub-faults into a second set;

[0008] Second data unit: obtain all second power curve graphs in the first data unit and process them to obtain processing results, obtain a third power curve graph under the common influence of multiple sub-faults based on the processing results, record multiple sub-faults corresponding to each third power curve graph as a fault set, establish a first mapping relationship between the third power curve graph and the fault set, and put all third power curve graphs and corresponding fault sets into a third set;

[0009] A first fault detection unit: inputs the standard power curve, the first set, the second set and the third set into a first fault detection model, trains the first fault detection model, and outputs a first fault detection result of the oil pumping unit based on the trained first fault detection model;

[0010] The second monitoring unit: acquires the operating parameters of the oil pumping unit in real time to obtain a first acquisition result, obtains first change information of each operating parameter based on the first acquisition result, and puts all the first change information into a fourth set;

[0011] The third data unit: obtains fault history data of the oil pumping unit, obtains second change information of all operating parameters under each fault type based on the fault history data, establishes a second mapping relationship between each fault type and the second change information of the corresponding operating parameters, and puts all fault types and the second change information of the corresponding operating parameters into a fifth set;

[0012] A second fault detection unit: inputting the fourth set and the fifth set into a second fault detection model, training the second fault detection model, and outputting a second fault detection result of the oil pumping unit based on the trained second fault detection model;

[0013] Fault result output unit: outputs the pumping unit fault detection result after processing the first fault detection result and the second fault detection result.

[0014] The present invention is implemented by the following technical scheme: In this scheme, two fault detection units are provided, namely a first fault detection unit and a second fault detection unit, the first fault detection unit outputs a first fault detection result, the second fault detection unit outputs a second fault detection result, and finally the first fault detection result and the second fault detection result are processed by the fault result output unit to output a final fault detection result;

[0015] The first fault detection unit implements fault detection of the oil pumping unit, including: firstly, obtaining the standard indicator diagram of the oil pumping unit through the standard unit and processing it to obtain the standard power curve diagram of the oil pumping unit; obtaining the indicator diagram of the oil pumping unit in real time through the first monitoring unit, and processing the indicator diagram of the oil pumping unit to obtain the first power curve diagram, and putting all the first power curve diagrams into the first set; obtaining multiple preset fault types through the first data unit, such as oil well sand production fault, oil well wax deposition fault and gas influence fault, and each preset fault type corresponds to multiple specific sub-faults, such as oil well sand production fault includes piston sand resistance, fixed valve stuck and fixed valve stuck in valve There are multiple sub-faults such as the cover, and each sub-fault has a corresponding power diagram, and the power diagram of each sub-fault is processed to obtain the corresponding second power curve diagram; all the second power curve diagrams are obtained and processed through the second data unit to obtain a third power curve diagram under the common influence of multiple sub-faults, and the multiple sub-faults are used as a fault set of a single third power curve diagram, and a first mapping relationship between the fault set and the third power curve diagram is established; finally, the standard power curve diagram, the first set, the second set and the third set are input into the first fault detection model and trained, and the trained first fault detection model is used to output the first fault detection result.

[0016] The second fault detection unit realizes fault detection of the oil pump, including: obtaining the operating parameters of the oil pump in real time through the second monitoring unit, and obtaining the first change information of each operating parameter, and placing all the first change information into a fourth set; obtaining the fault history data of the oil pump through the third data unit, and based on the fault history data of the oil pump, obtaining the second change information of each operating parameter of the oil pump under different faults, and establishing a second mapping relationship between each fault type and the second change information of the corresponding operating parameter; finally, inputting the fourth set and the fifth set into the second fault detection model and training it, and using the trained second fault detection model to output the second fault detection result.

