A deepwater drilling overflow early warning device and method

Abstract

The invention discloses an early warning device and method for deepwater drilling overflow, which relate to the field of petroleum engineering technology, and include: a data acquisition module, which is used to collect data during drilling, the data including downhole bottom pressure, temperature, acoustic logging data, gamma logging data, underwater gas content data and above-water well fluid flow data; an environmental compensation module, which is used to monitor environmental changes, and dynamically correct the data collected by the data acquisition module using a real-time correction algorithm based on a neural network; a data processing and analysis module, which is used to clean and process the data of the data acquisition module and extract features, and transmit the features to an early warning system module. The invention obtains accurate environmental data through the environmental compensation module, and these data are processed by the real-time correction algorithm to eliminate external interference, such as the influence of water flow fluctuations, temperature changes, etc. on sensor readings, thereby improving the accuracy and reliability of the data.

Description

Technical Field

[0001] The invention relates to the technical field of petroleum engineering, and in particular to an early warning device and method for deepwater drilling overflow. Background Art

[0002] Deepwater drilling refers to drilling operations conducted in deeper waters. Deepwater drilling has become an important means of exploring and developing marine oil resources. Deepwater drilling faces complex environmental challenges, including extreme conditions such as high pressure, low temperature, high salinity, and strong currents in the deep sea. During deepwater drilling, overflow is a common drilling risk. Overflow refers to the phenomenon that formation fluid enters the wellbore through formation cracks or pores during drilling. If measures are not taken in time to control the overflow, more serious drilling accidents such as blowouts may occur. Therefore, early warning of overflow is crucial to ensure the safety of deepwater drilling operations.

[0003] After searching, the invention patent with Chinese patent number CN114482990A discloses a method and device for early warning of drilling overflow;

[0004] Compared with the prior art, the invention patent with Chinese patent number CN114482990A establishes an overflow data set for neural network training and neural network input based on drilling logging data and drilling engineering data; establishes Neural network model; using training sample data set, The neural network model is trained to obtain an overflow warning neural network model; the test sample data set is input into the overflow warning neural network model to obtain the first drilling overflow probability; the wellbore pressure is calculated according to the drilling logging data and the drilling engineering data to obtain the second drilling overflow probability; the first drilling overflow probability and the second drilling overflow probability are weighted to obtain the weighted probability of drilling overflow; when the weighted probability exceeds the preset value, the drilling overflow warning information is issued. The present invention can improve the accuracy and processing efficiency of drilling overflow warning.

[0005] However, in actual use of the above-mentioned device, due to the complex marine environment, water flow, pressure and temperature changes may interfere with the sensor readings and affect the monitoring accuracy. In addition, the threshold is not adjusted dynamically in time according to the data situation, which may lead to false alarms or missed alarms. Therefore, a deepwater drilling overflow early warning device and method are proposed. Summary of the invention

[0006] The purpose of the present invention is to solve the shortcomings of the prior art, such as the complex marine environment, the water flow, pressure and temperature changes may interfere with the sensor readings and affect the monitoring accuracy, and the threshold is not adjusted dynamically in time according to the data situation, which may lead to false alarms or missed alarms. An early warning device and method for deepwater drilling overflow is proposed to solve the shortcomings of the prior art, such as the complex marine environment, the water flow, pressure and temperature changes may interfere with the sensor readings and affect the monitoring accuracy, and the threshold is not adjusted dynamically in time according to the data situation, which may lead to false alarms or missed alarms.

