Intelligent monitoring system and method for drilling overflow and well leakage

Abstract

The present application provides a drilling overflow and well leakage intelligent monitoring system and method, which relates to the field of intelligent monitoring. It first collects the mass flow at the inlet and outlet of the drilling circulation system and transmits it to the data processing cloud, and then uses artificial intelligence-based data analysis and processing algorithms to extract features from the collected data, so as to monitor the overflow and well leakage of the drilling by utilizing the time series coordinated changes between the mass flow time series data at the inlet and the mass flow time series data at the outlet. In this way, by comprehensively analyzing the time series characteristics of the inlet and outlet mass flow rates, the dynamic correlation of flow changes during the drilling process can be fully captured, and the monitoring coverage can be improved, thereby improving the intelligence of drilling overflow and well leakage monitoring, so as to improve the safety and efficiency of drilling operations.

Description

Technical Field

[0001] The present application relates to the field of intelligent monitoring, and more specifically, to a drilling overflow and well leakage intelligent monitoring system and method. Background Art

[0002] Drilling is a complex and delicate operation, during which you may encounter a variety of problems, the most common of which are overflow and lost circulation. If these problems are not dealt with in a timely manner, they may lead to serious safety problems and economic losses. For example, overflow may lead to blowout, causing environmental pollution, equipment damage and casualties; while lost circulation will cause a large amount of drilling fluid to be lost, destroy the stability of the well wall and increase drilling costs.

[0003] Chinese patent CN108729868B proposes a method for monitoring overflow and well leakage in deep-sea drilling. Mass flow meters are set at the inlet and outlet of the drilling circulation system to collect and transmit mass flow data to the data processing and analysis end in real time. The system calculates the overflow volume, total overflow volume, overflow rate, leakage volume, total leakage volume and leakage rate at different times, and determines the overflow and well leakage by monitoring the changes in mass flow.

[0004] In the above patent, the overflow and well leakage are calculated and judged based on the mass flow rate at the inlet and outlet of the drilling circulation system. However, the patent takes into account the trend of data changes in time series, that is, the overflow and well leakage are judged by calculating the difference in mass flow rate at different time points. But there is also an interactive relationship between the mass flow rate at the inlet and outlet in time. If the calculation is directly based on the values ​​of the two, it may cause the monitoring system to fail to capture the dynamic correlation of flow changes during the drilling process. For example, in some cases, the increase in inlet flow is not immediately reflected in the outlet flow, which may be due to a delayed response caused by well leakage or overflow. If the system only relies on data changes at a single point in time without considering the interaction between the flow rates at the previous and subsequent time points, it is difficult to accurately identify this delayed response, thereby increasing the risk of false alarms and missed reports. In addition, the monitoring method that lacks dynamic correlation cannot fully reflect the complex changes in the drilling process, resulting in reduced accuracy and reliability of the system in real-time response and prediction of future trends.

[0005] Therefore, an optimized drilling overflow and lost circulation monitoring solution is desired. Summary of the invention

[0006] In order to solve the above technical problems, the present application is proposed. In order to improve the safety and efficiency of drilling operations, according to one aspect of the present application, a drilling overflow and well leakage intelligent monitoring method is provided, which includes:

[0007] A first mass flow meter and a second mass flow meter are arranged at the inlet of the drilling circulation system to respectively collect the mass flow time series data at the inlet and the mass flow time series data at the outlet of the drilling circulation system;

[0008] Transmitting the mass flow rate time series data at the inlet and the mass flow rate time series data at the outlet of the drilling circulation system to the data processing cloud;

[0009] In the data processing cloud, the overflow and leakage conditions of the drilling are monitored based on the time series coordinated changes between the mass flow time series data at the inlet and the mass flow time series data at the outlet, including the steps of: arranging the mass flow time series data at the inlet and the mass flow time series data at the outlet in time series, and obtaining the drilling inlet-outlet mass flow time series difference vector based on the arranged drilling inlet mass flow time series input vector and the drilling outlet mass flow time series input vector; The inlet-outlet mass flow time difference vectors are respectively time-encoded to obtain the drilling inlet time characteristics, the drilling outlet time characteristics and the drilling inlet-outlet time difference characteristics; the drilling inlet-outlet time difference characteristics are respectively matched and interacted with the drilling inlet time characteristics and the drilling outlet time characteristics to obtain the drilling inlet influence difference change representation and the drilling outlet influence difference change representation; based on the energy synergistic response aggregation between the drilling inlet influence difference change representation and the drilling outlet influence difference change representation, the multi-modal interactive aggregation representation of the drilling inlet and outlet mass flow is obtained to obtain the monitoring result.

[0010] In the above-mentioned drilling overflow and lost circulation intelligent monitoring method, the mass flow time series data at the inlet and the mass flow time series data at the outlet are arranged in time series, and based on the arranged drilling inlet mass flow time series input vector and the drilling outlet mass flow time series input vector, a drilling inlet-outlet mass flow time series difference vector is obtained, including:

[0011] Arranging the mass flow time series data at the inlet and the mass flow time series data at the outlet according to the time dimension to obtain the drilling inlet mass flow time series input vector and the drilling outlet mass flow time series input vector;

[0012] The position-by-position difference between the wellbore inlet mass flow timing input vector and the wellbore outlet mass flow timing input vector is calculated to obtain the wellbore inlet-outlet mass flow timing difference vector.

[0013] In the above-mentioned drilling overflow and lost circulation intelligent monitoring method, the drilling inlet mass flow timing input vector, the drilling outlet mass flow timing input vector and the drilling inlet-outlet mass flow timing difference vector are respectively time-coded to obtain the drilling inlet timing feature, the drilling outlet timing feature and the drilling inlet-outlet timing difference feature, including:

[0014] The drilling inlet mass flow timing input vector, the drilling outlet mass flow timing input vector and the drilling inlet-outlet mass flow timing difference vector are respectively passed through a timing feature extractor based on a Bi-LSTM model to obtain a drilling inlet mass flow timing feature vector as the drilling inlet timing feature, a drilling outlet mass flow timing feature vector as the drilling outlet timing feature and a drilling inlet-outlet mass flow timing difference feature vector as the drilling inlet-outlet timing difference feature.

[0015] In the above-mentioned drilling overflow and lost circulation intelligent monitoring method, the drilling entry-exit time series difference feature is respectively subjected to drilling feature fine-grained matching interaction with the drilling entry time series feature and the drilling exit time series feature to obtain a drilling entry influence difference change representation and a drilling exit influence difference change representation, including:

[0016] Performing feature fine-grained optimal matching characterization on the drilling inlet mass flow time series feature vector and the drilling inlet-outlet mass flow time series difference feature vector to obtain a drilling inlet mass flow impact difference change characterization vector as the drilling inlet impact difference change characterization;

[0017] Performing feature fine-grained optimal matching characterization on the well outlet mass flow time series feature vector and the well inlet-outlet mass flow time series difference feature vector to obtain a well outlet mass flow impact difference change characterization vector as the well outlet impact difference change characterization;

[0018] Among them, the drilling inlet mass flow time series feature vector and the drilling inlet-outlet mass flow time series difference feature vector are characterized by fine-grained optimal matching to obtain a drilling inlet mass flow impact difference change characterization vector, including:

[0019] Based on a predetermined scale, fine-grained feature decoupling is performed on the drilling inlet mass flow time series feature vector and the drilling inlet-outlet mass flow time series difference feature vector to obtain a set of drilling inlet mass flow local time series feature vectors and a set of drilling inlet-outlet mass flow local time series feature vectors;

[0020] Calculate the hyperbolic space distance metric factor between any two local time series feature vectors of the well inlet mass flow and the local time series feature vectors of the well inlet-outlet mass flow in the set of the well inlet mass flow local time series feature vectors and the set of the well inlet-outlet mass flow local time series feature vectors;

[0021] Based on the hyperbolic space distance metric factor, the set of the local time series feature vectors of the drilling inlet mass flow and the set of the local time series feature vectors of the drilling inlet-outlet mass flow are screened for optimal feature pairing to obtain the set of optimal pairing of the local time series feature vectors of the drilling inlet mass flow and the local time series feature vectors of the drilling inlet-outlet mass flow;

[0022] Pairing each optimal drilling inlet mass flow local time series feature vector and drilling inlet-outlet mass flow local time series feature vector in the set of paired optimal drilling inlet mass flow local time series feature vector and drilling inlet-outlet mass flow local time series feature vector into a feature multi-dimensional interaction module to obtain a set of drilling inlet mass flow time series feature paired interaction fusion feature vectors;

[0023] The set of the paired interactive fusion feature vectors of the drilling inlet mass flow time series features is cascaded to obtain a characterization vector of the drilling inlet mass flow impact difference change.

