Control method of gas-driven foam scrubbing filling pry based on neural network algorithm

Abstract

The invention relates to the technical field of control methods, and particularly discloses a control method of a gas-driven foam scrubbing filling pry based on a neural network algorithm, which comprises the following steps: step 1, in the filling process of the gas-driven foam scrubbing filling pry, detecting a foam scrubbing agent in an oil-gas well to obtain detection data; 2, calculating an evaluation coefficient based on the detection data; 3, whether the filling process of the gas-driven foam scrubbing filling pry is abnormal or not is evaluated, and a state evaluation signal is generated; 4, detecting the gas injection pressure based on the state early warning signal, evaluating the relevance between the gas injection pressure and the state of the gas-driven foam scrubbing filling skid in the filling process, and generating a relevance evaluation signal; 5, adjusting the gas injection pressure based on the correlated signals; real-time monitoring, accurate evaluation and intelligent regulation and control of the filling process are achieved, the safety and efficiency of oil and gas production are effectively improved, the fault occurrence probability is reduced by reducing human intervention, and the problem response speed is increased.

Description

Technical Field

[0001] The invention relates to the technical field of control methods, and in particular to a control method for a gas drive foam pumping filling skid based on a neural network algorithm. Background Art

[0002] The gas drive foam injection skid is a special equipment used in the field of oil and gas extraction. It combines gas drive technology and foam injection technology to improve the extraction efficiency of oil and gas wells. The gas drive foam injection skid is usually designed to automatically inject foam balls or foam injection agents into oil and gas wells to help improve the drainage and gas recovery capacity of gas wells, thereby increasing the production of gas wells.

[0003] However, in the prior art, the detection of the filling process status of the gas-driven foaming filling skid is mainly based on passive detection or judgment based on experience, and there is no periodic detection and real-time data analysis to quickly capture the subtle changes in the foaming agent content, resulting in untimely and inaccurate identification of abnormal conditions. Summary of the invention

[0004] The purpose of the present invention is to provide a control method for a gas drive foam injection skid based on a neural network algorithm to solve the technical problems in the above background.

[0005] The purpose of the present invention can be achieved through the following technical solutions: The present invention provides a control method for a gas drive foaming filling skid based on a neural network algorithm, comprising the following steps: Step 1: During the filling process of the gas drive foam filling skid, the foaming agent in the oil and gas well is tested to obtain the test data; Among them, the test data include: abnormal degree coefficient and content fluctuation coefficient; Step 2: Calculate the evaluation coefficient based on the test data; Step 3: Based on the evaluation coefficient, evaluate whether the filling process of the gas drive foam filling skid is abnormal, and generate a state evaluation signal; Among them, the status assessment signal includes: normal status signal, status warning signal and abnormal status signal; Step 4: Based on the status warning signal, the gas injection pressure is detected, the correlation between the gas injection pressure and the state of the gas drive foam filling skid filling process is evaluated, and a correlation evaluation signal is generated; The correlation evaluation signal includes: a correlation signal and an uncorrelated signal; Step 5: Adjust the gas injection pressure based on the associated signal.

[0006] As a further solution of the present invention: the process of acquiring the detection data is as follows: Preset the detection cycle, and detect the foam expelling agent in the oil and gas wells to obtain the deviation characterization value of the foam expelling agent in the oil and gas wells in each detection cycle within the target period; Based on the deviation characterization values ​​of each detection cycle, the abnormality degree coefficient and content fluctuation coefficient were calculated respectively.

[0007] As a further solution of the present invention: the process of obtaining the deviation characterization value is: Based on any detection period, extract the average content of the foaming agent in the oil and gas wells in the detection period to obtain the detection content of the foaming agent; The difference between the detected content of the foaming agent and the set content of the foaming agent is calculated to obtain the content deviation value, and then the ratio of the content deviation value to the set content of the foaming agent is calculated to obtain the deviation characterization value.