[0007] In order to achieve the above object, the present invention adopts the following technical solutions:

[0008] A deepwater drilling overflow early warning device and method, comprising:

[0009] A data acquisition module is used to collect data during drilling, including downhole bottom pressure, temperature, acoustic logging data, gamma logging data, underwater gas content data and above-water well fluid flow data;

[0010] The environmental compensation module is used to monitor environmental changes and dynamically correct the data collected by the data acquisition module using a real-time correction algorithm based on a neural network;

[0011] The data processing and analysis module is used to clean and extract features from the data of the data acquisition module, and the features are transmitted to the early warning system module;

[0012] The threshold setting and adjustment module is used to set the threshold and adjust the threshold through the incremental learning model according to the data fluctuation amplitude;

[0013] The early warning system module receives the threshold value set by the threshold setting and adjustment module and the data collected by the data acquisition module after dynamic correction by the environmental compensation module. When the data is identified to be abnormal, the early warning system module automatically triggers the early warning signal;

[0014] Data recording module, used to archive all monitoring data, correction data and warning records;

[0015] The data acquisition module monitors the data during drilling in real time, while the environmental compensation module monitors environmental changes, and then dynamically corrects the data of the data acquisition module. The data processing and analysis module cleans and extracts features from the data of the data acquisition module, and then transmits the features to the early warning system module. The threshold setting and adjustment module adjusts the threshold according to the data fluctuation amplitude after setting the threshold. The early warning system module receives the data after cleaning and feature extraction from the data processing and analysis module, and compares it with the threshold adjusted by the threshold setting and adjustment module. The data recording module archives all monitoring data, correction data and early warning records at the same time;

[0016] The threshold setting and adjustment module includes a threshold setting unit and a dynamic threshold adjustment unit. The threshold setting unit sets the initial monitoring threshold according to the research method and historical data, and the dynamic threshold adjustment unit sets the threshold adjustment rule and automatically adjusts it according to the data fluctuation amplitude or frequency;

[0017] The threshold adjustment process of the dynamic threshold adjustment unit is as follows:

[0018] Real-time trend monitoring:

[0019] Data collection: real-time collection of data from the data collection module;

[0020] Trend Analysis:

[0021] Sliding window method: set a time window and calculate the mean and standard deviation of these data;

[0022] Anomaly detection algorithm: Use the Z-score statistical method to identify data points that are beyond the normal range. If the current pressure exceeds the past average + 2 times the standard deviation, it is marked as abnormal. The algorithm formula of the Z-score statistical method is:

[0023]

[0024] in:

[0025] Z represents the Z-score of data point X, that is, the standardized position of the data point relative to the mean of the data set, X represents the original value or observed value of the data point, μ represents the mean of the data set, and σ represents the standard deviation of the data set;

[0026] Dynamic adjustment mechanism:

[0027] Threshold setting: The initial threshold is set based on historical data and industry standards, and the threshold is updated regularly to reflect changes in real-time data to ensure sensitivity and specificity;

[0028] Adaptive Adjustment:

[0029] Incremental learning: Establish an incremental learning model, continuously learn new data, and use new data points to adjust the prediction model so that it can adapt to different drilling environments;

[0030] Feedback mechanism: When an abnormality is detected, the event is recorded and the threshold is automatically adjusted according to the subsequent situation. If the alarm is triggered three times in a row, the threshold is increased to reduce the false alarm rate;

[0031] Example of early warning process:

[0032] Data monitoring:

[0033] Collect and upload real-time data every minute;

[0034] Trend Analysis:

[0035] The system analyzes the data of the last 10 minutes every 5 minutes and calculates the mean and standard deviation;

[0036] Anomaly Detection:

[0037] If the pressure value collected at a certain time exceeds the currently calculated threshold, an alarm is triggered;

[0038] Dynamic adjustment of threshold:

[0039] When alarms occur frequently, analyze the causes of the alarms and adjust the thresholds appropriately.