[0017] The first fault detection unit in the present scheme converts the complex analysis and processing of the oil pumping unit indicator diagram into the research of the oil pumping unit power curve, and not only establishes the second power curve diagram of the oil pumping unit corresponding to a single fault, but also establishes the third power curve diagram of the oil pumping unit corresponding to the influence of multiple faults. Then, the first power curve diagram of the oil pumping unit is obtained by using the oil pumping unit indicator diagram obtained in real time. Finally, the standard power curve diagram, the first power curve diagram, the second power curve diagram and the third power curve diagram are processed by the first fault detection unit to output the first fault detection result, which not only ensures the timeliness of the oil pumping unit fault detection, but also can output the detection results of a single fault or multiple faults, greatly improving the accuracy of the oil pumping unit fault detection.

[0018] The second fault detection unit in this scheme uses the fault history data of the oil pump. Based on the premise that there are differences in the changes in the operating parameters of the oil pump under different faults, it obtains the change information of each operating parameter of the oil pump in real time, and outputs the second fault detection result through the second fault detection model, thereby improving the accuracy of the oil pump fault detection.

[0019] As an optional technical solution, real-time acquisition of the pumping unit's power diagram and processing to obtain a first power curve diagram include:

[0020] The power diagram of the pumping unit has the stroke of the pumping unit rod as the abscissa, the load on the pumping unit rod as the ordinate, and the area of ​​the curve enclosed region is the power of the pumping unit rod;

[0021] In the dynamometer diagram of the oil pumping unit, the stroke of the oil pumping unit rod is divided into a number of equal intervals, and the area of ​​the corresponding curve closed region in each interval is calculated to obtain a calculation result;

[0022] Based on the calculation results, the power of the pumping rod corresponding to each interval is obtained;

[0023] Use a preset number series to numerically name each interval;

[0024] The first power curve is constructed by taking all the intervals named with numerical values ​​as the horizontal coordinate and the power of the pumping unit rod corresponding to each interval as the vertical coordinate.

[0025] As an optional technical solution, processing the power diagram included in each sub-fault to obtain the corresponding second power curve diagram includes:

[0026] In the dynamometer diagram included in the sub-fault, the stroke of the pumping unit rod is divided into a number of equal intervals, and the area of ​​the corresponding curve closed region in each interval is calculated to obtain a calculation result;

[0027] Based on the calculation results, the power of the pumping rod corresponding to each interval is obtained;

[0028] Use a preset number series to numerically name each interval;

[0029] The second power curve is constructed by taking all the intervals named with numerical values ​​as the horizontal coordinates and the power of the pumping unit rod corresponding to each interval as the vertical coordinate.

[0030] As an optional technical solution, each preset fault type includes several sub-faults, specifically:

[0031]

[0032]

[0033] …

[0034]

[0035] , ,……and Corresponding to different preset fault types, , ,……and Corresponding to the sub-faults under different preset fault types, , ,……and The number of sub-faults corresponding to different preset fault types.

[0036] As an optional technical solution, obtaining a third power curve diagram under the common influence of multiple sub-faults includes:

[0037] The number of multiple sub-faults is preset to R, 2≤R≤m;

[0038] Randomly determine R fault types from a plurality of preset fault types;

[0039] Among the determined R fault types, a single sub-fault in each fault type and a corresponding second power curve graph are randomly obtained to obtain a second acquisition result;

[0040] Based on the second acquisition result shown, second power curve graphs corresponding to R sub-faults are acquired and put into a preset set;

[0041] Draw an average power curve graph of all the second power curve graphs in the preset set, wherein the average power curve graph uses the abscissa of any second power curve graph as the abscissa and uses the average power of the corresponding pumping rods in all the second power curve graphs as the ordinate;

[0042] The average power curve graph that is drawn is the third power curve graph.

[0043] As an optional technical solution, the following formula is used to calculate the average power of the corresponding pumping rods in all the second power curves:

[0044]

[0045] is the average power of the pumping rod corresponding to the first horizontal coordinate in all the second power curves, In each second power curve diagram, the first abscissa corresponds to the power of the pumping rod.