[0008] As a further solution of the present invention: the process of obtaining the abnormal degree coefficient is: Based on the deviation characterization values ​​of each detection cycle within the target period, each deviation characterization value is compared and analyzed with the deviation characterization threshold value; If the deviation characterization value is less than or equal to the deviation characterization threshold, a normal period content signal is generated; If the deviation characterization value is greater than the deviation characterization threshold, a cycle content abnormality signal is generated, and the detected cycle is marked as an abnormal cycle; Extract the deviation characterization values ​​of all abnormal cycles, calculate the mean of the deviation characterization values ​​of all abnormal cycles, obtain the abnormal characterization mean, and then calculate the ratio of the abnormal characterization mean to the deviation characterization mean to obtain the abnormal deviation ratio; At the same time, the number of all abnormal cycles is extracted to obtain the number of abnormal cycles, and then the ratio of the number of abnormal cycles to the number of detection cycles is calculated to obtain the ratio of the number of abnormal cycles; The abnormal deviation ratio is multiplied by the abnormal cycle number ratio to obtain the abnormal deviation coefficient, and then the abnormal deviation coefficient is multiplied by the deviation representation mean to obtain the abnormal degree coefficient.

[0009] As a further solution of the present invention: the process of obtaining the content fluctuation coefficient is: Based on the deviation characterization values ​​of all detection cycles within the target period, the variance of the deviation characterization values ​​of each detection cycle is calculated to obtain the content fluctuation coefficient.

[0010] As a further solution of the present invention: the process of obtaining the evaluation coefficient is: By formula: , calculate the evaluation coefficient PG, where a1 and a2 are preset proportional factors, and both are greater than 0, CD is the abnormality degree coefficient, and BD is the content fluctuation coefficient.

[0011] As a further solution of the present invention: the process of generating the state evaluation signal is: Preset a first threshold and a second threshold, and compare and analyze the evaluation coefficient with the first threshold and the second threshold respectively; If the evaluation coefficient is ≤ the first threshold, a normal state signal is generated; If the first threshold value < the evaluation coefficient ≤ the second threshold value, a status warning signal is generated; If the evaluation coefficient>the second threshold, a state abnormality signal is generated.

[0012] As a further solution of the present invention: the process of generating the relevance evaluation signal is: Based on the status warning signal, the gas injection pressure of the gas drive foam filling skid during the filling process is detected, and the pressure deviation ratio of each abnormal period within the target period is obtained respectively; Based on any abnormal period, the pressure deviation ratio of the abnormal period is calculated with the deviation characterization value to obtain the associated parameter; In a two-dimensional coordinate system established with the abnormal period as the X-axis and the associated parameter as the Y-axis, a trend line of the associated parameter is drawn, and with the Y-axis value as the mean value of the associated parameter, a straight line parallel to the X-axis is drawn to obtain the mean straight line of the associated parameter; Among them, the mean of the associated parameters is the average value of the associated parameters of all abnormal cycles in the target period; The area of ​​the region enclosed by the trend line of the associated parameter and the straight line of the mean value of the associated parameter is extracted, and a ratio calculation is performed at a preset area threshold to obtain a floating representation value of the associated parameter; At the same time, the line between each two adjacent abnormal cycles in the trend line of the associated parameter is marked as a trend sub-line, the slope of each trend sub-line is obtained, and the absolute value is taken to obtain the representation value of the associated parameter change; All the correlation parameter change representation values ​​are compared and analyzed with the correlation parameter change representation thresholds; If the associated parameter change representation value is less than or equal to the associated parameter change representation threshold, then an associated parameter gradual change signal is generated; If the correlation parameter change representation value is greater than the correlation parameter change representation threshold, a correlation parameter mutation signal is generated, and the trend sub-fold line corresponding to the correlation parameter change representation value is marked as a mutation sub-fold line; The area of ​​the region enclosed by each mutation sub-polyline and the associated parameter mean straight line is extracted respectively, and the ratio is calculated at the preset area threshold to obtain the associated parameter mutation characterization value; Then the parameter floating representation value and the associated parameter mutation representation value are summed up to obtain the correlation coefficient; Preset a correlation coefficient threshold, and compare and analyze the correlation coefficient with the correlation coefficient threshold; If the correlation coefficient is less than the correlation coefficient threshold, a correlation signal is generated; If the correlation coefficient is ≥ the correlation coefficient threshold, an uncorrelated signal is generated.

[0013] As a further solution of the present invention: the process of obtaining the pressure deviation ratio is: Based on any abnormal period, the average gas injection pressure of the abnormal period is obtained, the difference between the average gas injection pressure and the preset pressure is calculated, and the absolute value is taken to obtain the pressure deviation value, and then the ratio of the pressure deviation value to the preset pressure is calculated to obtain the pressure deviation ratio.