[0040] The above technical solution further includes:

[0041] The data acquisition module includes a downhole data acquisition unit, an underwater data acquisition unit and an above-water data acquisition unit. The downhole data acquisition unit uses measurement while drilling and logging while drilling technology to monitor the key parameters of bottom hole pressure, temperature, acoustic logging data and gamma logging data in real time. The underwater data acquisition unit obtains gas content through a monitor and obtains flow data of drilling fluid in the annulus through a flow measurement device, and uses underwater cameras and sensors to monitor underwater environmental parameters and conditions. The above-water data acquisition unit monitors the flow of drilling fluid in the platform section in real time through a drilling fluid flow monitor and a liquid level monitor. The environmental compensation module includes an environmental factor monitoring unit, a real-time correction algorithm unit and a meteorological and sea state data correction unit. The environmental factor monitoring unit monitors and records environmental parameter changes in real time through environmental monitoring sensors. The real-time correction algorithm unit uses a real-time correction algorithm to cope with the impact of environmental changes on sensor readings. The meteorological and sea state data correction unit performs correction processing on the collected meteorological and sea state data.

[0042] The real-time correction algorithm unit is a real-time correction algorithm based on a neural network, and the specific steps are as follows:

[0043] Data collection and preprocessing:

[0044] Data collection: Obtaining a large number of data samples from various sensors and environmental monitoring devices, including sensor readings and environmental parameters recorded simultaneously, such as temperature, humidity, and pressure;

[0045] Data cleaning: Carefully check and clean the collected data to remove noise, outliers, and missing data;

[0046] Data division: The cleaned data is divided into training set, validation set and test set. The training set is used for the model training process, that is, to minimize the loss function by adjusting the model parameters. The validation set is used to evaluate the performance of the model during the training process. The test set is used to evaluate the generalization ability of the model after the training is completed.

[0047] Neural network model design:

[0048] Determine the network structure: Select the convolutional neural network for real-time correction. The convolutional layer is the core of the convolutional neural network. The features of the input data are extracted through the convolution operation. The output feature map size calculation formula of the convolutional layer is: , where N is the size of the output feature map, W is the size of the input image, F is the size of the convolution kernel, P is the padding size, which is used to add extra zero pixels on the border of the input image, and S is the stride, that is, the distance the convolution kernel slides on the input image;

[0049] Activation function: Select the ReLU activation function to enhance the nonlinear mapping ability of the model and define the loss function. The ReLU activation function algorithm formula is: , in this formula, Represents the output of the ReLU activation function, and x represents the input value;

[0050] Loss function: The difference between the model prediction value and the actual value is evaluated by the mean square error. The algorithm formula is:

[0051] , where: MSE is the mean square error, n is the total number of data points, represents the i-th actual measurement value, represents the i-th estimated value;

[0052] Model training:

[0053] Initialization parameters: Randomly initialize the weights and bias parameters of the neural network;

[0054] Forward propagation: input the training data into the neural network and calculate the output value layer by layer. The output value of each layer will be passed as the input value of the next layer until the final prediction result is obtained;

[0055] Back propagation: Calculate the difference between the predicted result and the actual value according to the loss function, and pass the error to each layer through the back propagation algorithm to update the weight and bias parameters to reduce the value of the loss function;

[0056] Iterative training: Repeat the process of forward propagation and back propagation until the stopping condition is met;

[0057] Model validation and testing;

[0058] Verify the model: Use the validation set to verify the trained model and evaluate the generalization ability and performance of the model;

[0059] Test model: Use the test set to test the model to ensure its accuracy and reliability in practical applications;

[0060] Deployment and application:

[0061] Model deployment: deploy the trained model to the actual application scenario;

[0062] Real-time correction: Use the deployed model to make real-time corrections to sensor readings, eliminating external interference and improving data accuracy and reliability.

[0063] The data processing and analysis module includes a data cleaning unit and a feature extraction unit. The data cleaning unit cleans the data of the data acquisition module to remove noise and abnormal values ​​therein, and the feature extraction unit extracts key features such as wind speed, wind direction, and wave height from the data cleaned by the data cleaning unit.

[0064] The data cleaning unit is cleaned and processed by Gaussian filtering, and the algorithm formula of the Gaussian filtering is:

[0065]

[0066] in, is the pixel after filtering The gray value of The original pixel The gray value in the field, m and n are the sizes of the filter window, It represents the sum of all pixels in the area.