[0046] As an optional technical solution, based on the trained first fault detection model, outputting the first fault detection result of the oil pump includes:

[0047] Calculating a first correlation between the first power curve graph and the standard power curve graph in the first set;

[0048] Determine whether the first correlation is greater than a first preset threshold value, if not, record the first power curve graph corresponding to the first correlation being greater than the first preset threshold value as a first power curve label graph, calculate the second correlation between the first power curve label graph and all second power curve graphs in the second set, and if so, continue to calculate the first correlation between the first power curve graph in the first set and the standard power curve graph;

[0049] Determine whether the second correlation exists and whether it is greater than the second preset threshold value. If so, obtain the corresponding second power curve diagram and sub-fault and output them as the first fault detection result; if not, calculate the third correlation between the first power curve mark diagram and all third power curve diagrams in the third set, obtain the third power curve diagram and fault set corresponding to the third correlation being greater than the third preset threshold value, and output them as the first fault detection result.

[0050] As an optional technical solution, based on the trained second fault detection model, outputting the second fault detection result of the oil pump includes:

[0051] Obtaining first change information of each operating condition parameter in the fourth set;

[0052] Traversing the second change information of each operating condition parameter in the fifth set, and filtering to obtain the second change information corresponding to the first change information;

[0053] Based on the second mapping relationship and the filtered second change information, the second fault detection result is output.

[0054] As an optional technical solution, processing the first fault detection result and the second fault detection result includes:

[0055] Determine whether the first fault detection result is consistent with the second fault detection result. If so, output the first fault detection result or the second fault detection result as the pumping unit fault detection result; if not, output the same fault detection result of the first fault detection result and the second fault detection result, and mark and troubleshoot the different fault detection results of the first fault detection result and the second fault detection result.

[0056] As an optional technical solution, all preset fault types include oil well sand production fault, oil well wax deposition fault, gas impact fault, leakage fault, heavy oil fault, liquid supply fault, oil well pumping and spraying fault and oil pumping equipment fault.

[0057] One or more technical solutions provided by the present invention have at least the following technical effects or advantages:

[0058] This scheme establishes two fault detection units, namely a first fault detection unit and a second fault detection unit. The first fault detection unit outputs a first fault detection result, and the second fault detection unit outputs a second fault detection result. Finally, the first fault detection result and the second fault detection result are processed by the fault result output unit to output a final fault detection result. The fault detection results of the two fault detection units are used to verify and supplement each other, thereby improving the accuracy of the entire pumping unit fault detection system in pumping unit fault detection.

[0059] This scheme converts the complex analysis and processing of the oil pumping unit indicator diagram into the research of the oil pumping unit power curve. It not only establishes the second power curve diagram of the oil pumping unit corresponding to a single fault, but also establishes the third power curve diagram of the oil pumping unit corresponding to the influence of multiple faults. Then, the first power curve diagram of the oil pumping unit is obtained by using the oil pumping unit indicator diagram obtained in real time. Finally, the standard power curve diagram, the first power curve diagram, the second power curve diagram and the third power curve diagram are processed by the first fault detection unit to output the first fault detection result. This not only ensures the timeliness of the oil pumping unit fault detection, but also can output the detection results of a single fault or multiple faults, greatly improving the accuracy of the oil pumping unit fault detection.

[0060] This solution utilizes the fault history data of the oil pump and, based on the premise that there are differences in the changes in the operating parameters of the oil pump under different faults, obtains the change information of each operating parameter of the oil pump in real time, and outputs the second fault detection result through the second fault detection model, thereby improving the accuracy of the oil pump fault detection. BRIEF DESCRIPTION OF THE DRAWINGS

[0061] The drawings described herein are used to provide a further understanding of the embodiments of the present invention, constitute a part of the present invention, and do not constitute a limitation on the embodiments of the present invention;

[0062] Figure 1 It is a schematic diagram of the composition of the intelligent detection system for oil pumping unit failure in the present invention;

[0063] Figure 2 It is the standard indicator diagram of the oil pumping unit in the present invention. DETAILED DESCRIPTION

[0064] In order to more clearly understand the above-mentioned purpose, features and advantages of the present invention, the present invention is further described in detail below in conjunction with the accompanying drawings and specific embodiments. It should be noted that the embodiments of the present invention and the features in the embodiments can be combined with each other without conflict.

