Accelerometer fault detection method for rotary steerable drill tool attitude measurement device

Abstract

The application discloses a kind of accelerometer fault detection methods of rotary steerable drill posture measuring device, comprising: step S1, the Kalman filter of constructing rotary steerable drill posture measuring device system;Step S2, according to Kalman filter, the output estimation error of rotary steerable drill posture measuring device system and its digital features are calculated;Step S3, according to the output estimation error of rotary steerable drill posture measuring device system, residual error, evaluation function and threshold value are obtained;Step S4, according to residual error, evaluation function and threshold value, accelerometer fault detection decision is executed.The present application can realize the accurate detection of accelerometer fault under the condition that the probability distribution of noise is unknown, thereby effectively avoiding the rotary steerable drill along the wrong direction drilling, improve the reliability of rotary steerable drill posture measuring device, improve the work efficiency of rotary steerable drill.

Description

Technical Field

[0001] This invention belongs to the field of fault diagnosis, specifically relating to a fault detection method for the accelerometer of a rotary drilling tool attitude measurement device under unknown noise probability distribution. Background Technology

[0002] Rotary directional drilling systems are highly sophisticated directional drilling equipment, typically used in drilling operations in complex geological environments such as deep earth and deep sea. The rotary directional drilling tool is a crucial component of this system, its core task being to control the drill bit to drill along a predetermined attitude and direction. The rotary directional drilling tool's attitude closed-loop control system includes: an attitude measurement device, an attitude control device, and an attitude deviation directional device. The attitude measurement device is primarily responsible for converting real-time measurement data from attitude sensors such as gyroscopes and accelerometers into real-time attitude information of the drilling tool, which is then used for subsequent attitude control. Therefore, the accuracy of the attitude measurement device's calculation directly affects the attitude control performance, and consequently, the drilling efficiency of the rotary directional drilling system.

[0003] Accelerometers are essential sensors in the attitude measurement devices of rotary drilling tools. However, in complex geological environments such as deep earth and deep sea, accelerometers are often subjected to high temperatures, high pressures, and strong vibrations, making them highly susceptible to failure. According to the working principle of attitude measurement devices, an accelerometer malfunction can cause significant deviations in attitude calculations, leading to the drill bit drilling in the wrong direction and delaying the drilling progress. Therefore, it is urgent to design suitable fault detection methods to detect accelerometer malfunctions promptly and effectively.

[0004] Most existing methods for accelerometer fault detection in rotary steerable drilling tools assume that the noise distribution is known. For example, the literature "Accelerometer fault detection for rotary steerable drilling tool systems under strong noises" (Niu Yichun, Sheng Li, Gao Ming, et al. IEEE Transactions on Instrumentation & Measurement, 2022) and "Particle filter-based fault detection for Toolface measurement of rotary steerable systems" (Sheng Li, Niu Yichun, Gao Ming, et al. IEEE Transactions on Instrumentation & Measurement, 2023) have achieved accelerometer fault detection under the assumption that the noise follows a zero-mean Gaussian distribution.

[0005] However, rotary drilling tools typically operate in complex geological environments such as deep earth and deep sea. Factors such as geological structure, rock strata properties, and drilling depth all affect the probability distribution of noise, making it almost impossible to accurately establish a probability distribution model for noise. Therefore, it is necessary to develop a method for detecting accelerometer faults in the attitude measurement device of rotary drilling tools under unknown noise probability distribution conditions. This will ensure accurate detection of accelerometer faults even in complex and unknown drilling environments, thereby ensuring the drilling efficiency of the rotary drilling system. Summary of the Invention

[0006] To address the problems existing in the prior art, the present invention provides a method and apparatus for detecting accelerometer faults in a rotary drilling tool attitude measurement device.

[0007] To achieve the above objectives, the present invention provides the following solution:

[0008] A method for detecting accelerometer faults in a rotary drilling tool attitude measurement device includes:

[0009] Step S1: Construct the Kalman filter for the attitude measurement device system of the rotary drilling tool;

[0010] Step S2: Calculate the output estimation error and its digital characteristics of the rotary drilling tool attitude measurement device system based on the Kalman filter;

[0011] Step S3: Based on the output estimation error of the rotary drilling tool attitude measurement device system, obtain the residual, evaluation function and threshold;

[0012] Step S4: Based on the residual, evaluation function and threshold, execute the accelerometer fault detection decision.

[0013] Preferably, the Kalman filter constructed in step S1 is:

[0014]

[0015] in, express Sampling time The posterior estimate, express Sampling time The prior estimate, express The Kalman filter gain matrix at the sampling time;

[0016] when hour, and It is calculated using the following formula:

[0017]

[0018]

[0019] in, express Sampling time The posterior estimate, express The system coefficient matrix at the sampling time, express The prior estimation error covariance matrix at the sampling time.