[0067] The early warning system module includes a data monitoring unit and an alarm triggering unit. The data monitoring unit continuously monitors real-time data and compares it with the dynamic threshold. The alarm triggering unit receives the comparison result of the data monitoring unit. When the real-time data exceeds the set dynamic threshold, an alarm is immediately triggered. The data recording module includes a data archiving unit and a feedback mechanism unit. The data archiving unit archives all monitoring data, correction data and early warning records for subsequent analysis and improvement.

[0068] A deepwater drilling overflow early warning method comprises the following steps:

[0069] Step 1: Use the data acquisition module to monitor the downhole data, underwater data and surface data during drilling in real time, and use the environmental compensation module to monitor environmental changes and correct the sensor readings of the data acquisition module;

[0070] Step 2: The data collected by the data acquisition module is processed and analyzed by the data processing and analysis module, and then transmitted to the early warning system module. The threshold value of the early warning system module is dynamically adjusted according to the real-time data trend monitored by the threshold setting and adjustment module, and then the early warning system module issues an alarm according to the comparison result;

[0071] Step 3: The data recording module records all data generated by the early warning work.

[0072] The present invention has the following beneficial effects:

[0073] In the present invention, when carrying out early warning work, due to the complex marine environment, water flow, pressure and temperature changes may interfere with the sensor readings and affect the monitoring accuracy. Accurate environmental data can be obtained through the environmental compensation module. These data are processed by the real-time correction algorithm to eliminate external interference, such as the impact of water flow fluctuations, temperature changes, etc. on sensor readings, thereby improving the accuracy and reliability of the data, which is crucial for accurately judging abnormal conditions during the drilling process.

[0074] In the present invention, the threshold setting and adjustment module can identify the normal operating range through in-depth analysis of historical data, thereby setting a reasonable monitoring threshold and reducing the possibility of false alarms and missed alarms. Moreover, through accurate threshold setting and dynamic adjustment, an alarm can be issued in time before overflow occurs, thereby taking emergency measures and reducing potential economic losses. BRIEF DESCRIPTION OF THE DRAWINGS

[0075] Figure 1 A system block diagram of a deepwater drilling overflow early warning device and method proposed by the present invention.

[0076] In the figure: 1. Data acquisition module; 2. Environmental compensation module; 3. Data processing and analysis module; 4. Threshold setting and adjustment module; 5. Early warning system module; 6. Data recording module. DETAILED DESCRIPTION

[0077] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention. Example

[0078] like Figure 1 As shown, the present invention proposes a deepwater drilling overflow early warning device, comprising:

[0079] Data acquisition module 1, used to collect data during drilling, including downhole bottom hole pressure, temperature, acoustic logging data, gamma logging data, underwater gas content data and above-water well fluid flow data;

[0080] Environmental compensation module 2, used to monitor environmental changes and dynamically correct the data collected by data acquisition module 1 using a real-time correction algorithm based on a neural network;

[0081] The data processing and analysis module 3 is used to clean and extract features from the data of the data acquisition module 1, and transmit the features to the early warning system module 5;

[0082] Threshold setting and adjustment module 4, used to set the threshold and adjust the threshold through the incremental learning model according to the data fluctuation amplitude;

[0083] The early warning system module 5 receives the threshold value set by the threshold setting and adjustment module 4 and the data collected by the data collection module 1 after dynamic correction by the environmental compensation module 2. When the data is identified to be abnormal, the early warning system module 5 automatically triggers an early warning signal;

[0084] Data recording module 6, used to archive all monitoring data, calibration data and warning records;

[0085] The data acquisition module 1 monitors the data during drilling in real time, while the environmental compensation module 2 monitors environmental changes, and then dynamically corrects the data of the data acquisition module 1. The data processing and analysis module 3 cleans and extracts features from the data of the data acquisition module 1, and then transmits the features to the early warning system module 5. The threshold setting and adjustment module 4 adjusts the threshold according to the data fluctuation amplitude after setting the threshold. The early warning system module 5 receives the data after cleaning and feature extraction from the data processing and analysis module 3, and compares it with the threshold adjusted by the threshold setting and adjustment module 4. The data recording module 6 archives all monitoring data, correction data and early warning records at the same time;