[0024] In the above-mentioned drilling overflow and lost circulation intelligent monitoring method, based on the energy synergistic response aggregation between the drilling inlet influence difference change representation and the drilling outlet influence difference change representation, the multi-modal interactive aggregation representation of the drilling inlet and outlet mass flow is obtained to obtain the monitoring results, including:

[0025] Performing feature interactive aggregation based on energy collaborative response on the drilling inlet mass flow rate impact difference change characterization vector and the drilling outlet mass flow rate impact difference change characterization vector to obtain a drilling inlet and outlet mass flow rate multimodal interactive aggregation characterization vector as the drilling inlet and outlet mass flow rate multimodal interactive aggregation characterization;

[0026] The monitoring result is obtained based on the multi-modal interactive aggregation characterization of the mass flow rate at the inlet and outlet of the drilling well, and the monitoring result is used to indicate whether overflow and well leakage occur in the drilling well.

[0027] In the above-mentioned drilling overflow and well leakage intelligent monitoring method, the drilling inlet mass flow rate influence difference change characterization vector and the drilling outlet mass flow rate influence difference change characterization vector are subjected to feature interactive aggregation based on energy synergistic response to obtain the drilling inlet and outlet mass flow multimodal interactive aggregation characterization vector, including:

[0028] Calculate the positional division between the drilling inlet mass flow rate impact difference change characterization vector and the drilling outlet mass flow rate impact difference change characterization vector to obtain the drilling inlet and outlet mass flow rate change response characterization vector;

[0029] Calculating the energy synergistic response value of each position in the drilling inlet and outlet mass flow rate change response characterization vector to obtain the drilling inlet and outlet mass flow rate energy synergistic response value vector;

[0030] Performing normalization processing based on a sigmoid function on the drilling inlet and outlet mass flow energy coordinated response value vector to obtain a drilling inlet and outlet mass flow energy coordinated response value weight value vector;

[0031] Based on the drilling inlet and outlet mass flow energy coordinated response value weight value vector, the drilling inlet mass flow impact difference change characterization vector and the drilling outlet mass flow impact difference change characterization vector are interactively aggregated to obtain the drilling inlet and outlet mass flow multimodal interactive aggregation characterization vector.

[0032] In the above-mentioned drilling overflow and lost circulation intelligent monitoring method, the energy synergistic response value of each position in the drilling inlet and outlet mass flow change response characterization vector is calculated to obtain the drilling inlet and outlet mass flow energy synergistic response value vector, including:

[0033] Respectively calculating the mean and variance of the drilling inlet and outlet mass flow rate change response characterization vector to obtain the drilling inlet and outlet mass flow rate change mean and the drilling inlet and outlet mass flow rate change variance;

[0034] The value obtained by adding the variance of the mass flow rate change at the well inlet and outlet and the preset hyperparameter is multiplied by four to obtain a first well inlet and outlet mass flow rate energy coordinated response value;

[0035] Calculate the square of each value after subtracting the drilling inlet and outlet mass flow rate change response characterization vector from the drilling inlet and outlet mass flow rate change mean value by position to obtain the drilling inlet and outlet mass flow rate change difference value vector;

[0036] The modulated drilling inlet and outlet mass flow rate change variance obtained by multiplying the drilling inlet and outlet mass flow rate change variance by two, the value obtained by multiplying the preset hyperparameter by two, and the drilling inlet and outlet mass flow rate change difference value vector are added by position to obtain a second drilling inlet and outlet mass flow rate energy coordinated response value vector;

[0037] The first wellbore inlet and outlet mass flow energy coordinated response value is divided by the second wellbore inlet and outlet mass flow energy coordinated response value vector by position to obtain the wellbore inlet and outlet mass flow energy coordinated response value vector.

[0038] In the above-mentioned drilling overflow and well leakage intelligent monitoring method, based on the drilling inlet and outlet mass flow energy coordinated response value weight value vector, the drilling inlet mass flow impact difference change characterization vector and the drilling outlet mass flow impact difference change characterization vector are interactively aggregated to obtain the drilling inlet and outlet mass flow multi-modal interactive aggregation characterization vector, including:

[0039] Multiplying the drilling inlet and outlet mass flow energy coordinated response value weight value vector by the drilling inlet mass flow impact difference change characterization vector according to the position point to obtain the drilling inlet mass flow impact difference change characterization weighted vector;

[0040] Subtract the constant 1 from the weight value vector of the energy coordinated response value of the mass flow rate at the inlet and outlet of the well by position, and multiply the vector after the subtraction by the vector representing the difference change in the mass flow rate at the outlet of the well by position to obtain a weighted vector representing the difference change in the mass flow rate at the outlet of the well;

[0041] The weighted vector representing the difference in impact of the well inlet mass flow and the weighted vector representing the difference in impact of the well outlet mass flow are added by position to obtain the multi-modal interaction aggregation representation vector of the well inlet and outlet mass flow.

[0042] In the above-mentioned drilling overflow and well leakage intelligent monitoring method, the monitoring result is obtained based on the multimodal interactive aggregation characterization of the drilling inlet and outlet mass flow, which is used to: pass the multimodal interactive aggregation characterization vector of the drilling inlet and outlet mass flow through a classifier-based drilling overflow and well leakage monitor to obtain the monitoring result.

[0043] According to another aspect of the present application, a drilling overflow and lost circulation intelligent monitoring system is provided, which comprises:

[0044] A mass flow time series data acquisition module is used to set a first mass flow meter and a second mass flow meter at the inlet of the drilling circulation system to respectively acquire the mass flow time series data at the inlet and the mass flow time series data at the outlet of the drilling circulation system;

[0045] A mass flow time series data transmission module, used to transmit the mass flow time series data at the inlet and the mass flow time series data at the outlet of the drilling circulation system to a data processing cloud;

[0046] The drilling overflow and well leakage monitoring module is used to monitor the overflow and well leakage of the drilling on the data processing cloud based on the time series coordinated changes between the mass flow time series data at the inlet and the mass flow time series data at the outlet, wherein the drilling overflow and well leakage monitoring module includes: a mass flow time series data arrangement unit, which is used to arrange the mass flow time series data at the inlet and the mass flow time series data at the outlet in time series, and obtain the drilling inlet-outlet mass flow time series difference vector based on the arranged drilling inlet mass flow time series input vector and the drilling outlet mass flow time series input vector; a drilling mass flow feature time series encoding unit, which is used to arrange the drilling inlet mass flow time series input vector and the drilling outlet mass flow time series input vector; The flow rate time series input vector and the drilling inlet-outlet mass flow time series difference vector are respectively time-coded to obtain the drilling inlet time series feature, the drilling outlet time series feature and the drilling inlet-outlet time series difference feature; a drilling feature fine-grained matching interaction unit is used to perform drilling feature fine-grained matching interaction on the drilling inlet-outlet time series difference feature with the drilling inlet time series feature and the drilling outlet time series feature respectively to obtain the drilling inlet influence difference change representation and the drilling outlet influence difference change representation; an energy synergistic response aggregation unit is used to obtain the drilling inlet and outlet mass flow multi-modal interactive aggregation representation based on the energy synergistic response aggregation between the drilling inlet influence difference change representation and the drilling outlet influence difference change representation to obtain the monitoring result.

[0047] Compared with the prior art, the present application provides an intelligent monitoring system and method for drilling overflow and well leakage, which first collects the mass flow at the inlet and outlet of the drilling circulation system and transmits it to the data processing cloud, and then uses artificial intelligence-based data analysis and processing algorithms to extract features from the collected data, so as to monitor the overflow and well leakage of the drilling by utilizing the time series coordinated changes between the mass flow time series data at the inlet and the mass flow time series data at the outlet. In this way, by comprehensively analyzing the time series characteristics of the inlet and outlet mass flow rates, the dynamic correlation of flow changes during the drilling process can be fully captured, and the monitoring coverage can be improved, thereby improving the intelligence of drilling overflow and well leakage monitoring, so as to improve the safety and efficiency of drilling operations. BRIEF DESCRIPTION OF THE DRAWINGS

[0048] By describing the embodiments of the present application in more detail in conjunction with the accompanying drawings, the above and other purposes, features and advantages of the present application will become more apparent. The accompanying drawings are used to provide a further understanding of the embodiments of the present application and constitute a part of the specification. Together with the embodiments of the present application, they are used to explain the present application and do not constitute a limitation of the present application. In the accompanying drawings, the same reference numerals generally represent the same components or steps.

[0049] Figure 1It is a flow chart of the intelligent monitoring method of drilling overflow and well leakage according to an embodiment of the present application.

[0050] Figure 2 This is a flowchart for monitoring the overflow and leakage conditions of the drilling based on the coordinated timing changes between the mass flow timing data at the inlet and the mass flow timing data at the outlet in the data processing cloud in the intelligent monitoring method for drilling overflow and leakage according to an embodiment of the present application.

[0051] Figure 3 This is a data flow diagram for monitoring the overflow and leakage conditions of the drilling based on the coordinated timing changes between the mass flow timing data at the inlet and the mass flow timing data at the outlet in the data processing cloud in the intelligent monitoring method for drilling overflow and leakage according to an embodiment of the present application.