[0014] As a further solution of the present invention: the process of adjusting the gas injection pressure is: Based on the relevant signals, the mean value of abnormal characteristics and the mean pressure deviation ratio of the target period are obtained; By formula: , calculate the pressure adjustment value YT, where YJ is the mean value of abnormality representation, PC is the deviation representation threshold, YP is the pressure deviation ratio, and SY is the preset pressure; Based on the pressure adjustment value, the gas injection pressure of the next target period in the filling process of the gas drive foam filling skid is adjusted.

[0015] Beneficial effects of the present invention: (1) During the filling process, the present invention uses an optical sensor to periodically detect the content of the foaming agent in the oil and gas well, and compares it with the preset foaming agent set content to calculate the deviation characterization value, which reflects the degree of deviation between the foaming agent content and the set value; next, by statistically analyzing the deviation characterization values ​​in the target time period, the abnormal degree coefficient is calculated, which combines the severity (abnormal deviation ratio) and occurrence frequency (abnormal cycle number ratio) of the foaming agent content deviation, and combines the average level of the overall deviation to comprehensively evaluate the abnormal conditions in the filling process; at the same time, the content fluctuation coefficient is calculated to measure the fluctuation degree of the foaming agent content in different detection cycles, reflecting the stability of the filling process; finally, Based on the abnormality degree coefficient and the content fluctuation coefficient, the evaluation coefficient is calculated by a preset formula and compared with the set threshold value to judge the state of the filling process. If the evaluation coefficient is lower than the first threshold value, it means that the filling process is normal; if it is between the two threshold values, a warning signal is issued; if it is higher than the second threshold value, it is considered that the filling process is abnormal. The present invention realizes comprehensive real-time monitoring and accurate quantitative evaluation of the filling process of the gas-driven foaming filling skid. Through periodic detection and real-time data analysis, it can not only quickly capture the subtle changes in the content of the foaming agent, but also accurately quantify the impact of these changes on production stability. The present invention surpasses traditional passive detection or judgment based on experience, making the identification of abnormal conditions more timely and accurate. (2) After receiving the status warning of the filling process, the present invention conducts a detailed analysis of the gas injection pressure. By calculating the pressure deviation ratio in each abnormal period and performing a correlation analysis with the deviation characterization value of the filling process, the system can accurately determine the correlation between the gas injection pressure and the filling status. This process not only relies on real-time data, but also combines preset thresholds and empirical values ​​to ensure the accuracy and reliability of the analysis. When it is found that there is a significant correlation between the gas injection pressure and the filling status, the pressure adjustment value is calculated. This adjustment strategy aims to quickly respond to and improve abnormal conditions in the filling process by fine-tuning the gas injection pressure, thereby ensuring the continuity and efficiency of oil and gas well production. The present invention realizes real-time monitoring, precise evaluation and intelligent regulation of the filling process, effectively improving the safety and efficiency of oil and gas production, reducing the probability of failure by reducing human intervention, and accelerating the response to problems. BRIEF DESCRIPTION OF THE DRAWINGS

[0016] The present invention will be further described below in conjunction with the accompanying drawings.

[0017] Figure 1 It is a flowchart of the present invention; Figure 2 It is a system block diagram of the present invention. DETAILED DESCRIPTION