[0086] The threshold setting and adjustment module 4 includes a threshold setting unit and a dynamic threshold adjustment unit. The threshold setting unit sets the initial monitoring threshold according to the research method and historical data, and the dynamic threshold adjustment unit sets the threshold adjustment rule and automatically adjusts it according to the data fluctuation amplitude or frequency;

[0087] The threshold adjustment process of the dynamic threshold adjustment unit is:

[0088] Real-time trend monitoring:

[0089] Data collection: real-time collection of data from data collection module 1;

[0090] Trend Analysis:

[0091] Sliding window method: set a time window and calculate the mean and standard deviation of these data;

[0092] Anomaly detection algorithm: Use the Z-score statistical method to identify data points that are beyond the normal range. If the current pressure exceeds the past average + 2 times the standard deviation, it is marked as abnormal. The algorithm formula of the Z-score statistical method is:

[0093]

[0094] in:

[0095] Z represents the Z-score of data point X, that is, the standardized position of the data point relative to the mean of the data set, X represents the original value or observed value of the data point, μ represents the mean of the data set, and σ represents the standard deviation of the data set;

[0096] Dynamic adjustment mechanism:

[0097] Threshold setting: The initial threshold is set based on historical data and industry standards, and the threshold is updated regularly to reflect changes in real-time data to ensure sensitivity and specificity;

[0098] Adaptive Adjustment:

[0099] Incremental learning: Establish an incremental learning model, continuously learn new data, and use new data points to adjust the prediction model so that it can adapt to different drilling environments;

[0100] Feedback mechanism: When an abnormality is detected, the event is recorded and the threshold is automatically adjusted according to the subsequent situation. If the alarm is triggered three times in a row, the threshold is increased to reduce the false alarm rate;

[0101] Example of early warning process:

[0102] Data monitoring:

[0103] Collect and upload real-time data every minute;

[0104] Trend Analysis:

[0105] The system analyzes the data of the last 10 minutes every 5 minutes and calculates the mean and standard deviation;

[0106] Anomaly Detection:

[0107] If the pressure value collected at a certain time exceeds the currently calculated threshold, an alarm is triggered;

[0108] Dynamic adjustment of threshold:

[0109] When alarms occur frequently, analyze the causes of the alarms and adjust the thresholds appropriately.

[0110] In this embodiment, a deepwater drilling overflow early warning device and method, the working principle is that when performing overflow early warning work during deepwater drilling, firstly, the bottom hole pressure, temperature, acoustic logging data, gamma logging data key parameters, underwater gas content, drilling fluid flow data and the flow of well fluid above water in the platform section are monitored in real time, and the environment is monitored in real time through the environmental compensation module 2, the changes of external conditions are captured in time, and the correction algorithm is applied to dynamically correct the sensor readings according to the monitored environmental changes to eliminate the influence of external interference, thereby improving the accuracy and reliability of the data;

[0111] Then the data processing and analysis module 3 cleans the corrected data to remove noise, outliers and incomplete data to ensure data quality, and extracts key features from the processed data. At the same time, the threshold setting and adjustment module 4 conducts in-depth analysis of historical data to determine a reasonable monitoring threshold, reduce the possibility of false alarms and missed alarms, monitor the trend of real-time data, identify abnormal fluctuations, and promptly reflect potential risks. According to the changes in real-time data, the monitoring threshold is automatically adjusted to adapt the system to new environmental conditions. Then the early warning system module 5 implements the early warning algorithm based on the corrected data and dynamic thresholds to analyze whether there is an overflow risk. Once it is detected that the data exceeds the set threshold, the alarm system is immediately triggered, and an alarm can be issued in time before the overflow occurs, so that emergency measures can be taken to reduce potential economic losses. Finally, the monitoring data, correction data and early warning records are archived through the data recording module 6.