[0020] Preferably, in step S2, the output estimation error of the Kalman filter is:

[0021]

[0022] in, express Sampling time measurement output The estimation error;

[0023] The numerical characteristics of the output estimation error are shown in the following formula:

[0024]

[0025]

[0026]

[0027] in, express Under fault-free sampling conditions The covariance matrix, express Sampling time fault When it happens The mean, express Sampling time fault When it happens The covariance matrix, express Sampling time fault The mean of the state estimation error when it occurs. express The List, Indicates a fault The mean, express Sampling time fault The covariance matrix of the state estimation error at the time of occurrence. Indicates a fault The covariance matrix, and These are known parameters.

[0028] As a preferred embodiment, in step S3, the residual, evaluation function, and threshold are as shown in the following formulas:

[0029]

[0030]

[0031]

[0032] in, , and They represent Sampling time is used to detect faults Residuals, evaluation function and threshold, It is the column vector to be calculated. These are the parameters to be calculated.

[0033] Preferably, in step S4, if If so, it is considered a fault. Occurs; if and If no fault occurs, it is considered that no fault has occurred.

[0034] Compared with the prior art, the beneficial effects of the present invention are as follows:

[0035] This invention enables accelerometer fault detection in the attitude measurement device of a rotary drilling tool under unknown noise probability distribution. It can design appropriate residuals and thresholds under unknown noise probability distribution to accurately detect the time of fault occurrence, thereby preventing the rotary drilling tool from drilling in the wrong direction. Attached Figure Description

[0036] To more clearly illustrate the technical solution of the present invention, the drawings used in the embodiments are briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0037] Figure 1 This is a flowchart of an accelerometer fault detection method according to an embodiment of the present invention;

[0038] Figure 2 For numerical simulation Trajectory graphs of the fault, evaluation function, and threshold at the time of occurrence;

[0039] Figure 3 For numerical simulation Trajectory graphs of the fault, evaluation function, and threshold at the time of occurrence;

[0040] Figure 4 An experimental platform for rotary steerable drilling tools;

[0041] Figure 5 These are the raw measurement data from the gyroscope used in the experiment;

[0042] Figure 6 These are the raw measurement data from the accelerometer during the experiment;

[0043] Figure 7 For the experiment Trajectory graphs of the fault, evaluation function, and threshold at the time of occurrence;

[0044] Figure 8 For the experiment Trajectory graph of fault, evaluation function and threshold when it occurs. Detailed Implementation

[0045] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0046] To make the above-mentioned objects, features and advantages of the present invention more apparent and understandable, the present invention will be further described in detail below with reference to the accompanying drawings and specific embodiments.

[0047] Example 1

[0048] like Figure 1 As shown, the present invention provides a method for detecting accelerometer faults in a rotary drilling tool attitude measurement device under unknown noise probability distribution, comprising the following steps:

[0049] Step 1: Construct a Kalman filter; the specific steps are as follows:

[0050] Step 1.1: Based on the working mechanism of the rotary drilling tool attitude measurement device, the system equation of the attitude measurement device is established as shown in formula (1):

[0051] (1);

[0052] Among them, subscript Indicates the sampling point. express The state variables of the attitude measurement device at the sampling time. express The state variables at the sampling time, express The measured output of the attitude measurement device at the sampling time. , express Zero-mean noise that is uncorrelated at each sampling time. and They represent and The covariance matrix, express Accelerometer malfunction at the sampling time The fault coefficient matrix, express The system coefficient matrix at the sampling time, for The coefficient matrix is ​​measured at the sampling time. Indicates the sampling interval. express Gyroscope measurement value at the sampling time The state variable at the initial sampling time follows a mean of . The covariance matrix is The unknown distribution;

[0053] Step 1.2: Construct the Kalman filter for system (1), as shown in equation (2):

[0054] (2);

[0055] in, express Sampling time The posterior estimate, express Sampling time The prior estimate, express The Kalman filter gain matrix at the sampling time. When hour, and The results are obtained by calculation using formulas (3) and (4):

[0056] (3);

[0057] (4);

[0058] in, express Sampling time The posterior estimate, express The system coefficient matrix at the sampling time, express The prior estimation error covariance matrix at the sampling time. When hour, The result is obtained by formula (5):

[0059] (5);

[0060] in, express Sampling time noise The covariance matrix, express The posterior estimation error covariance matrix at the sampling time. When hour, The result is obtained by formula (6):

[0061] (6)

[0062] in, express Kalman filter gain matrix at sampling time, express The prior estimation error covariance matrix at the sampling time;