[0052] Figure 4 A flowchart of the drilling entrance-exit timing differential feature in the intelligent monitoring method for drilling overflow and well leakage according to an embodiment of the present application is provided for performing fine-grained matching and interaction of drilling features with the drilling entrance timing feature and the drilling exit timing feature to obtain a drilling entrance influence difference change representation and a drilling exit influence difference change representation.

[0053] Figure 5 It is a system block diagram of the drilling overflow and well leakage intelligent monitoring system according to an embodiment of the present application. DETAILED DESCRIPTION

[0054] Various exemplary embodiments, features and aspects of the present application will be described in detail below with reference to the accompanying drawings. The same reference numerals in the accompanying drawings represent elements with the same or similar functions. Although various aspects of the embodiments are shown in the accompanying drawings, the drawings are not necessarily drawn to scale unless otherwise specified.

[0055] The word “exemplary” is used exclusively herein to mean “serving as an example, example, or illustration.” Any embodiment described herein as “exemplary” is not necessarily to be construed as preferred or advantageous over other embodiments.

[0056] In addition, in order to better illustrate the present application, numerous specific details are given in the following specific embodiments. It should be understood by those skilled in the art that the present application can also be implemented without certain specific details. In some examples, methods, means, components and circuits well known to those skilled in the art are not described in detail in order to highlight the subject matter of the present application.

[0057] In addition, the terms "first" and "second" are used for descriptive purposes only and should not be understood as indicating or implying relative importance or implicitly indicating the number of the indicated technical features. Thus, a feature defined as "first" or "second" may explicitly or implicitly include one or more of the features. In the description of this application, the meaning of "plurality" is two or more, unless otherwise clearly and specifically defined.

[0058] Drilling refers to the process of drilling holes on the ground or underwater by mechanical or other technical means to reach deep into the earth to obtain underground resources (such as oil, natural gas, minerals, etc.) or conduct geological exploration. During the drilling process, you may face a variety of challenges, among which the most common problems include overflow and well leakage. Traditional drilling overflow and well leakage monitoring methods ignore the temporal relationship between inlet and outlet flow rates and cannot effectively capture the dynamic correlation of flow changes during the drilling process. This will increase the risk of false alarms and missed alarms, and it is difficult to fully reflect the complex changes in the drilling process, reducing the accuracy and reliability of the system in real-time response and future trend prediction.

[0059] In response to the above technical problems, the present application proposes an intelligent monitoring method for drilling overflow and well leakage. Figure 1 Flow chart of the intelligent monitoring method for drilling overflow and well leakage according to an embodiment of the present application. Figure 1 As shown, according to the embodiment of the present application, the intelligent monitoring method for drilling overflow and well leakage includes: S110, setting a first mass flow meter and a second mass flow meter at the inlet of the drilling circulation system to respectively collect the mass flow timing data at the inlet and the mass flow timing data at the outlet of the drilling circulation system; S120, transmitting the mass flow timing data at the inlet and the mass flow timing data at the outlet of the drilling circulation system to the data processing cloud; S130, in the data processing cloud, monitoring the overflow and well leakage of the drilling based on the timing coordinated changes between the mass flow timing data at the inlet and the mass flow timing data at the outlet.

[0060] In the above-mentioned drilling overflow and well leakage intelligent monitoring method, step S110, a first mass flowmeter and a second mass flowmeter are set at the inlet of the drilling circulation system to respectively collect the mass flow time series data at the inlet and the mass flow time series data at the outlet of the drilling circulation system. It should be understood that by monitoring the flow at the inlet and outlet, a flow balance analysis can be performed to determine whether the flow of the fluid in the system is normal. Among them, the first mass flowmeter is usually installed at the inlet of the drilling circulation system, and this position allows it to monitor the mass flow of the drilling fluid delivered to the wellhead from the storage tank or mud pump. The second mass flowmeter is set at the outlet of the drilling circulation system to monitor the drilling fluid flow returned from the wellhead. Specifically, the flowmeter should be installed on a horizontal pipeline, and the fluid flow of the horizontal pipeline is smooth to avoid flow fluctuations caused by pipeline elbows or valves. In addition, the flowmeter should be connected to the control system to transmit data in real time for analysis.

[0061] In the above-mentioned drilling overflow and well leakage intelligent monitoring method, step S120, the mass flow time series data at the inlet and the mass flow time series data at the outlet of the drilling circulation system are transmitted to the data processing cloud. It should be understood that in the drilling circulation system, the mass flow time series data at the inlet usually includes the flow rate, pressure and density parameters of the fluid, while the mass flow time series data at the outlet can reflect the changes in the fluid after the drilling operation. By collecting these two data points in real time, comprehensive monitoring of the fluid flow can be achieved, so that abnormal conditions can be discovered in time during the operation process and the risk of accidents can be reduced. Furthermore, the cloud platform has powerful data storage and processing capabilities. By transmitting data to the data processing cloud, the cloud platform can deeply explore the potential patterns in the data, which helps to discover abnormal conditions in time. Specifically, the mass flow data at the inlet and outlet are uploaded to the cloud in real time via a wireless network.

[0062] In the above-mentioned drilling overflow and well leakage intelligent monitoring method, step S130, in the data processing cloud, the overflow and well leakage of the drilling are monitored based on the time series coordinated changes between the mass flow time series data at the inlet and the mass flow time series data at the outlet. It should be understood that due to the law of conservation of flow, that is, under ideal conditions, the amount of fluid entering the wellbore should be equal to the amount of fluid flowing out of the wellbore. When overflow or well leakage occurs, this balance relationship will be broken, and the change in flow will reflect potential safety hazards. Therefore, in the technical solution of the present application, dynamic monitoring is achieved by analyzing the time series coordinated changes in flow data. Specifically, if the flow at the outlet is significantly lower than the flow at the inlet, it may indicate the presence of well leakage; on the contrary, if the outlet flow significantly exceeds the inlet flow, an overflow may occur.

[0063] Accordingly, in monitoring the overflow and leakage of drilling based on the time series coordinated changes between the mass flow time series data at the inlet and the mass flow time series data at the outlet, the technical concept of the present application is to use artificial intelligence-based data analysis and processing algorithms to perform time series arrangement and differential calculation of the mass flow at the inlet and outlet, and to perform time series encoding and fine-grained matching interaction on the inlet, outlet and differential mass flow, so as to obtain the time series characterization information of the three. Then, the interaction semantics between the drilling inlet mass flow time series characteristics and the drilling outlet mass flow time series characteristics and the inlet and outlet mass flow difference time series characteristics are calculated, respectively, to characterize the outlet and inlet flow time series characteristics under the influence of the time series changes of the drilling inlet and outlet mass flow difference, and then infer the time series delay response characterizing this drilling mass flow through the interaction response aggregation information between the inlet mass flow change characterization and the outlet mass flow change characterization after feature optimization expression, so as to intelligently judge whether the drilling has overflow and leakage. In this way, by comprehensively analyzing the timing characteristics of the inlet and outlet mass flow rates, the dynamic correlation of flow changes during the drilling process can be fully captured, the monitoring coverage can be improved, and the intelligence of drilling overflow and well leakage monitoring can be enhanced to improve the safety and efficiency of drilling operations.

[0064] Figure 2 This is a flowchart for monitoring the overflow and leakage conditions of the drilling based on the coordinated timing changes between the mass flow timing data at the inlet and the mass flow timing data at the outlet in the data processing cloud in the intelligent monitoring method for drilling overflow and leakage according to an embodiment of the present application. Figure 3 The data flow diagram of the intelligent monitoring method for drilling overflow and well leakage according to the embodiment of the present application is a data processing cloud, which monitors the overflow and well leakage of the drilling based on the time series coordinated changes between the mass flow time series data at the inlet and the mass flow time series data at the outlet. Figure 2 and Figure 3As shown, in step S130, it includes: S131, time-series arranging the mass flow time series data at the inlet and the mass flow time series data at the outlet, and obtaining the drilling inlet-outlet mass flow time series difference vector based on the arranged drilling inlet mass flow time series input vector and the drilling outlet mass flow time series input vector; S132, time-series encoding the drilling inlet mass flow time series input vector, the drilling outlet mass flow time series input vector and the drilling inlet-outlet mass flow time series difference vector respectively to obtain the drilling inlet time series. The drilling entrance-exit timing characteristics, the drilling exit timing characteristics and the drilling entrance-exit timing difference characteristics; S133, the drilling entrance-exit timing difference characteristics are respectively matched and interacted with the drilling entrance timing characteristics and the drilling exit timing characteristics to obtain the drilling entrance influence difference change representation and the drilling exit influence difference change representation; S134, based on the energy synergistic response aggregation between the drilling entrance influence difference change representation and the drilling exit influence difference change representation, a multi-modal interactive aggregation representation of the drilling entrance and exit mass flow is obtained to obtain the monitoring result.