[0112] In one embodiment, the data acquisition module 1 includes a downhole data acquisition unit, an underwater data acquisition unit and an above-water data acquisition unit. The downhole data acquisition unit uses measurement while drilling and logging while drilling technology to monitor the key parameters of bottom hole pressure, temperature, acoustic logging data, and gamma logging data in real time. The underwater data acquisition unit obtains gas content through a monitor and obtains flow data of drilling fluid in the annulus through a flow measuring device, and uses underwater cameras and sensors to monitor underwater environmental parameters and conditions. The above-water data acquisition unit uses a drilling fluid flow monitor and a liquid level monitor to monitor the flow of drilling fluid in the platform section in real time.

[0113] In one embodiment, the environmental compensation module 2 includes an environmental factor monitoring unit, a real-time correction algorithm unit and a meteorological and sea condition data correction unit. The environmental factor monitoring unit monitors and records changes in environmental parameters in real time through environmental monitoring sensors. The real-time correction algorithm unit uses a real-time correction algorithm to cope with the impact of environmental changes on sensor readings. The meteorological and sea condition data correction unit performs correction processing on the collected meteorological and sea condition data.

[0114] In one embodiment, the real-time correction algorithm unit is a real-time correction algorithm based on a neural network, and the specific steps are as follows:

[0115] Data collection and preprocessing:

[0116] Data collection: Obtain a large number of data samples from various sensors and environmental monitoring devices, including sensor readings and environmental parameters recorded simultaneously;

[0117] Data cleaning: Carefully check and clean the collected data to remove noise, outliers, and missing data;

[0118] Data division: The cleaned data is divided into training set, validation set and test set. The training set is used for model training, the validation set is used to evaluate the performance of the model during training, and the test set is used to evaluate the generalization ability of the model after training.

[0119] Neural network model design:

[0120] Determine the network structure: Select the convolutional neural network for real-time correction. The convolutional layer is the core of the convolutional neural network. The features of the input data are extracted through the convolution operation. The output feature map size calculation formula of the convolutional layer is: , where N is the size of the output feature map, W is the size of the input image, F is the size of the convolution kernel, P is the padding size, which is used to add extra zero pixels on the border of the input image, and S is the stride, that is, the distance the convolution kernel slides on the input image;

[0121] Activation function: Select the ReLU activation function to enhance the nonlinear mapping ability of the model and define the loss function. The ReLU activation function algorithm formula is: , in this formula, Represents the output of the ReLU activation function, and x represents the input value;

[0122] Loss function: The difference between the model prediction value and the actual value is evaluated by the mean square error. The algorithm formula is:

[0123] , where: MSE is the mean square error, n is the total number of data points, represents the i-th actual measurement value, represents the i-th estimated value;

[0124] Model training:

[0125] Initialization parameters: Randomly initialize the weights and bias parameters of the neural network;

[0126] Forward propagation: input the training data into the neural network and calculate the output value layer by layer. The output value of each layer will be passed as the input value of the next layer until the final prediction result is obtained;

[0127] Back propagation: Calculate the difference between the predicted result and the actual value according to the loss function, and pass the error to each layer through the back propagation algorithm to update the weight and bias parameters to reduce the value of the loss function;

[0128] Iterative training: Repeat the process of forward propagation and back propagation until the stopping condition is met;

[0129] Model validation and testing;

[0130] Verify the model: Use the validation set to verify the trained model and evaluate the generalization ability and performance of the model;

[0131] Test model: Use the test set to test the model to ensure its accuracy and reliability in practical applications;

[0132] Deployment and application:

[0133] Model deployment: deploy the trained model to the actual application scenario;

[0134] Real-time correction: Use the deployed model to make real-time corrections to sensor readings to eliminate external interference.