[0018] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0019] Embodiment 1: See also Figure 1 As shown, a control method for a gas drive foaming filling skid based on a neural network algorithm according to an embodiment of the present invention comprises the following steps: Step 1: During the filling process of the gas drive foam filling skid, the foaming agent in the oil and gas well is tested to obtain the test data; Among them, the test data include: abnormal degree coefficient and content fluctuation coefficient; In some embodiments, a detection cycle is preset (the duration of the detection cycle can be 1 minute or 5 minutes), and the foam expelling agent in the oil and gas well is detected to obtain the deviation characterization value of the foam expelling agent in the oil and gas well in each detection cycle within a target period (the target period includes multiple detection cycles, and the number of detection cycles can be 30 or 60); Based on the deviation characterization values ​​of each detection cycle, the abnormality degree coefficient and the content fluctuation coefficient are calculated respectively; First, specifically, the process of obtaining the deviation characterization value is: Based on any detection period, the average content of the foaming agent in the oil and gas wells in the detection period is extracted by using an optical sensor to obtain the detection content of the foaming agent; The difference between the detected content of the foaming agent and the set content of the foaming agent is calculated to obtain a content deviation value, and then the ratio of the content deviation value to the set content of the foaming agent is calculated to obtain a deviation characterization value; It should be explained that the set content of the foaming agent is the content of the foaming agent in the oil and gas well pre-set by those skilled in the art based on the actual situation of the oil and gas well and production requirements, and this value is an empirical value; It needs to be further explained that the deviation characterization value indicates the degree of deviation between the foaming agent content and the set value within the detection period. The larger the value of the deviation characterization value, the higher the degree of deviation between the foaming agent content and the set value, and the higher the degree of abnormality. Second, specifically, the process of obtaining the abnormality degree coefficient is: Based on the deviation characterization values ​​of each detection cycle within the target period, each deviation characterization value is compared and analyzed with the deviation characterization threshold value; If the deviation characterization value is less than or equal to the deviation characterization threshold, it means that the content deviation of the foaming agent in the detection cycle is within an acceptable range, that is, a normal cycle content signal is generated; If the deviation characterization value is greater than the deviation characterization threshold, it means that the content deviation of the foaming agent in the detection cycle exceeds the normal range, that is, a cycle content abnormality signal is generated, and the detection cycle is marked as an abnormal cycle; Extract the deviation characterization values ​​of all abnormal cycles, calculate the average of the deviation characterization values ​​of all abnormal cycles to obtain the abnormal characterization mean, and then calculate the ratio of the abnormal characterization mean to the deviation characterization mean (the deviation characterization mean is the average value of the deviation characterization values ​​of all detection cycles in the target period) to obtain the abnormal deviation ratio; At the same time, the number of all abnormal cycles is extracted to obtain the number of abnormal cycles, and then the ratio of the number of abnormal cycles to the number of detection cycles (the number of detection cycles is the total number of all detection cycles in the target period) is calculated to obtain the ratio of the number of abnormal cycles; The abnormal deviation ratio is multiplied by the abnormal cycle number ratio to obtain the abnormal deviation coefficient, and then the abnormal deviation coefficient is multiplied by the deviation representation mean to obtain the abnormal degree coefficient; It should be explained that the abnormality degree coefficient represents the comprehensive quantitative index of the overall abnormality degree and impact range of the deviation of the foaming agent content in the oil and gas wells during the entire target period. The abnormality degree coefficient combines the severity of the deviation of the foaming agent content (reflected by the abnormal deviation ratio) and the frequency of such deviation (reflected by the abnormal cycle number ratio), and further combines these two factors with the average level of the foaming agent content in the entire target period (reflected by the deviation characterization mean) to obtain an indicator that can comprehensively reflect the abnormal conditions in the foaming