[0065] Specifically, in step S131, the mass flow timing data at the inlet and the mass flow timing data at the outlet are arranged in time series, and based on the arranged drilling inlet mass flow timing input vector and the drilling outlet mass flow timing input vector, a drilling inlet-outlet mass flow timing difference vector is obtained, including: arranging the mass flow timing data at the inlet and the mass flow timing data at the outlet according to the time dimension to obtain the drilling inlet mass flow timing input vector and the drilling outlet mass flow timing input vector; calculating the position-by-position difference between the drilling inlet mass flow timing input vector and the drilling outlet mass flow timing input vector to obtain the drilling inlet-outlet mass flow timing difference vector.

[0066] It should be understood that, considering that during the drilling process, the mass flow data is continuously generated over time, that is, the mass flow time series data at the inlet and the mass flow time series data at the outlet both have time series characteristics in the time dimension. Based on this, in order to better reflect the time correlation of these data, the mass flow time series data at the inlet and the mass flow time series data at the outlet are arranged according to the time dimension to obtain the drilling inlet mass flow time series input vector and the drilling outlet mass flow time series input vector. In this way, the trend, periodic pattern, seasonal fluctuation and other characteristics of the drilling inlet and outlet mass flow data changing over time during the drilling process can be better captured.

[0067] Next, in order to be able to more intuitively see the difference between the mass flow rates at the inlet and outlet. That is, if the inlet flow rate is significantly greater than the outlet flow rate in a certain period of time, this may mean that liquid has entered the drilling circulation system (overflow), otherwise it may be that the liquid has leaked into the formation (well leakage). The position-by-position difference between the drilling inlet mass flow timing input vector and the drilling outlet mass flow timing input vector is calculated to capture the timing changes more carefully and subtly, and the drilling inlet-outlet mass flow timing difference vector is obtained.

[0068] Then, considering that the drilling inlet mass flow timing input vector, the drilling outlet mass flow timing input vector and the drilling inlet-outlet mass flow timing difference vector have different timing feature representations in different time periods, therefore, in step S132 of the technical solution of the present application, the drilling inlet mass flow timing input vector, the drilling outlet mass flow timing input vector and the drilling inlet-outlet mass flow timing difference vector are respectively time-encoded to obtain the drilling inlet timing characteristics, the drilling outlet timing characteristics and the drilling inlet-outlet timing difference characteristics. Specifically, the drilling inlet mass flow timing input vector, the drilling outlet mass flow timing input vector and the drilling inlet-outlet mass flow timing difference vector are respectively passed through a timing feature extractor based on a Bi-LSTM model to mine and refine the mass flow timing changes between different time spans, and the drilling inlet mass flow timing feature vector is obtained as the drilling inlet timing feature, the drilling outlet mass flow timing feature vector is obtained as the drilling outlet timing feature and the drilling inlet-outlet mass flow timing difference feature vector is obtained as the drilling inlet-outlet timing feature.

[0069] Figure 4 The present invention is a flow chart of performing fine-grained matching interaction of the drilling entrance-exit time series difference feature with the drilling entrance time series feature and the drilling exit time series feature in the drilling overflow and well leakage intelligent monitoring method according to the embodiment of the present application to obtain the drilling entrance influence difference change representation and the drilling exit influence difference change representation. Figure 4As shown, in step S133, the drilling inlet-outlet timing differential feature is subjected to drilling feature fine-grained matching interaction with the drilling inlet timing feature and the drilling outlet timing feature respectively to obtain a drilling inlet impact difference change representation and a drilling outlet impact difference change representation, including: S1331, performing feature fine-grained optimal matching representation on the drilling inlet mass flow timing feature vector and the drilling inlet-outlet mass flow timing differential feature vector to obtain a drilling inlet mass flow impact difference change representation vector as the drilling inlet impact difference change representation; S1332, performing feature fine-grained optimal matching representation on the drilling outlet mass flow timing feature vector and the drilling inlet-outlet mass flow timing differential feature vector to obtain a drilling outlet mass flow impact difference change representation vector as the drilling outlet impact difference change representation.

[0070] More specifically, it is considered that the drilling inlet mass flow time series feature vector reflects the change of the inlet flow over time. The drilling inlet-outlet mass flow time series difference feature vector emphasizes the time series difference change between the inlet and outlet flows. Therefore, in order to integrate information from two different perspectives to obtain a more comprehensive and detailed feature representation, in step S1331 of the technical solution of the present application, the drilling inlet mass flow time series feature vector and the drilling inlet-outlet mass flow time series difference feature vector are characterized by fine-grained optimal matching to obtain the drilling inlet mass flow impact difference change characterization vector as the drilling inlet impact difference change characterization.

[0071] More specifically, in step S1331, the drilling inlet mass flow time series feature vector and the drilling inlet-outlet mass flow time series difference feature vector are characterized by fine-grained optimal matching to obtain a drilling inlet mass flow impact difference change characterization vector, which can be expressed by the formula: ; ; ; ; ; ;in, is the time series characteristic vector of the drilling inlet mass flow rate, are the first, second,...,th in the set of local time series feature vectors of the mass flow rate at the drilling entrance. ,..., The local time series characteristic vector of the mass flow rate at the drilling entrance, is the time series difference characteristic vector of the well inlet-outlet mass flow rate, are the first, second,...,th in the set of local time series feature vectors of the drilling inlet-outlet mass flow rate. ,..., The local time series characteristic vector of the mass flow rate from the well inlet to the outlet, Decouple operations for fine-grained features, For calculation and The hyperbolic space distance metric factor between To return the minimum value value, is the minimum approximate matching value, To calculate the square of the one-norm of a vector, is the inverse hyperbolic cosine function, , and They are position difference, position dot product and position addition respectively. , and is the corresponding weighting parameter, is the first in the set of local time series characteristic vectors of the drilling inlet-outlet mass flow rate The local time series characteristic vector of the mass flow rate from the well inlet to the outlet, The set of the interactive fusion feature vectors of the drilling inlet mass flow time series feature pairing The time series features of the mass flow rate at the drilling entrance are paired and interactively fused into feature vectors. , and are the first, second and third pairs of the interactive fusion feature vectors of the drilling inlet mass flow time series feature pairing. The time series features of the mass flow rate at the drilling entrance are paired and interactively fused into feature vectors. For cascade processing, is a vector representing the difference in the impact of the drilling inlet mass flow rate.

[0072] More specifically, step S1331, performs fine-grained optimal matching characterization on the drilling inlet mass flow timing feature vector and the drilling inlet-outlet mass flow timing difference feature vector to obtain a drilling inlet mass flow impact difference change characterization vector, including: performing fine-grained feature decoupling on the drilling inlet mass flow timing feature vector and the drilling inlet-outlet mass flow timing difference feature vector based on a predetermined scale to obtain a set of drilling inlet mass flow local timing feature vectors and a set of drilling inlet-outlet mass flow local timing feature vectors; calculating a hyperbolic space distance measurement factor between any two drilling inlet mass flow local timing feature vectors and drilling inlet-outlet mass flow local timing feature vectors in the set of drilling inlet mass flow local timing feature vectors and the set of drilling inlet-outlet mass flow local timing feature vectors; based on the hyperbolic space distance measurement factor, the drilling inlet mass flow local timing feature vector is used to measure the distance between any two drilling inlet mass flow local timing feature vectors and the drilling inlet-outlet mass flow local timing feature vectors. The curved space distance measurement factor is used to perform optimal feature pairing screening on the set of local time series feature vectors of the drilling inlet mass flow and the set of local time series feature vectors of the drilling inlet-outlet mass flow to obtain the set of optimal drilling inlet mass flow local time series feature vectors and drilling inlet-outlet mass flow local time series feature vector pairs; each optimal drilling inlet mass flow local time series feature vector and drilling inlet-outlet mass flow local time series feature vector pair in the set of optimal drilling inlet mass flow local time series feature vector pairs is input into the feature multi-dimensional interaction module to obtain a set of drilling inlet mass flow time series feature pairing interactive fusion feature vectors; the set of drilling inlet mass flow time series feature pairing interactive fusion feature vectors is cascaded to obtain the drilling inlet mass flow impact difference change characterization vector.