[0135] In one embodiment, the data processing and analysis module 3 includes a data cleaning unit and a feature extraction unit. The data cleaning unit cleans the data of the data acquisition module 1 to remove noise and abnormal values, and the feature extraction unit extracts key features from the data cleaned by the data cleaning unit.

[0136] In one embodiment, the data cleaning unit performs cleaning and processing by Gaussian filtering, and the algorithm formula of Gaussian filtering is:

[0137]

[0138] in, is the pixel after filtering The gray value of The original pixel The gray value in the field, m and n are the sizes of the filter window, It represents the sum of all pixels in the area.

[0139] In one embodiment, the early warning system module 5 includes a data monitoring unit and an alarm triggering unit. The data monitoring unit continuously monitors the real-time data and compares it with the dynamic threshold. The alarm triggering unit receives the comparison result of the data monitoring unit. When the real-time data exceeds the set dynamic threshold, an alarm is immediately triggered.

[0140] In one embodiment, the data recording module 6 includes a data archiving unit and a feedback mechanism unit. The data archiving unit archives all monitoring data, correction data and warning records.

[0141] In one embodiment, a deepwater drilling overflow early warning method comprises the following steps:

[0142] Step 1: Use the data acquisition module 1 to monitor the downhole data, underwater data and surface data during drilling in real time, and use the environmental compensation module 2 to monitor environmental changes and correct the sensor readings of the data acquisition module 1;

[0143] Step 2: The data collected by the data acquisition module 1 is processed and analyzed by the data processing and analysis module 3, and then transmitted to the early warning system module 5. The threshold value of the early warning system module 5 is dynamically adjusted according to the real-time data trend monitored by the threshold setting and adjustment module 4, and then the early warning system module 5 issues an alarm according to the comparison result;

[0144] Step 3: The data recording module 6 records all data generated by the early warning work.

[0145] Although embodiments of the present invention have been shown and described, it will be appreciated by those skilled in the art that various changes, modifications, substitutions and variations may be made to the embodiments without departing from the principles and spirit of the present invention, and that the scope of the present invention is defined by the appended claims and their equivalents.

Claims

1. A deepwater drilling overflow early warning device, characterized in that: include: A data acquisition module is used to collect data during drilling, including downhole bottom pressure, temperature, acoustic logging data, gamma logging data, underwater gas content data and above-water well fluid flow data; The environmental compensation module is used to monitor environmental changes and dynamically correct the data collected by the data acquisition module using a real-time correction algorithm based on a neural network; The data processing and analysis module is used to clean and extract features from the data of the data acquisition module, and the features are transmitted to the early warning system module; The threshold setting and adjustment module is used to set the threshold and adjust the threshold through the incremental learning model according to the data fluctuation amplitude; The early warning system module receives the threshold value set by the threshold setting and adjustment module and the data collected by the data acquisition module after dynamic correction by the environmental compensation module. When the data is identified to be abnormal, the early warning system module automatically triggers the early warning signal; Data recording module, used to archive all monitoring data, correction data and warning records; The data acquisition module monitors the data during drilling in real time, while the environmental compensation module monitors environmental changes, and then dynamically corrects the data of the data acquisition module. The data processing and analysis module cleans and extracts features from the data of the data acquisition module, and then transmits the features to the early warning system module. The threshold setting and adjustment module adjusts the threshold according to the data fluctuation amplitude after setting the threshold. The early warning system module receives the data after cleaning and feature extraction by the data processing and analysis module, and compares it with the threshold adjusted by the threshold setting and adjustment module. The data recording module archives all monitoring data, correction data and early warning records at the same time; The threshold setting and adjustment module includes a threshold setting unit and a dynamic threshold adjustment unit. The threshold setting unit sets the initial monitoring threshold according to the research method and historical data, and the dynamic threshold adjustment unit sets the threshold adjustment rule and automatically adjusts it according to the data fluctuation amplitude or frequency; The threshold adjustment process of the dynamic threshold adjustment unit is as follows: S1. Real-time trend monitoring, the specific steps are: Collect data from data acquisition module in real time; S2. Trend analysis, this step is as follows: Set a time window, calculate the mean and standard deviation of these data, and use the Z-score statistical method to identify data points that are beyond the normal range. If the current pressure exceeds the sum of the past average value and 2 times the standard deviation, it is marked as abnormal. The algorithm formula of the Z-score statistical method is: Where: Z represents the Z-score of data point X, that is, the standardized position of the data point relative to the mean of the data set, X represents the original value or observed value of the data point, μ represents the mean of the data set, and σ represents the standard deviation of the data set; S3. Dynamic adjustment of threshold, this step is as follows: Initial thresholds are set based on historical data and industry standards, and thresholds are regularly updated to reflect changes in real-time data to ensure sensitivity and specificity; S4. Adaptive adjustment, this step is specifically as follows: S41. Incremental learning: Establish an incremental learning model to continuously learn new data and use new data points to adjust the prediction model so that it can adapt to different drilling environments; S42. Feedback adjustment: When an abnormality is detected, the event is recorded and the threshold is automatically adjusted according to subsequent situations. If the alarm is triggered three times in a row, the threshold is increased to reduce the false alarm rate.