agent injection process; the larger the value of the abnormality degree coefficient, the more serious the deviation of the foaming agent content in the process of foaming agent injection in the oil and gas wells, and the higher the frequency of such deviation, that is, the more significant the overall abnormal condition; Thirdly, specifically, the process of obtaining the content fluctuation coefficient is as follows: Based on the deviation characterization values ​​of all detection cycles within the target period, the variance of the deviation characterization values ​​of each detection cycle is calculated to obtain the content fluctuation coefficient; It should be explained that the content fluctuation coefficient indicates the degree of fluctuation of the deviation value of the foaming agent content relative to its average level during the target period; the larger the content fluctuation coefficient, the more drastic the fluctuation of the foaming agent content during different detection cycles, that is, the worse the stability of the foaming agent filling process, and the more significant the overall abnormal condition; Step 2: Calculate the evaluation coefficient based on the test data; In some embodiments, by the formula: , calculate the evaluation coefficient PG, where a1 and a2 are preset proportional factors, both greater than 0, CD is the abnormality coefficient, and BD is the content fluctuation coefficient; Step 3: Based on the evaluation coefficient, evaluate whether the filling process of the gas drive foam filling skid is abnormal, and generate a state evaluation signal; Among them, the status assessment signal includes: normal status signal, status warning signal and abnormal status signal; In some embodiments, a first threshold and a second threshold are preset, and the evaluation coefficient is compared and analyzed with the first threshold and the second threshold respectively; If the evaluation coefficient is ≤ the first threshold value, it means that the filling process of the gas drive foam injection skid is in a normal state, that is, the content deviation and fluctuation of the foam injection agent are within an acceptable range, and will not have a significant impact on the production of the oil and gas well, and a normal state signal is generated; If the first threshold value < evaluation coefficient ≤ second threshold value, it means that the filling process begins to show abnormal trends, but has not yet reached the level of serious abnormality, and a status warning signal is generated; If the evaluation coefficient is greater than the second threshold, it means that the filling process of the gas flooding foam injection skid has been abnormal, and the content deviation and fluctuation of the foam injection agent are beyond the normal range, which may have a serious impact on the production of the oil and gas well and generate a state abnormality signal; The technical solution of the embodiment of the present invention is mainly as follows: during the filling process, the content of the foaming agent in the oil and gas well is periodically detected by an optical sensor, and compared with the preset foaming agent set content, and a deviation characterization value is calculated, and the deviation characterization value reflects the degree of deviation between the foaming agent content and the set value; next, the deviation characterization value within the target time period is statistically analyzed to calculate the abnormal degree coefficient, which combines the severity (abnormal deviation ratio) and the frequency of occurrence (abnormal cycle number ratio) of the foaming agent content deviation, and combines the average level of the overall deviation to comprehensively evaluate the abnormal conditions during the filling process; at the same time, the content fluctuation coefficient is calculated to measure the fluctuation degree of the foaming agent content in different detection cycles, reflecting the stability of the filling process; finally, Afterwards, based on the abnormality degree coefficient and the content fluctuation coefficient, the evaluation coefficient is calculated by a preset formula, and compared with the set threshold value to judge the state of the filling process. If the evaluation coefficient is lower than the first threshold value, it means that the filling process is normal; if it is between the two threshold values, a warning signal is issued; if it is higher than the second threshold value, it is regarded as abnormal in the filling process. The embodiment of the present invention realizes comprehensive real-time monitoring and accurate quantitative evaluation of the filling process of the gas drive foam drainage filling skid. Through periodic detection and real-time data analysis, it can not only quickly capture the subtle changes in the foam drainage agent content, but also accurately quantify the impact of these changes on production stability. The embodiment of the present invention surpasses the traditional passive detection or judgment based on experience, making the identification of abnormal conditions more timely and accurate.