[0073] It should be understood that in step S1331, the drilling inlet mass flow timing feature vector and the drilling inlet-outlet mass flow timing difference feature vector are firstly subjected to fine-grained feature decoupling to decompose the timing features of the entire time domain into a set of multiple drilling inlet mass flow local timing feature vectors containing multiple local times and a set of drilling inlet-outlet mass flow local timing feature vectors. Next, the hyperbolic space distance metric factor between any drilling inlet mass flow local timing feature vector and drilling inlet-outlet mass flow local timing feature vector in the set is calculated. That is, the hyperbolic space is a non-Euclidean geometric space that can better quantify the similarity or difference between the two feature vectors. Then, based on the hyperbolic space distance metric factor, the two sets are optimally paired to obtain the optimal drilling inlet mass flow local timing feature vector and the drilling inlet-outlet mass flow local timing feature vector pairing set. That is, based on the metric factor, the pair of feature vectors with the closest distance in the hyperbolic space is found to ensure that each pair of matching pairs is the most representative of similarity. Afterwards, in order to strengthen the correlation and complex interaction between each matching pair, they are input into the feature multi-dimensional interaction module to obtain a set of drilling inlet mass flow rate time series feature pairing interaction fusion feature vectors. Finally, these fused feature vectors are cascaded to obtain the drilling inlet mass flow rate impact difference change representation vector to obtain a comprehensive feature representation, which contains not only the change information of the flow rate itself, but also the change information relative to the outlet flow rate.

[0074] Specifically, the drilling multi-parameter hyperbolic space distance metric factor between any two local time series feature vectors of drilling inlet mass flow and the local time series feature vectors of drilling inlet-outlet mass flow in the set of local time series feature vectors of drilling inlet mass flow and the set of local time series feature vectors of drilling inlet-outlet mass flow is calculated, including:

[0075] Extracting a predetermined drilling inlet mass flow local time series feature vector from the set of drilling inlet mass flow local time series feature vectors;

[0076] Calculating the hyperbolic space distance measurement factor between the predetermined drilling inlet mass flow local time series feature vector and each drilling inlet-outlet mass flow local time series feature vector in the set of the drilling inlet-outlet mass flow local time series feature vector to obtain a set of drilling multi-parameter hyperbolic space distance measurement factors of the predetermined drilling inlet mass flow local time series feature vector, including: calculating the position difference between the predetermined drilling inlet mass flow local time series feature vector and the drilling inlet-outlet mass flow local time series feature vector, and calculating the square of the norm of the obtained difference vector to obtain the drilling multi-parameter difference norm value; calculating the constant one and the predetermined drilling inlet mass flow local time series feature vector The method comprises the following steps: calculating the difference between the square of the one-norm of the quantity to obtain the drilling inlet mass flow timing difference; calculating the difference between the constant one and the square of the one-norm of the local timing characteristic vector of the drilling inlet-outlet mass flow to obtain the drilling inlet-outlet mass flow timing difference; calculating the product between the drilling inlet mass flow timing difference and the drilling inlet-outlet mass flow timing difference to obtain the product value; dividing the drilling inlet mass flow timing difference by the product value, multiplying the obtained value by the constant two and then adding it to the constant one to obtain the drilling multi-parameter distance measurement value; inputting the drilling multi-parameter distance measurement value into the arccosh function to obtain the drilling multi-parameter hyperbolic space distance measurement factor.

[0077] Specifically, based on the drilling multi-parameter hyperbolic space distance measurement factor, the set of the drilling inlet mass flow local timing feature vectors and the set of the drilling inlet-outlet mass flow local timing feature vectors are screened for optimal feature pairing to obtain the set of optimal drilling inlet mass flow local timing feature vectors and drilling inlet-outlet mass flow local timing feature vector pairings, including: taking the drilling inlet mass flow local timing feature vector and the drilling inlet-outlet mass flow local timing feature vector corresponding to the smallest drilling multi-parameter hyperbolic space distance measurement factor in the set of the drilling inlet-outlet mass flow local timing feature vectors as the optimal drilling inlet mass flow local timing feature vector and drilling inlet-outlet mass flow local timing feature vector pairing.

[0078] Specifically, it is considered that the drilling outlet mass flow time series feature vector provides information about the change of the outlet flow over time. The drilling inlet-outlet mass flow time series difference feature vector reflects the difference between the inlet and outlet flows. Therefore, in order to more accurately capture and characterize the characteristics of the change of the outlet flow during the drilling process, and combine the differential characteristics to enhance the detection capability of abnormal conditions (such as overflow or well leakage), in step S1332 of the technical solution of the present application, the drilling outlet mass flow time series feature vector and the drilling inlet-outlet mass flow time series difference feature vector are characterized by fine-grained optimal matching to obtain the drilling outlet mass flow impact difference change characterization vector as the drilling outlet impact difference change characterization.

[0079] Specifically, step S134, based on the multimodal interactive aggregation representation of the drilling inlet and outlet mass flow obtained by energy synergistic response aggregation between the drilling inlet influence difference change representation and the drilling outlet influence difference change representation, obtains a monitoring result, including: performing feature interactive aggregation based on energy synergistic response on the drilling inlet mass flow influence difference change representation vector and the drilling outlet mass flow influence difference change representation vector to obtain a drilling inlet and outlet mass flow multimodal interactive aggregation representation vector as the drilling inlet and outlet mass flow multimodal interactive aggregation representation; based on the drilling inlet and outlet mass flow multimodal interactive aggregation representation, obtains the monitoring result, and the monitoring result is used to indicate whether overflow and well leakage occur in the drilling.

[0080] More specifically, it is considered that there are different degrees of response interaction between the characterization of the difference change in the impact of the drilling inlet mass flow and the characterization of the difference change in the impact of the drilling outlet mass flow at different time points. If overflow occurs, the fluid in the formation may enter the drilling fluid, resulting in a sudden increase in the outlet flow. In this case, even if the inlet flow does not increase significantly, the outlet flow may rise rapidly. Therefore, it is possible to interactively respond to the time series characteristics of the inlet and outlet mass flow at different times, so as to improve the comprehensive characterization of the entire drilling inlet and outlet mass flow, and perform feature interactive aggregation based on energy synergistic response on the drilling inlet mass flow impact difference change characterization vector and the drilling outlet mass flow impact difference change characterization vector to obtain a drilling inlet and outlet mass flow multimodal interactive aggregation characterization vector as the drilling inlet and outlet mass flow multimodal interactive aggregation characterization.

[0081] More specifically, the drilling inlet mass flow rate influence difference change characterization vector and the drilling outlet mass flow rate influence difference change characterization vector are subjected to characteristic interactive aggregation based on energy synergistic response to obtain the drilling inlet and outlet mass flow multimodal interactive aggregation characterization vector, which can be expressed as: ; ; ; ;in, is the characterization vector of the difference in the impact of the drilling inlet mass flow rate, is the characterization vector of the difference in the impact of the mass flow rate at the well outlet, is the characterization vector of the change response of the mass flow rate at the well inlet and outlet, and They are The mean and variance of is a preset hyperparameter, yes The eigenvalues ​​at each position in , is the energy synergistic response value of each position in the energy synergistic response value vector of the mass flow rate at the inlet and outlet of the drilling, yes function, is the characteristic value of each position in the weight value vector of the coordinated response value of the mass flow energy at the inlet and outlet of the drilling, yes The eigenvalues ​​at each position in , yes The eigenvalues ​​at each position in , It is the eigenvalue of each position in the multi-modal interactive aggregation characterization vector of the well inlet and outlet mass flow.

[0082] More specifically, the drilling inlet mass flow impact difference change characterization vector and the drilling outlet mass flow impact difference change characterization vector are subjected to feature interactive aggregation based on energy synergistic response to obtain a drilling inlet and outlet mass flow multimodal interactive aggregation characterization vector, including: calculating the positional division between the drilling inlet mass flow impact difference change characterization vector and the drilling outlet mass flow impact difference change characterization vector to obtain a drilling inlet and outlet mass flow change response characterization vector; calculating the energy synergistic response value of each position in the drilling inlet and outlet mass flow change response characterization vector to obtain a drilling inlet and outlet mass flow energy synergistic response value vector; performing normalization processing based on the sigmoid function on the drilling inlet and outlet mass flow energy synergistic response value vector to obtain a drilling inlet and outlet mass flow energy synergistic response value weight value vector; based on the drilling inlet and outlet mass flow energy synergistic response value weight value vector, interactively aggregate the drilling inlet mass flow impact difference change characterization vector and the drilling outlet mass flow impact difference change characterization vector to obtain the drilling inlet and outlet mass flow multimodal interactive aggregation characterization vector. It should be understood that first, by calculating the positional division between the drilling inlet mass flow rate impact difference change characterization vector and the drilling outlet mass flow rate impact difference change characterization vector, the drilling inlet and outlet mass flow rate change response characterization vector is obtained to understand the inlet and outlet response interaction relationship at different time points in a more detailed manner. After that, the energy synergy response value of each position of the response characterization vector is calculated to obtain the drilling inlet and outlet mass flow rate energy synergy response value vector. Here, the energy synergy response value can quantify the degree of influence of the inlet mass flow rate change at different time points on the outlet mass flow rate change, which helps to understand the dynamic response characteristics of the flow rate change at different time points during the drilling process and provides support for subsequent feature weighting. Then, in order to highlight and strengthen the key response features and suppress unimportant features, the drilling inlet and outlet mass flow rate energy synergy response value vector is input into the sigmoid function for normalization to convert each response value into a weight to obtain the drilling inlet and outlet mass flow rate energy synergy response value weight value vector. In particular, the Sigmoid function can map the original energy synergy response value to a fixed interval [0, 1], thereby achieving standardization and strengthening key and significant features. Finally, based on the weight value vector, the drilling inlet mass flow rate impact difference change characterization vector and the drilling outlet mass flow rate impact difference change characterization vector are interactively aggregated to ensure that those time points that play a key role in the drilling process receive more attention, thereby highlighting the key features and obtaining the drilling inlet and outlet mass flow multimodal interactive aggregation characterization vector.