2. The deepwater drilling overflow early warning device according to claim 1 is characterized in that: The data acquisition module includes a downhole data acquisition unit, an underwater data acquisition unit and an above-water data acquisition unit. The downhole data acquisition unit uses measurement while drilling and logging while drilling technology to monitor the key parameters of bottom hole pressure, temperature, acoustic logging data, and gamma logging data in real time. The underwater data acquisition unit obtains gas content through a monitor, obtains flow data of drilling fluid in the annulus through a flow measuring device, and uses underwater cameras and sensors to monitor underwater environmental parameters and conditions. The above-water data acquisition unit uses a drilling fluid flow monitor and a liquid level monitor to monitor the flow of drilling fluid in the platform section in real time.

3. The deepwater drilling overflow early warning device according to claim 1 is characterized in that: The environmental compensation module includes an environmental factor monitoring unit, a real-time correction algorithm unit and a meteorological and sea condition data correction unit. The environmental factor monitoring unit monitors and records changes in environmental parameters in real time through environmental monitoring sensors. The real-time correction algorithm unit uses a real-time correction algorithm to cope with the impact of environmental changes on sensor readings. The meteorological and sea condition data correction unit performs correction processing on the collected meteorological and sea condition data.

4. The deepwater drilling overflow early warning device according to claim 1, characterized in that: The data processing and analysis module includes a data cleaning unit and a feature extraction unit. The data cleaning unit cleans the data of the data acquisition module to remove noise and abnormal values ​​therein, and the feature extraction unit extracts key features from the data cleaned by the data cleaning unit.

5. The deepwater drilling overflow early warning device according to claim 1, characterized in that: The early warning system module includes a data monitoring unit and an alarm triggering unit. The data monitoring unit continuously monitors real-time data and compares it with the dynamic threshold. The alarm triggering unit receives the comparison result of the data monitoring unit. When the real-time data exceeds the set dynamic threshold, an alarm is immediately triggered. The data recording module includes a data archiving unit and a feedback mechanism unit. The data archiving unit archives all monitoring data, correction data and early warning records.

6. An overflow early warning method of the deepwater drilling overflow early warning device according to claim 1, characterized in that: The following steps are involved: Step 1: Use the data acquisition module to monitor the downhole data, underwater data and surface data during drilling in real time, and use the environmental compensation module to monitor environmental changes and correct the sensor readings of the data acquisition module; Step 2: The data collected by the data acquisition module is processed and analyzed by the data processing and analysis module, and then transmitted to the early warning system module. The threshold value of the early warning system module is dynamically adjusted according to the real-time data trend monitored by the threshold setting and adjustment module, and then the early warning system module issues an alarm according to the comparison result; Step 3: The data recording module records all data generated by the early warning work.

Images