[0020] Embodiment 2: Based on Example 1, please refer to Figure 1 As shown, the control method of the gas drive foam filling skid based on the neural network algorithm described in the embodiment of the present invention also includes the following steps: Step 4: Based on the status warning signal, the gas injection pressure is detected, the correlation between the gas injection pressure and the state of the gas drive foam filling skid filling process is evaluated, and a correlation evaluation signal is generated; The correlation evaluation signal includes: a correlation signal and an uncorrelated signal; In some practical schemes, based on the status warning signal, the gas injection pressure of the gas drive foam filling skid during the filling process is detected, and the pressure deviation ratio of each abnormal period within the target period is obtained respectively; Based on any abnormal period, the pressure deviation ratio of the abnormal period is calculated with the deviation characterization value to obtain the associated parameter; In a two-dimensional coordinate system established with the abnormal period as the X-axis and the associated parameter as the Y-axis, a trend line of the associated parameter is drawn, and with the Y-axis value as the mean value of the associated parameter, a straight line parallel to the X-axis is drawn to obtain the mean straight line of the associated parameter; Among them, the mean of the associated parameters is the average value of the associated parameters of all abnormal cycles in the target period; The area of ​​the region enclosed by the trend line of the associated parameter and the straight line of the mean value of the associated parameter is extracted, and a ratio calculation is performed at a preset area threshold to obtain a floating representation value of the associated parameter; At the same time, the line between each two adjacent abnormal cycles in the trend line of the associated parameter is marked as a trend sub-line, the slope of each trend sub-line is obtained, and the absolute value is taken to obtain the representation value of the associated parameter change; All the correlation parameter change representation values ​​are compared and analyzed with the correlation parameter change representation thresholds; If the correlation parameter change representation value is less than or equal to the correlation parameter change representation threshold, it means that the correlation parameter fluctuation between the two abnormal cycles corresponding to the correlation parameter change representation value is small, that is, the correlation parameter gradual change signal is generated; If the correlation parameter change representation value is greater than the correlation parameter change representation threshold, it means that the correlation parameter fluctuates greatly between the two abnormal periods corresponding to the correlation parameter change representation value, that is, a correlation parameter mutation signal is generated, and the trend sub-line corresponding to the correlation parameter change representation value is marked as a mutation sub-line; The area of ​​the region enclosed by each mutation sub-polyline and the associated parameter mean straight line is extracted respectively, and the ratio is calculated at the preset area threshold to obtain the associated parameter mutation characterization value; Then the parameter floating representation value and the associated parameter mutation representation value are summed up to obtain the correlation coefficient; Preset a correlation coefficient threshold, and compare and analyze the correlation coefficient with the correlation coefficient threshold; If the correlation coefficient is less than the correlation coefficient threshold, it is determined that the gas injection pressure has a large correlation with the state of the gas flooding bubble drainage filling skid filling process, and a correlation signal is generated; If the correlation coefficient is ≥ the correlation coefficient threshold, it is determined that the gas injection pressure has a small correlation with the state of the gas flooding bubble drainage filling skid filling process, and an uncorrelated signal is generated; Specifically, the process of obtaining the pressure deviation ratio is: Based on any abnormal period, the average gas injection pressure of the abnormal period is obtained, the difference between the average gas injection pressure and the preset pressure is calculated, and the absolute value is taken to obtain the pressure deviation value, and then the ratio of the pressure deviation value to the preset pressure is calculated to obtain the pressure deviation ratio; It should be explained that the preset pressure is the gas injection pressure preset by those skilled in the art to meet the filling process of the gas drive foam filling skid, and the value is an empirical value; Step 5: Adjust the gas injection pressure based on the relevant signal; In some embodiments, based on the associated signal, the average value of the abnormal characterization and the average pressure deviation ratio of the target period are obtained (the average pressure deviation ratio is the average value of the pressure deviation ratios of all abnormal periods in the target period); By formula: , calculate the pressure adjustment value YT, where YJ is the mean value of abnormality representation, PC is the deviation representation threshold, YP is the pressure deviation ratio, and SY is the preset pressure; Based on the pressure adjustment value, the gas injection pressure of the next target period in the filling process of the gas drive foam filling skid is adjusted; The technical solution of the embodiment of the present invention is mainly as follows: after receiving the status warning of the filling process, the gas injection pressure is analyzed in detail, and by calculating the pressure deviation ratio in each abnormal period and performing correlation analysis with the deviation characterization value of the filling process, the system can accurately judge the correlation between the gas injection pressure and the filling status. This process not only relies on real-time data, but also combines preset thresholds and experience values ​​to ensure the accuracy and reliability of the analysis. When it is found that there is a significant correlation between the gas injection pressure and the filling status, the pressure adjustment value is calculated. The adjustment strategy aims to quickly respond to and improve abnormal conditions in the filling process by fine-tuning the gas injection pressure, thereby ensuring the continuity and efficiency of oil and gas well production. The embodiment of the present invention realizes real-time monitoring, precise evaluation and intelligent regulation of the filling process, effectively improving the safety and efficiency of oil and gas production, reducing the probability of failure by reducing human intervention, and accelerating the response to problems.