[0083] More specifically, calculating the energy synergistic response value of each position in the drilling entrance and exit mass flow rate change response characterization vector to obtain the drilling entrance and exit mass flow rate energy synergistic response value vector includes: respectively calculating the mean and variance of the drilling entrance and exit mass flow rate change response characterization vector to obtain the drilling entrance and exit mass flow rate change mean and the drilling entrance and exit mass flow rate change variance; multiplying the value obtained by adding the drilling entrance and exit mass flow rate change variance and a preset hyperparameter by four to obtain a first drilling entrance and exit mass flow rate energy synergistic response value; calculating the drilling entrance and exit mass flow rate change response characterization vector and the drilling entrance and exit mass flow rate change characterization vector. The squares of the values ​​after subtracting the mean of the quantity change by position are obtained to obtain the drilling inlet and outlet mass flow change difference value vector; the modulated drilling inlet and outlet mass flow change variance obtained by multiplying the drilling inlet and outlet mass flow change variance by two, the value obtained by multiplying the preset hyperparameter by two and the drilling inlet and outlet mass flow change difference value vector are added by position to obtain the second drilling inlet and outlet mass flow energy coordinated response value vector; the first drilling inlet and outlet mass flow energy coordinated response value and the second drilling inlet and outlet mass flow energy coordinated response value vector are divided by position to obtain the drilling inlet and outlet mass flow energy coordinated response value vector.

[0084] More specifically, based on the drilling inlet and outlet mass flow energy coordinated response value weight value vector, the drilling inlet mass flow impact difference change characterization vector and the drilling outlet mass flow impact difference change characterization vector are interactively aggregated to obtain the drilling inlet and outlet mass flow multimodal interactive aggregation characterization vector, including: multiplying the drilling inlet and outlet mass flow energy coordinated response value weight value vector and the drilling inlet mass flow impact difference change characterization vector by position point to obtain the drilling inlet mass flow impact difference change characterization weighted vector; subtracting a constant one from the drilling inlet and outlet mass flow energy coordinated response value weight value vector by position point, and multiplying the subtracted vector by the drilling outlet mass flow impact difference change characterization vector by position point to obtain the drilling outlet mass flow impact difference change characterization weighted vector; adding the drilling inlet mass flow impact difference change characterization weighted vector and the drilling outlet mass flow impact difference change characterization weighted vector by position to obtain the drilling inlet and outlet mass flow multimodal interactive aggregation characterization vector.

[0085] More specifically, based on the multimodal interactive aggregation characterization of the drilling inlet and outlet mass flow, the monitoring result is obtained, which is used to: pass the multimodal interactive aggregation characterization vector of the drilling inlet and outlet mass flow through a classifier-based drilling overflow and well leakage monitor to obtain the monitoring result. It should be understood that the multimodal interactive aggregation characterization of the drilling inlet and outlet mass flow obtained by energy synergistic response aggregation using the drilling inlet mass flow impact difference change characterization vector and the drilling outlet mass flow impact difference change characterization vector can be used to intelligently determine whether overflow and well leakage occur in the drilling. In this way, by comprehensively analyzing the time series characteristics of the inlet and outlet mass flow rates, the dynamic correlation of flow changes during the drilling process can be fully captured, and the monitoring coverage can be improved, thereby improving the intelligence of drilling overflow and well leakage monitoring, so as to improve the safety and efficiency of drilling operations.

[0086] Preferably, in an example of the present application, the multimodal interactive aggregation characterization vector of the drilling inlet and outlet mass flow is passed through a classifier-based drilling overflow well leakage monitor to obtain a monitoring result, including: calculating the absolute value sum of each eigenvalue of the multimodal interactive aggregation characterization vector of the drilling inlet and outlet mass flow to obtain a first drilling inlet and outlet mass flow multimodal interactive aggregation spatial structure value, and calculating the square root of the sum of the squares of each eigenvalue of the multimodal interactive aggregation characterization vector of the drilling inlet and outlet mass flow to obtain a second drilling inlet and outlet mass flow multimodal interactive aggregation spatial structure value; Each eigenvalue of the multimodal interaction aggregation characterization vector is multiplied by the first well inlet and outlet mass flow multimodal interaction aggregation spatial structure value and the second well inlet and outlet mass flow multimodal interaction aggregation spatial structure value to obtain the first well inlet and outlet mass flow multimodal interaction aggregation structure reference value and the second well inlet and outlet mass flow multimodal interaction aggregation structure reference value corresponding to each eigenvalue; each eigenvalue of the well inlet and outlet mass flow multimodal interaction aggregation characterization vector is multiplied by the length of the well inlet and outlet mass flow multimodal interaction aggregation characterization vector and the square root of the length to obtain the first well inlet and outlet mass flow multimodal interaction aggregation structure reference value and the second well inlet and outlet mass flow multimodal interaction aggregation structure reference value corresponding to each eigenvalue. The first drilling entrance and exit mass flow multimodal interactive aggregation scale transformation value and the second drilling entrance and exit mass flow multimodal interactive aggregation scale transformation value corresponding to the first drilling entrance and exit mass flow multimodal interactive aggregation structure reference value are divided by the difference between the first drilling entrance and exit mass flow multimodal interactive aggregation spatial structure value and the first drilling entrance and exit mass flow multimodal interactive aggregation scale transformation value to obtain the first drilling entrance and exit mass flow multimodal interactive aggregation transformation adjustment value; the second drilling entrance and exit mass flow multimodal interactive aggregation structure reference value is divided by the second drilling entrance and exit mass flow multimodal interactive aggregation scale transformation value The difference between the spatial structure value and the second drilling inlet and outlet mass flow multimodal interactive aggregation scale transformation value is used to obtain the second drilling inlet and outlet mass flow multimodal interactive aggregation transformation adjustment value; the weighted sum of the first drilling inlet and outlet mass flow multimodal interactive aggregation transformation adjustment value and the second drilling inlet and outlet mass flow multimodal interactive aggregation transformation adjustment value is calculated to obtain each eigenvalue of the optimized drilling inlet and outlet mass flow multimodal interactive aggregation characterization vector; and the optimized drilling inlet and outlet mass flow multimodal interactive aggregation characterization vector is passed through a classifier-based drilling overflow well leakage monitor to obtain a monitoring result.

[0087] Here, the multimodal interactive aggregation representation vector of the well inlet and outlet mass flow rate (denoted as ) is expressed as: ; ; ; ; ; ; ; ;in, Represents the eigenvalues ​​of the multimodal interactive aggregation characterization vector of the mass flow at the well inlet and outlet, It represents the multi-modal interactive aggregation transformation adjustment value of the mass flow rate at the inlet and outlet of the first drilling well, It represents the multi-modal interactive aggregation transformation adjustment value of the mass flow rate at the inlet and outlet of the second drilling well, represents the multimodal interactive aggregation spatial structure value of the mass flow at the inlet and outlet of the first drilling well, represents the multimodal interactive aggregation spatial structure value of the mass flow at the inlet and outlet of the second drilling well, represents the length of the multi-modal interaction aggregation characterization vector of the mass flow at the well inlet and outlet, represents the summation function, represents the absolute value function, represents the set of real numbers, represents a first wellbore inlet and outlet mass flow multimodal interactive aggregation transformation adjustment feature vector obtained by quantizing a plurality of the first wellbore inlet and outlet mass flow multimodal interactive aggregation transformation adjustment values, A second wellbore inlet and outlet mass flow multimodal interactive aggregation transformation adjustment feature vector representing a plurality of second wellbore inlet and outlet mass flow multimodal interactive aggregation transformation adjustment values ​​obtained by vectorizing and arranging the second wellbore inlet and outlet mass flow multimodal interactive aggregation transformation adjustment feature vectors, represents the interactive aggregation transformation adjustment weighted value, It means adding by position. represents positional multiplication, represents the multi-modal interaction aggregation characterization vector of the mass flow at the well inlet and outlet, Represents the optimized drilling inlet and outlet mass flow multi-modal interaction aggregation characterization vector.

[0088] That is, the present application takes into account that the drilling inlet mass flow rate impact difference change characterization vector and the drilling outlet mass flow rate impact difference change characterization vector respectively represent the temporal fine-grained fusion characteristics of the inlet mass flow time series characteristics and the outlet mass flow time series characteristics relative to the inlet-outlet mass flow difference time series characteristics. When performing feature interactive aggregation based on energy synergistic response, the uneven temporal distribution of the inlet-outlet mass flow difference in the time series direction will also make the temporal fusion representation of the multimodal interactive aggregation characterization vector of the drilling inlet and outlet mass flow have a complex aggregation spatial structure based on different local energy synergistic response distribution structures, thereby expecting to improve its classification regression convergence and generalization effects under the complex aggregation spatial structure.