[0065] In the following description, many specific details are set forth to facilitate a full understanding of the present invention. However, the present invention may also be implemented in other ways different from those within the scope of this description. Therefore, the protection scope of the present invention is not limited to the specific embodiments disclosed below.

[0066] Example

[0067] Please refer to Figure 1 and Figure 2 , Figure 1 It is the intelligent detection system for oil pumping unit failure in the present invention. Figure 2 It is a standard indicator diagram of the oil pumping unit in the present invention, and the system comprises:

[0068] Standard unit: obtain the standard power diagram of the pumping unit and process it to obtain the standard power curve diagram;

[0069] The first monitoring unit is used to obtain the power diagram of the oil pumping unit in real time and process it to obtain a first power curve diagram, and put all the first power curve diagrams into a first set;

[0070] First data unit: obtaining multiple preset fault types, each preset fault type includes several sub-faults, each sub-fault includes a corresponding power diagram, processing the power diagram included in each sub-fault to obtain a corresponding second power curve diagram, and putting all the second power curve diagrams and corresponding sub-faults into a second set;

[0071] Second data unit: obtain all second power curve graphs in the first data unit and process them to obtain processing results, obtain a third power curve graph under the common influence of multiple sub-faults based on the processing results, record multiple sub-faults corresponding to each third power curve graph as a fault set, establish a first mapping relationship between the third power curve graph and the fault set, and put all third power curve graphs and corresponding fault sets into a third set;

[0072] A first fault detection unit: inputs the standard power curve, the first set, the second set and the third set into a first fault detection model, trains the first fault detection model, and outputs a first fault detection result of the oil pumping unit based on the trained first fault detection model;

[0073] The second monitoring unit: acquires the operating parameters of the oil pumping unit in real time to obtain a first acquisition result, obtains first change information of each operating parameter based on the first acquisition result, and puts all the first change information into a fourth set;

[0074] The third data unit: obtains fault history data of the oil pumping unit, obtains second change information of all operating parameters under each fault type based on the fault history data, establishes a second mapping relationship between each fault type and the second change information of the corresponding operating parameters, and puts all fault types and the second change information of the corresponding operating parameters into a fifth set;

[0075] A second fault detection unit: inputting the fourth set and the fifth set into a second fault detection model, training the second fault detection model, and outputting a second fault detection result of the oil pumping unit based on the trained second fault detection model;

[0076] Fault result output unit: outputs the pumping unit fault detection result after processing the first fault detection result and the second fault detection result.

[0077] The specific embodiment of the present invention is as follows: This embodiment is divided into three parts, the first part is that the first fault detection unit realizes fault detection, the second part is that the second fault detection unit realizes fault detection, and the third part is the final pumping unit fault detection result output;

[0078] The first part specifically includes:

[0079] Standard unit: obtain the standard power diagram of the pumping unit and process it to obtain the standard power curve diagram; please refer to Figure 2 , Figure 2 It is the standard indicator diagram of the oil pumping unit. Figure 2 In the figure, the abscissa S is the stroke of the pumping rod, the ordinate P is the load on the pumping rod, and the closed area ABCD is the power of the pumping rod;

[0080] Processing the standard dynamometer diagram to obtain a standard power curve diagram includes: dividing the stroke of the pumping unit rod into a number of equal intervals, calculating the area of ​​the corresponding curve closed region in each interval to obtain a calculation result; based on the calculation result, obtaining the power of the pumping unit rod corresponding to each interval; using a preset series to numerically name each interval, the preset series can be an arithmetic series with a first term of 0 and an arithmetic progression of 1, that is, after numerical naming, each interval corresponds to 0, 1, 2, ... and z in sequence; using all the numerically named intervals as the horizontal coordinates and the power of the pumping unit rod corresponding to each interval as the vertical coordinate to construct a standard power curve diagram. It should be noted that in this embodiment, all dynamometer diagrams are converted into corresponding power curve diagrams using the above method and the above standard, which will not be elaborated in detail later.