[0021] Embodiment 3: Based on Example 1 and Example 2, please refer to Figure 2 As shown, a control system of a gas drive foaming and filling skid based on a neural network algorithm according to an embodiment of the present invention includes: Data acquisition module: During the filling process of the gas drive foam injection skid, the foam injection agent in the oil and gas well is tested and the test data is obtained; Among them, the test data include: abnormal degree coefficient and content fluctuation coefficient; Data processing module: calculate the evaluation coefficient based on the detection data; Status evaluation signal: Based on the evaluation coefficient, it evaluates whether the filling process of the gas drive foam filling skid is abnormal, and generates a status evaluation signal; Among them, the status assessment signal includes: normal status signal, status warning signal and abnormal status signal; Correlation evaluation signal: Based on the status warning signal, the gas injection pressure is detected, the correlation between the gas injection pressure and the state of the gas drive foam filling skid filling process is evaluated, and a correlation evaluation signal is generated; The correlation evaluation signal includes: a correlation signal and an uncorrelated signal; Control module: adjusts the gas injection pressure based on relevant signals.

[0022] The above thresholds are set for the convenience of comparison. The thresholds depend on the amount of sample data and the number of bases set by technicians in this field for each set of sample data. The above formulas are obtained by collecting a large amount of data and performing software simulation to select a formula close to the actual value. The factors in the formula are set by technicians in this field according to actual conditions; for example: Formula: ; A technician in this field collects multiple groups of test data and sets a corresponding evaluation coefficient for each group of test data; Substitute the set evaluation coefficient and the collected test data into the formula, any two formulas constitute a set of two-variable linear equations, screen the calculated factors and take the average, and obtain the values ​​of a1 and a2 as 0.67 and 0.33 respectively; The size of the factor is to quantify each parameter to obtain a specific value for subsequent comparison. The size of the factor depends on the amount of test data and the preliminary setting of the corresponding evaluation coefficient for each set of test data by technical personnel in this field; as long as it does not affect the proportional relationship between the parameter and the quantified value, such as the evaluation coefficient is proportional to the value of the abnormality degree coefficient.

[0023] The above is a detailed description of an embodiment of the present invention, but the content is only a preferred embodiment of the present invention and cannot be considered to limit the scope of implementation of the present invention. All equivalent changes and improvements made within the scope of the present invention should still fall within the scope of the patent coverage of the present invention.

Claims

1. A control method for a gas-driven foaming filling skid based on a neural network algorithm, characterized in that: The following steps are involved: Step 1: During the filling process of the gas drive foam filling skid, the foaming agent in the oil and gas well is tested to obtain the test data; Among them, the test data include: abnormal degree coefficient and content fluctuation coefficient; Step 2: Calculate the evaluation coefficient based on the test data; Step 3: Based on the evaluation coefficient, evaluate whether the filling process of the gas drive foam filling skid is abnormal, and generate a state evaluation signal; Among them, the status assessment signal includes: normal status signal, status warning signal and abnormal status signal; Step 4: Based on the status warning signal, the gas injection pressure is detected, the correlation between the gas injection pressure and the state of the gas drive foam filling skid filling process is evaluated, and a correlation evaluation signal is generated; The correlation evaluation signal includes: a correlation signal and an uncorrelated signal; Step 5: Adjust the gas injection pressure based on the associated signal.

2. The control method of the gas flooding foam filling skid based on the neural network algorithm according to claim 1 is characterized in that: The process of acquiring detection data is as follows: Preset the detection cycle, and detect the foam expelling agent in the oil and gas wells to obtain the deviation characterization value of the foam expelling agent in the oil and gas wells in each detection cycle within the target period; Based on the deviation characterization values ​​of each detection cycle, the abnormality degree coefficient and content fluctuation coefficient were calculated respectively.

3. The control method of the gas flooding foam filling skid based on the neural network algorithm according to claim 2 is characterized in that: The process of obtaining the deviation characterization value is as follows: Based on any detection period, extract the average content of the foaming agent in the oil and gas wells in the detection period to obtain the detection content of the foaming agent; The difference between the detected content of the foaming agent and the set content of the foaming agent is calculated to obtain the content deviation value, and then the ratio of the content deviation value to the set content of the foaming agent is calculated to obtain the deviation characterization value.

4. The control method of the gas flooding foam filling skid based on the neural network algorithm according to claim 3 is characterized in that: The process of obtaining the abnormality degree coefficient is as follows: Based on the deviation characterization values ​​of each detection cycle within the target period, each deviation characterization value is compared and analyzed with the deviation characterization threshold value; If the deviation characterization value is less than or equal to the deviation characterization threshold, a normal period content signal is generated; If the deviation characterization value is greater than the deviation characterization threshold, a cycle content abnormality signal is generated, and the detected cycle is marked as an abnormal cycle; Extract the deviation characterization values ​​of all abnormal cycles, calculate the mean of the deviation characterization values ​​of all abnormal cycles, obtain the abnormal characterization mean, and then calculate the ratio of the abnormal characterization mean to the deviation characterization mean to obtain the abnormal deviation ratio; At the same time, the number of all abnormal cycles is extracted to obtain the number of abnormal cycles, and then the ratio of the number of abnormal cycles to the number of detection cycles is calculated to obtain the ratio of the number of abnormal cycles; The abnormal deviation ratio is multiplied by the abnormal cycle number ratio to obtain the abnormal deviation coefficient, and then the abnormal deviation coefficient is multiplied by the deviation representation mean to obtain the abnormal degree coefficient.

5. The control method of the gas flooding foam filling skid based on the neural network algorithm according to claim 3 is characterized in that: The process of obtaining the content fluctuation coefficient is: Based on the deviation characterization values ​​of all detection cycles within the target period, the variance of the deviation characterization values ​​of each detection cycle is calculated to obtain the content fluctuation coefficient.