[0089] Therefore, in the preferred example, with respect to the spatial structure information of the feature set of the multimodal interactive aggregation characterization vector of the drilling entrance and exit mass flow in the high-dimensional space, the scale-based frame transformation of each eigenvalue of the multimodal interactive aggregation characterization vector of the drilling entrance and exit mass flow is performed by taking the class-norm space structured representation of the multimodal interactive aggregation characterization vector of the drilling entrance and exit mass flow as a reference window, and the spatial structure-based frame attention weight adjustment of each eigenvalue of the multimodal interactive aggregation characterization vector of the drilling entrance and exit mass flow is implemented to ensure the invariance of the spatial transformation (translation, scaling and rotation) of the multimodal interactive aggregation characterization vector of the drilling entrance and exit mass flow under the feature space interaction, thereby improving the convergence and generalization effect of the classification regression of the feature set of the multimodal interactive aggregation characterization vector of the drilling entrance and exit mass flow under the complex spatial structure representation, and improving the accuracy of the monitoring results obtained by passing it through the classifier-based drilling overflow well leakage monitor. In this way, through comprehensive analysis of the time series characteristics of the inlet and outlet mass flow rates, the dynamic correlation of flow changes during the drilling process can be fully captured, the monitoring coverage can be improved, and the intelligence level of drilling overflow and well leakage monitoring can be enhanced to improve the safety and efficiency of drilling operations.

[0090] In summary, the drilling overflow and well leakage intelligent monitoring method based on the embodiment of the present application is explained, which first collects the mass flow at the inlet and outlet of the drilling circulation system and transmits it to the data processing cloud, and then uses the data analysis and processing algorithm based on artificial intelligence to extract features from the collected data, so as to monitor the overflow and well leakage of the drilling by utilizing the time series coordinated changes between the mass flow time series data at the inlet and the mass flow time series data at the outlet. In this way, by comprehensively analyzing the time series characteristics of the inlet and outlet mass flow rates, the dynamic correlation of flow changes during the drilling process can be fully captured, and the monitoring coverage can be improved, thereby improving the intelligence of drilling overflow and well leakage monitoring, so as to improve the safety and efficiency of drilling operations.

[0091] Figure 5 FIG. 1 is a system block diagram of an intelligent monitoring system for drilling overflow and well leakage according to an embodiment of the present application. Figure 5As shown, according to the intelligent monitoring system 100 for drilling overflow and well leakage of the embodiment of the present application, it includes: a mass flow timing data acquisition module 110, which is used to set a first mass flow meter and a second mass flow meter at the inlet of the drilling circulation system to respectively collect the mass flow timing data at the inlet and the mass flow timing data at the outlet of the drilling circulation system; a mass flow timing data transmission module 120, which is used to transmit the mass flow timing data at the inlet and the mass flow timing data at the outlet of the drilling circulation system to the data processing cloud; a drilling overflow and well leakage monitoring module 130, which is used to monitor the overflow and well leakage conditions of the drilling based on the timing coordinated changes between the mass flow timing data at the inlet and the mass flow timing data at the outlet on the data processing cloud.

[0092] Here, those skilled in the art can understand that the specific operations of each step in the above-mentioned drilling overflow well leakage intelligent monitoring system have been referred to above. Figures 1 to 4 The intelligent monitoring method for drilling overflow and well leakage has been introduced in detail, and therefore, its repeated description will be omitted.

[0093] As described above, the drilling overflow and well leakage intelligent monitoring system 100 according to the embodiment of the present application can be implemented in various terminal devices. In one example, the drilling overflow and well leakage intelligent monitoring system 100 can be integrated into the terminal device as a software module and / or a hardware module. For example, the drilling overflow and well leakage intelligent monitoring system 100 can be a software module in the operating system of the terminal device, or can be an application developed for the terminal device; of course, the drilling overflow and well leakage intelligent monitoring system 100 can also be one of the many hardware modules of the terminal device.

[0094] Alternatively, in another example, the drilling overflow and well leakage intelligent monitoring system 100 and the terminal device may also be separate devices, and the drilling overflow and well leakage intelligent monitoring system 100 may be connected to the terminal device via a wired and / or wireless network and transmit interactive information in accordance with an agreed data format.

[0095] In summary, the drilling overflow and well leakage intelligent monitoring system based on the embodiment of the present application is explained, which first collects the mass flow at the inlet and outlet of the drilling circulation system and transmits it to the data processing cloud, and then uses the data analysis and processing algorithm based on artificial intelligence to extract features from the collected data, so as to monitor the overflow and well leakage of the drilling by utilizing the time series coordinated changes between the mass flow time series data at the inlet and the mass flow time series data at the outlet. In this way, by comprehensively analyzing the time series characteristics of the inlet and outlet mass flow rates, the dynamic correlation of flow changes during the drilling process can be fully captured, and the monitoring coverage can be improved, thereby improving the intelligence of drilling overflow and well leakage monitoring, so as to improve the safety and efficiency of drilling operations.

Claims

1. A drilling overflow and well leakage intelligent monitoring method, characterized in that: include: A first mass flow meter and a second mass flow meter are arranged at the inlet of the drilling circulation system to respectively collect the mass flow time series data at the inlet and the mass flow time series data at the outlet of the drilling circulation system; Transmitting the mass flow rate time series data at the inlet and the mass flow rate time series data at the outlet of the drilling circulation system to the data processing cloud; In the data processing cloud, the overflow and leakage conditions of the drilling are monitored based on the time series coordinated changes between the mass flow time series data at the inlet and the mass flow time series data at the outlet, including the steps of: arranging the mass flow time series data at the inlet and the mass flow time series data at the outlet in time series, and obtaining a drilling inlet-outlet mass flow time series difference vector based on the arranged drilling inlet mass flow time series input vector and the drilling outlet mass flow time series input vector; respectively performing time series encoding on the drilling inlet mass flow time series input vector, the drilling outlet mass flow time series input vector and the drilling inlet-outlet mass flow time series difference vector to obtain a drilling inlet time series feature, a drilling outlet time series feature and a drilling inlet-outlet time series difference feature; The drilling entrance-exit time series difference feature is respectively matched and interacted with the drilling entrance time series feature and the drilling exit time series feature to obtain a drilling entrance influence difference change representation and a drilling exit influence difference change representation; based on the energy synergistic response aggregation between the drilling entrance influence difference change representation and the drilling exit influence difference change representation, a multi-modal interactive aggregation representation of the drilling entrance and exit mass flow is obtained to obtain a monitoring result; Among them, based on the energy synergistic response aggregation between the drilling entrance impact difference change representation and the drilling exit impact difference change representation, the multi-modal interactive aggregation representation of the drilling entrance and exit mass flow is obtained, and the monitoring results are obtained, including: Performing feature interactive aggregation based on energy collaborative response on the drilling inlet mass flow rate impact difference change characterization vector and the drilling outlet mass flow rate impact difference change characterization vector to obtain a drilling inlet and outlet mass flow rate multimodal interactive aggregation characterization vector as the drilling inlet and outlet mass flow rate multimodal interactive aggregation characterization; Based on the multi-modal interactive aggregation characterization of the mass flow rate at the inlet and outlet of the drilling well, the monitoring result is obtained, and the monitoring result is used to indicate whether overflow and well leakage occur in the drilling well; Among them, the drilling inlet mass flow rate impact difference change characterization vector and the drilling outlet mass flow rate impact difference change characterization vector are subjected to feature interactive aggregation based on energy collaborative response to obtain the drilling inlet and outlet mass flow multimodal interactive aggregation characterization vector, including: Calculate the positional division between the drilling inlet mass flow rate impact difference change characterization vector and the drilling outlet mass flow rate impact difference change characterization vector to obtain the drilling inlet and outlet mass flow rate change response characterization vector; Calculating the energy synergistic response value of each position in the drilling inlet and outlet mass flow rate change response characterization vector to obtain the drilling inlet and outlet mass flow rate energy synergistic response value vector; Performing normalization processing based on a sigmoid function on the drilling inlet and outlet mass flow energy coordinated response value vector to obtain a drilling inlet and outlet mass flow energy coordinated response value weight value vector; Based on the drilling inlet and outlet mass flow energy coordinated response value weight value vector, the drilling inlet mass flow impact difference change characterization vector and the drilling outlet mass flow impact difference change characterization vector are interactively aggregated to obtain the drilling inlet and outlet mass flow multimodal interactive aggregation characterization vector.