[0081] The first monitoring unit is used to obtain the power diagram of the oil pumping unit in real time and process it to obtain a first power curve diagram, and put all the first power curve diagrams into a first set;

[0082] First data unit: obtaining multiple preset fault types, each preset fault type includes several sub-faults, each sub-fault includes a corresponding power diagram, processing the power diagram included in each sub-fault to obtain a corresponding second power curve diagram, and putting all the second power curve diagrams and corresponding sub-faults into a second set;

[0083] Among them, all preset fault types include oil well sand production fault, oil well wax deposition fault, gas impact fault, leakage fault, heavy oil fault, fluid supply fault, oil well pumping and spraying fault and oil pumping equipment fault. Each preset fault type contains several sub-faults, which are as follows:

[0084]

[0085]

[0086] …

[0087]

[0088] , ,……and Corresponding to different preset fault types, , ,……and Corresponding to the sub-faults under different preset fault types, , ,……and The number of sub-faults corresponding to different preset fault types.

[0089] like Corresponding to the oil well sand production fault type, the oil well sand production fault type includes They correspond to the piston sand resistance fault, the fixed valve card dead fault, the fixed valve card dead on the valve cover fault and the floating valve failure fault. The corresponding value is 4. It should be noted that the number of sub-faults contained in each preset fault type is not exactly the same.

[0090] Second data unit: obtaining all second power curve graphs in the data unit and processing them to obtain processing results, obtaining a third power curve graph under the common influence of multiple sub-faults based on the processing results, recording multiple sub-faults corresponding to each third power curve graph as a fault set, establishing a first mapping relationship between the third power curve graph and the fault set, and placing all third power curve graphs and corresponding fault sets into a third set;

[0091] Obtaining the third power curve diagram under the common influence of multiple sub-faults includes:

[0092] The number of multiple sub-faults is preset to R, 2≤R≤m;

[0093] Randomly determine R fault types from a plurality of preset fault types;

[0094] Among the determined R fault types, a single sub-fault in each fault type and a corresponding second power curve graph are randomly obtained to obtain a second acquisition result;

[0095] Based on the second acquisition result shown, second power curve graphs corresponding to R sub-faults are acquired and put into a preset set;

[0096] Draw an average power curve graph of all the second power curve graphs in the preset set, wherein the average power curve graph uses the abscissa of any second power curve graph as the abscissa and uses the average power of the corresponding pumping rods in all the second power curve graphs as the ordinate;

[0097] The average power curve graph that is drawn is the third power curve graph.

[0098] The average power of the corresponding pumping rods in all the second power curves is calculated using the following formula:

[0099]

[0100] is the average power of the pumping rod corresponding to the first horizontal coordinate in all the second power curves, In each second power curve diagram, the first abscissa corresponds to the power of the pumping rod.

[0101] If the number of sub-faults is 2, it is necessary to randomly determine 2 fault types from multiple preset fault types. and , and then from and A single sub-fault is randomly obtained in each fault type, which are and , and obtain and The corresponding second power curve is or The horizontal coordinate of the corresponding second power curve is used as the horizontal coordinate of the average power curve. and The average power of the pumping rod in the corresponding second power curve is used as the ordinate of the average power curve, that is, =( + ) / 2. It should be noted that the vertical coordinates corresponding to all horizontal coordinates in the average power curve are processed by the above formula. At the same time, if the number of preset fault types is 8, then the possibility of randomly determining 2 fault types is 28. After determining the number of sub-faults and the value of R, it is necessary to obtain the third power curve under all combinations in the above way; if the number of multiple sub-faults is 3, it is necessary to randomly determine 3 fault types from multiple preset fault types, which are , and , and then from , and A single sub-fault is randomly obtained in each fault type, which are , and , and obtain , and The corresponding second power curve is , and The horizontal coordinate of the corresponding second power curve is used as the horizontal coordinate of the average power curve. , and The average power of the pumping rod in the corresponding second power curve is used as the ordinate of the average power curve, that is, =( + + ) / 3. Similarly, it is necessary to obtain the third power curve diagram under all sub-fault combinations when the value of R is 3.