6. The control method of the gas flooding foam pumping filling skid based on the neural network algorithm according to claim 1 is characterized in that: The process of obtaining the evaluation coefficient is: By formula: , calculate the evaluation coefficient PG, where a1 and a2 are preset proportional factors, and both are greater than 0, CD is the abnormality degree coefficient, and BD is the content fluctuation coefficient.

7. The control method of the gas flooding foam filling skid based on the neural network algorithm according to claim 1 is characterized in that: The process of generating the status evaluation signal is: Preset a first threshold and a second threshold, and compare and analyze the evaluation coefficient with the first threshold and the second threshold respectively; If the evaluation coefficient is ≤ the first threshold, a normal state signal is generated; If the first threshold value < the evaluation coefficient ≤ the second threshold value, a status warning signal is generated; If the evaluation coefficient>the second threshold, a state abnormality signal is generated.

8. The control method of the gas flooding foam pumping filling skid based on the neural network algorithm according to claim 1 is characterized in that: The process of generating the relevance evaluation signal is: Based on the status warning signal, the gas injection pressure of the gas drive foam filling skid during the filling process is detected, and the pressure deviation ratio of each abnormal period within the target period is obtained respectively; Based on any abnormal period, the pressure deviation ratio of the abnormal period is calculated with the deviation characterization value to obtain the associated parameter; In a two-dimensional coordinate system established with the abnormal period as the X-axis and the associated parameter as the Y-axis, a trend line of the associated parameter is drawn, and with the Y-axis value as the mean value of the associated parameter, a straight line parallel to the X-axis is drawn to obtain the mean straight line of the associated parameter; Among them, the mean of the associated parameters is the average value of the associated parameters of all abnormal cycles in the target period; The area of ​​the region enclosed by the trend line of the associated parameter and the straight line of the mean value of the associated parameter is extracted, and a ratio calculation is performed at a preset area threshold to obtain a floating representation value of the associated parameter; At the same time, the line between every two adjacent abnormal cycles in the trend line of the associated parameter is marked as a trend sub-line, the slope of each trend sub-line is obtained, and the absolute value is taken to obtain the representation value of the associated parameter change; Compare and analyze all the correlation parameter change representation values ​​with the correlation parameter change representation thresholds; If the associated parameter change representation value is less than or equal to the associated parameter change representation threshold, then an associated parameter gradual change signal is generated; If the correlation parameter change representation value is greater than the correlation parameter change representation threshold, a correlation parameter mutation signal is generated, and the trend sub-fold line corresponding to the correlation parameter change representation value is marked as a mutation sub-fold line; The area of ​​the region enclosed by each mutation sub-polyline and the associated parameter mean straight line is extracted respectively, and the ratio is calculated at the preset area threshold to obtain the associated parameter mutation characterization value; Then the parameter floating representation value and the associated parameter mutation representation value are summed up to obtain the correlation coefficient; Preset a correlation coefficient threshold, and compare and analyze the correlation coefficient with the correlation coefficient threshold; If the correlation coefficient is less than the correlation coefficient threshold, a correlation signal is generated; If the correlation coefficient is ≥ the correlation coefficient threshold, an uncorrelated signal is generated.

9. The control method of the gas flooding foam pumping filling skid based on the neural network algorithm according to claim 8 is characterized in that: The process of obtaining the pressure deviation ratio is: Based on any abnormal period, the average gas injection pressure of the abnormal period is obtained, the difference between the average gas injection pressure and the preset pressure is calculated, and the absolute value is taken to obtain the pressure deviation value, and then the ratio of the pressure deviation value to the preset pressure is calculated to obtain the pressure deviation ratio.

10. The control method of the gas flooding foam pumping filling skid based on the neural network algorithm according to claim 1 is characterized in that: The process of adjusting the gas injection pressure is as follows: Based on the relevant signals, the mean value of abnormal characteristics and the mean pressure deviation ratio of the target period are obtained; By formula: , calculate the pressure adjustment value YT, where YJ is the mean value of abnormality representation, PC is the deviation representation threshold, YP is the pressure deviation ratio, and SY is the preset pressure; Based on the pressure adjustment value, the gas injection pressure of the next target period in the filling process of the gas drive foam filling skid is adjusted.

Images