2. The drilling overflow and lost circulation intelligent monitoring method according to claim 1 is characterized in that: The mass flow time series data at the inlet and the mass flow time series data at the outlet are arranged in time series, and based on the arranged drilling inlet mass flow time series input vector and the drilling outlet mass flow time series input vector, a drilling inlet-outlet mass flow time series difference vector is obtained, including: Arranging the mass flow time series data at the inlet and the mass flow time series data at the outlet according to the time dimension to obtain the drilling inlet mass flow time series input vector and the drilling outlet mass flow time series input vector; The position-by-position difference between the wellbore inlet mass flow timing input vector and the wellbore outlet mass flow timing input vector is calculated to obtain the wellbore inlet-outlet mass flow timing difference vector.

3. The drilling overflow and lost circulation intelligent monitoring method according to claim 2 is characterized in that: The drilling inlet mass flow timing input vector, the drilling outlet mass flow timing input vector and the drilling inlet-outlet mass flow timing difference vector are respectively time-coded to obtain a drilling inlet timing feature, a drilling outlet timing feature and a drilling inlet-outlet timing difference feature, including: The drilling inlet mass flow timing input vector, the drilling outlet mass flow timing input vector and the drilling inlet-outlet mass flow timing difference vector are respectively passed through a timing feature extractor based on a Bi-LSTM model to obtain a drilling inlet mass flow timing feature vector as the drilling inlet timing feature, a drilling outlet mass flow timing feature vector as the drilling outlet timing feature and a drilling inlet-outlet mass flow timing difference feature vector as the drilling inlet-outlet timing difference feature.

4. The drilling overflow and lost circulation intelligent monitoring method according to claim 3 is characterized in that: The drilling entry-exit time series difference feature is respectively matched and interacted with the drilling entry time series feature and the drilling exit time series feature to obtain a drilling entry impact difference change representation and a drilling exit impact difference change representation, including: Performing feature fine-grained optimal matching characterization on the drilling inlet mass flow time series feature vector and the drilling inlet-outlet mass flow time series difference feature vector to obtain a drilling inlet mass flow impact difference change characterization vector as the drilling inlet impact difference change characterization; Performing feature fine-grained optimal matching characterization on the well outlet mass flow time series feature vector and the well inlet-outlet mass flow time series difference feature vector to obtain a well outlet mass flow impact difference change characterization vector as the well outlet impact difference change characterization; Among them, the drilling inlet mass flow time series feature vector and the drilling inlet-outlet mass flow time series difference feature vector are characterized by fine-grained optimal matching to obtain a drilling inlet mass flow impact difference change characterization vector, including: Based on a predetermined scale, fine-grained feature decoupling is performed on the drilling inlet mass flow time series feature vector and the drilling inlet-outlet mass flow time series difference feature vector to obtain a set of drilling inlet mass flow local time series feature vectors and a set of drilling inlet-outlet mass flow local time series feature vectors; Calculate the hyperbolic space distance metric factor between any two local time series feature vectors of the well inlet mass flow and the local time series feature vectors of the well inlet-outlet mass flow in the set of the well inlet mass flow local time series feature vectors and the set of the well inlet-outlet mass flow local time series feature vectors; Based on the hyperbolic space distance metric factor, the set of the local time series feature vectors of the well inlet mass flow and the set of the local time series feature vectors of the well inlet-outlet mass flow are screened for optimal feature pairing to obtain the set of optimal pairing of the local time series feature vectors of the well inlet mass flow and the local time series feature vectors of the well inlet-outlet mass flow; Pairing each optimal drilling inlet mass flow local time series feature vector and drilling inlet-outlet mass flow local time series feature vector in the set of paired optimal drilling inlet mass flow local time series feature vector and drilling inlet-outlet mass flow local time series feature vector into a feature multi-dimensional interaction module to obtain a set of drilling inlet mass flow time series feature paired interaction fusion feature vectors; The set of the paired interactive fusion feature vectors of the drilling inlet mass flow time series features is cascaded to obtain a characterization vector of the drilling inlet mass flow impact difference change.

5. The drilling overflow and lost circulation intelligent monitoring method according to claim 4 is characterized in that: Calculating the energy synergistic response value of each position in the drilling inlet and outlet mass flow rate change response characterization vector to obtain the drilling inlet and outlet mass flow rate energy synergistic response value vector, including: Respectively calculating the mean and variance of the drilling inlet and outlet mass flow rate change response characterization vector to obtain the drilling inlet and outlet mass flow rate change mean and the drilling inlet and outlet mass flow rate change variance; The value obtained by adding the variance of the mass flow rate change at the wellbore inlet and outlet and the preset hyperparameter is multiplied by four to obtain a first wellbore inlet and outlet mass flow rate energy coordinated response value; Calculate the square of each value after subtracting the drilling inlet and outlet mass flow rate change response characterization vector from the drilling inlet and outlet mass flow rate change mean value by position to obtain the drilling inlet and outlet mass flow rate change difference value vector; The modulated drilling inlet and outlet mass flow rate change variance obtained by multiplying the drilling inlet and outlet mass flow rate change variance by two, the value obtained by multiplying the preset hyperparameter by two, and the drilling inlet and outlet mass flow rate change difference value vector are added by position to obtain a second drilling inlet and outlet mass flow rate energy coordinated response value vector; The first wellbore inlet and outlet mass flow energy coordinated response value is divided by the second wellbore inlet and outlet mass flow energy coordinated response value vector by position to obtain the wellbore inlet and outlet mass flow energy coordinated response value vector.

6. The drilling overflow and lost circulation intelligent monitoring method according to claim 5 is characterized in that: Based on the drilling inlet and outlet mass flow energy coordinated response value weight value vector, the drilling inlet mass flow impact difference change characterization vector and the drilling outlet mass flow impact difference change characterization vector are interactively aggregated to obtain the drilling inlet and outlet mass flow multimodal interactive aggregation characterization vector, including: Multiplying the drilling inlet and outlet mass flow energy coordinated response value weight value vector by the drilling inlet mass flow impact difference change characterization vector according to the position point to obtain the drilling inlet mass flow impact difference change characterization weighted vector; Subtract the constant 1 from the weight value vector of the coordinated response value of the mass flow energy at the well inlet and outlet by position, and multiply the subtracted vector by the vector representing the difference in influence of the well outlet mass flow by position to obtain a weighted vector representing the difference in influence of the well outlet mass flow; The weighted vector representing the difference in impact of the well inlet mass flow and the weighted vector representing the difference in impact of the well outlet mass flow are added by position to obtain the multi-modal interaction aggregation representation vector of the well inlet and outlet mass flow.

7. The drilling overflow and lost circulation intelligent monitoring method according to claim 6 is characterized in that: Based on the multimodal interactive aggregation characterization of the drilling inlet and outlet mass flow, the monitoring result is obtained, which is used to: pass the multimodal interactive aggregation characterization vector of the drilling inlet and outlet mass flow through a classifier-based drilling overflow and well leakage monitor to obtain the monitoring result.

8. A drilling overflow and well leakage intelligent monitoring system, used to execute the drilling overflow and well leakage intelligent monitoring method according to any one of claims 1 to 7, characterized in that: include: A mass flow time series data acquisition module, used to set a first mass flow meter and a second mass flow meter at the inlet of the drilling circulation system to respectively acquire the mass flow time series data at the inlet and the mass flow time series data at the outlet of the drilling circulation system; A mass flow time series data transmission module, used to transmit the mass flow time series data at the inlet and the mass flow time series data at the outlet of the drilling circulation system to a data processing cloud; The drilling overflow and well leakage monitoring module is used to monitor the overflow and well leakage of the drilling on the data processing cloud based on the time series coordinated changes between the mass flow time series data at the inlet and the mass flow time series data at the outlet, wherein the drilling overflow and well leakage monitoring module includes: a mass flow time series data arrangement unit, which is used to arrange the mass flow time series data at the inlet and the mass flow time series data at the outlet in time series, and obtain the drilling inlet-outlet mass flow time series difference vector based on the arranged drilling inlet mass flow time series input vector and the drilling outlet mass flow time series input vector; a drilling mass flow feature time series encoding unit, which is used to arrange the drilling inlet mass flow time series input vector and the drilling outlet mass flow time series input vector; The flow rate time series input vector and the drilling inlet-outlet mass flow time series difference vector are respectively time-coded to obtain the drilling inlet time series feature, the drilling outlet time series feature and the drilling inlet-outlet time series difference feature; a drilling feature fine-grained matching interaction unit is used to perform drilling feature fine-grained matching interaction on the drilling inlet-outlet time series difference feature with the drilling inlet time series feature and the drilling outlet time series feature respectively to obtain the drilling inlet influence difference change representation and the drilling outlet influence difference change representation; an energy synergistic response aggregation unit is used to obtain the drilling inlet and outlet mass flow multi-modal interactive aggregation representation based on the energy synergistic response aggregation between the drilling inlet influence difference change representation and the drilling outlet influence difference change representation to obtain the monitoring result.

Images