[0102] In a more preferred embodiment, when calculating the average power of the pumping unit rods in the plurality of second power curves, different influence coefficients may be set for different preset fault types, that is, the following formula is evolved:

[0103]

[0104] are the influence coefficients of different preset fault types. The influence coefficients of all sub-faults under the same preset fault type are the same. By adding influence coefficients to different preset fault types, the third power curve diagram can further accurately reflect the fault condition of the oil pump under the influence of multiple sub-faults.

[0105] A first fault detection unit: inputs the standard power curve, the first set, the second set and the third set into a first fault detection model, trains the first fault detection model, and outputs a first fault detection result of the oil pumping unit based on the trained first fault detection model;

[0106] Based on the trained first fault detection model, outputting the first fault detection result of the oil pumping unit includes:

[0107] Calculating a first correlation between the first power curve graph and the standard power curve graph in the first set;

[0108] Determine whether the first correlation is greater than a first preset threshold value, if not, record the first power curve graph corresponding to the first correlation being greater than the first preset threshold value as a first power curve label graph, calculate the second correlation between the first power curve label graph and all second power curve graphs in the second set, and if so, continue to calculate the first correlation between the first power curve graph in the first set and the standard power curve graph;

[0109] Determine whether the second correlation exists and whether it is greater than the second preset threshold value. If so, obtain the corresponding second power curve diagram and sub-fault and output them as the first fault detection result; if not, calculate the third correlation between the first power curve mark diagram and all third power curve diagrams in the third set, obtain the third power curve diagram and fault set corresponding to the third correlation being greater than the third preset threshold value, and output them as the first fault detection result.

[0110] Wherein, if the first preset threshold, the second preset threshold and the third preset threshold are set to be , and , and the three preset thresholds can be adjusted according to actual conditions, and the present invention does not make specific limitations; firstly, the first correlation between all the first power curves in the first set and the standard power curve is calculated, if there is a first correlation value less than or equal to , indicating that the pumping unit may fail and further verification is needed. If all the first correlations are greater than , indicating that the pumping unit is in good working condition and only needs to be continuously monitored; when the pumping unit may fail, obtain the corresponding first power curve labeling diagram, calculate the second correlation between the first power curve labeling diagram and all the second power curve diagrams in the second set, and if there is a second correlation value greater than , indicating that the pumping unit is more likely to have a single sub-fault at this time, and the corresponding second power curve and the included sub-faults are output as the first fault detection result. If all the second correlations are less than or equal to , indicating that the failure of the pumping unit at this time is the result of the joint influence of multiple sub-faults. Therefore, the third correlation between the first power curve labeling diagram and all the third power curve diagrams in the third set is calculated, and the third correlation greater than The corresponding third power curve diagram and fault set are output as the first fault detection result.

[0111] The second part specifically includes:

[0112] The second monitoring unit: acquires the operating parameters of the oil pumping unit in real time to obtain a first acquisition result, obtains first change information of each operating parameter based on the first acquisition result, and puts all the first change information into a fourth set;

[0113] The third data unit: obtains fault history data of the oil pumping unit, obtains second change information of all operating parameters under each fault type based on the fault history data, establishes a second mapping relationship between each fault type and the second change information of the corresponding operating parameters, and puts all fault types and the second change information of the corresponding operating parameters into a fifth set;

[0114] The second fault detection unit: inputs the fourth set and the fifth set into a second fault detection model, trains the second fault detection model, and outputs a second fault detection result of the oil pump based on the trained second fault detection model.

[0115] Among them, the operating parameters include the page height of the pumping unit well, the casing pressure of the pumping unit well, the motor temperature and the vibration frequency, etc. This embodiment uses the fault history data of the pumping unit. Based on the premise that there are differences in the changes in the operating parameters of the pumping unit under different faults, the change information of each operating parameter of the pumping unit is obtained in real time, and the second fault detection result is output through the second fault detection model, thereby improving the accuracy of the pumping unit fault detection.

[0116] The third part includes:

[0117] Fault result output unit: outputs the pumping unit fault detection result after processing the first fault detection result and the second fault detection result.

[0118] Processing the first fault detection result and the second fault detection result includes:

[0119] Determine whether the first fault detection result is consistent with the second fault detection result. If so, output the first fault detection result or the second fault detection result as the pumping unit fault detection result; if not, output the same fault detection result of the first fault detection result and the second fault detection result, and mark and troubleshoot the different fault detection results of the first fault detection result and the second fault detection result.

[0120] Among them, this embodiment processes the first fault detection result and the second fault detection result through the fault result output unit to output the final fault detection result, and uses the fault detection results of the two fault detection units to verify and supplement each other, thereby improving the accuracy of the entire pumping unit fault detection system for pumping unit fault detection.

[0121] Although the preferred embodiments of the present invention have been described, those skilled in the art may make other changes and modifications to these embodiments once they have learned the basic creative concept. Therefore, the appended claims are intended to be interpreted as including the preferred embodiments and all changes and modifications that fall within the scope of the present invention.

[0122] Obviously, those skilled in the art can make various changes and modifications to the present invention without departing from the spirit and scope of the present invention. Thus, if these modifications and variations of the present invention fall within the scope of the claims of the present invention and their equivalents, the present invention is also intended to include these modifications and variations.

Claims

1. Intelligent detection system for oil pumping unit faults, characterized in that: The system comprises: Standard unit: obtain the standard power diagram of the pumping unit and process it to obtain the standard power curve diagram; The first monitoring unit is used to obtain the power diagram of the oil pumping unit in real time and process it to obtain a first power curve diagram, and put all the first power curve diagrams into a first set; First data unit: obtaining multiple preset fault types, each preset fault type includes several sub-faults, each sub-fault includes a corresponding power diagram, processing the power diagram included in each sub-fault to obtain a corresponding second power curve diagram, and putting all the second power curve diagrams and corresponding sub-faults into a second set; Second data unit: obtain all second power curve graphs in the first data unit and process them to obtain processing results, obtain a third power curve graph under the common influence of multiple sub-faults based on the processing results, record multiple sub-faults corresponding to each third power curve graph as a fault set, establish a first mapping relationship between the third power curve graph and the fault set, and put all third power curve graphs and corresponding fault sets into a third set; A first fault detection unit: inputs the standard power curve, the first set, the second set and the third set into a first fault detection model, trains the first fault detection model, and outputs a first fault detection result of the oil pumping unit based on the trained first fault detection model; The second monitoring unit: acquires the operating parameters of the oil pumping unit in real time to obtain a first acquisition result, obtains first change information of each operating parameter based on the first acquisition result, and puts all the first change information into a fourth set; The third data unit: obtains fault history data of the oil pumping unit, obtains second change information of all operating parameters under each fault type based on the fault history data, establishes a second mapping relationship between each fault type and the second change information of the corresponding operating parameters, and puts all fault types and the second change information of the corresponding operating parameters into a fifth set; A second fault detection unit: inputting the fourth set and the fifth set into a second fault detection model, training the second fault detection model, and outputting a second fault detection result of the oil pumping unit based on the trained second fault detection model; Fault result output unit: outputs the pumping unit fault detection result after processing the first fault detection result and the second fault detection result; Real-time acquisition of the pumping unit's power diagram and processing to obtain a first power curve diagram include: The power diagram of the pumping unit has the stroke of the pumping unit rod as the abscissa, the load on the pumping unit rod as the ordinate, and the area of ​​the curve enclosed region is the power of the pumping unit rod; In the dynamometer diagram of the oil pumping unit, the stroke of the oil pumping unit rod is divided into a number of equal intervals, and the area of ​​the corresponding curve closed region in each interval is calculated to obtain a calculation result; Based on the calculation results, the power of the pumping rod corresponding to each interval is obtained; Use a preset number series to numerically name each interval; The first power curve is constructed by taking all the intervals named with numerical values ​​as the horizontal coordinates and the power of the pumping rod corresponding to each interval as the vertical coordinate; The corresponding second power curve diagram obtained by processing the power diagram included in each sub-fault includes: In the dynamometer diagram included in the sub-fault, the stroke of the pumping unit rod is divided into a number of equal intervals, and the area of ​​the corresponding curve closed region in each interval is calculated to obtain a calculation result; Based on the calculation results, the power of the pumping rod corresponding to each interval is obtained; Use a preset number series to numerically name each interval; The second power curve is constructed by taking all the intervals named with numerical values ​​as the horizontal coordinates and the power of the pumping rod corresponding to each interval as the vertical coordinate; Each preset fault type contains several sub-faults, which are: …… , ,……and Corresponding to different preset fault types, , ,……and Corresponding to the sub-faults under different preset fault types, , ,……and The number of sub-faults corresponding to different preset fault types; Obtaining the third power curve diagram under the common influence of multiple sub-faults includes: The number of multiple sub-faults is preset to R, 2≤R≤m; Randomly determine R fault types from a plurality of preset fault types; Among the determined R fault types, a single sub-fault in each fault type and a corresponding second power curve graph are randomly obtained to obtain a second acquisition result; Based on the second acquisition result shown, second power curve graphs corresponding to R sub-faults are acquired and put into a preset set; Draw an average power curve graph of all the second power curve graphs in the preset set, wherein the average power curve graph uses the abscissa of any second power curve graph as the abscissa and uses the average power of the corresponding pumping rods in all the second power curve graphs as the ordinate; The average power curve graph that is drawn is the third power curve graph; The average power of the corresponding pumping rods in all the second power curves is calculated using the following formula: is the average power of the pumping rod corresponding to the first horizontal coordinate in all the second power curves, The first abscissa in each second power curve graph corresponds to the power of the pumping rod; Based on the trained first fault detection model, outputting the first fault detection result of the oil pumping unit includes: Calculating a first correlation between the first power curve graph and the standard power curve graph in the first set; Determine whether the first correlation is greater than a first preset threshold value, if not, record the first power curve graph corresponding to the first correlation being greater than the first preset threshold value as a first power curve label graph, calculate the second correlation between the first power curve label graph and all second power curve graphs in the second set, and if so, continue to calculate the first correlation between the first power curve graph in the first set and the standard power curve graph; Determine whether the second correlation is greater than a second preset threshold value, and if so, obtain the corresponding second power curve graph and sub-fault and output them as the first fault detection result; if not, calculate the third correlation between the first power curve mark graph and all third power curve graphs in the third set, obtain the third power curve graph and fault set corresponding to the third correlation being greater than the third preset threshold value, and output them as the first fault detection result; Processing the first fault detection result and the second fault detection result includes: Determine whether the first fault detection result is consistent with the second fault detection result. If so, output the first fault detection result or the second fault detection result as the pumping unit fault detection result; if not, output the same fault detection result of the first fault detection result and the second fault detection result, and mark and troubleshoot the different fault detection results of the first fault detection result and the second fault detection result.

2. The intelligent detection system for oil pumping unit failure according to claim 1 is characterized in that: Based on the trained second fault detection model, the second fault detection result of the pumping unit is outputted including: Obtaining first change information of each operating condition parameter in the fourth set; Traversing the second change information of each operating condition parameter in the fifth set, and filtering to obtain the second change information corresponding to the first change information; Based on the second mapping relationship and the filtered second change information, the second fault detection result is output.

3. The intelligent detection system for oil pumping unit failure according to claim 1 is characterized in that: All preset fault types include oil well sand production fault, oil well wax deposition fault, gas impact fault, leakage fault, heavy oil fault, fluid supply fault, oil well pumping and spraying fault and oil pumping equipment fault.

Images