[0063] Step 2: Calculate the output estimation error and its numerical characteristics. The specific steps are as follows:

[0064] Step 2.1: The output estimation error of the Kalman filter is shown in Equation (7):

[0065] (7);

[0066] in, express Sampling time measurement output The estimation error;

[0067] Step 2.2: Calculate the numerical characteristics of the output estimation error as shown in formulas (8)-(10):

[0068] (8);

[0069] (9);

[0070] (10);

[0071] in, express Under fault-free sampling conditions The covariance matrix, express Sampling time fault When it happens The mean, express Sampling time fault When it happens The covariance matrix, express Sampling time fault The mean of the state estimation error when it occurs. express The List, Indicates a fault The mean, express Sampling time fault The covariance matrix of the state estimation error at the time of occurrence. Indicates a fault The covariance matrix. and These are known parameters. and As shown in formulas (11) and (12):

[0072] (11);

[0073] (12);

[0074] Step 3: Design residuals, evaluation functions, and thresholds. The specific steps are as follows:

[0075] Step 3.1: The structure of the residual, evaluation function, and threshold is shown in formulas (13)-(15):

[0076] (13);

[0077] (14);

[0078] (15);

[0079] in, , and They represent Sampling time is used to detect faults Residuals, evaluation function and threshold, It is the column vector to be calculated. These are the parameters to be calculated;

[0080] Step 3.2: Calculated through the following steps and :

[0081] 1) Calculation and ,in, yes orthogonal subspaces .make and ;

[0082] 2) Initialization and ;

[0083] 3) Calculation

[0084] ;

[0085] ;

[0086] ;

[0087] ;

[0088] 4) Initialization , , and ;

[0089] 5) Calculation

[0090] ;

[0091] ;

[0092] ;

[0093] ;

[0094] ;

[0095] ;

[0096] ;

[0097] ;

[0098] 6) If ,make Otherwise, let (Return to step 5)

[0099] 7) If ,make and Otherwise, let Then return to step 3);

[0100] 8) Calculation and ;

[0101] 9) If and ,make , and Otherwise, let , Then return to step 3);

[0102] 10) Calculation ;

[0103] Step 4: Execute fault detection decisions, the specific steps are as follows:

[0104] like If so, it is considered a fault. Occurs; if and If no fault occurs, it is considered that no fault has occurred.

[0105] This invention proposes a method for accelerometer fault detection in a rotary drilling tool attitude measurement device under unknown noise probability distribution. This method employs a novel residual and threshold construction approach, enabling accurate fault detection even with unknown noise probability distributions. Compared to existing accelerometer fault detection methods, the method proposed in this invention maintains accurate fault detection even in complex and unknown downhole working environments.

[0106] To demonstrate the effectiveness and feasibility of the accelerometer fault detection method of the above-mentioned rotary drilling tool attitude measurement device, the present invention will be further explained below in conjunction with numerical simulation and experiments of the rotary drilling tool attitude measurement device.

[0107] Example 1:

[0108] The parameters of the rotary drilling tool attitude measurement device used are as follows: , , , , , , , .

[0109] The fault detection method described in this invention was used to detect accelerometer faults. The detection results are shown in [reference needed]. Figure 2 and 3 . Figure 2 Describes the fault The test results at the time of occurrence Figure 3 Describes the fault The detection results at the time of occurrence. As can be seen from the simulation diagram, when a fault occurs, the evaluation function designed in this invention will quickly exceed the threshold, and can effectively detect accelerometer faults even when the noise probability distribution is unknown.

[0110] To further verify the feasibility of the proposed algorithm, we used, for example... Figure 4 An experiment was conducted using the rotary drilling tool attitude measurement device shown, which simulated vibration interference during the drilling process using a vibration platform, leading to unknown downhole noise.

[0111] The sampling interval in the experiment was taken as The speed is set to The noise covariance matrix is ​​set to... , The remaining parameters are set to , , , .

[0112] Raw measurement data from gyroscopes and accelerometers, such as Figure 5 and 6 As shown. Figure 7 and 8 The results of the fault detection are presented. Figure 7 Describes the fault The test results at the time of occurrence Figure 8 Describes the fault The detection results at the time of occurrence. Therefore, the proposed method remains feasible and effective for experimental equipment.

[0113] The embodiments described above are merely preferred embodiments of the present invention and are not intended to limit the scope of the present invention. Various modifications and improvements made to the technical solutions of the present invention by those skilled in the art without departing from the spirit of the present invention should fall within the protection scope defined by the claims of the present invention.

Claims

1. A method for detecting accelerometer faults in a rotary drilling tool attitude measuring device, characterized in that, include: Step S1: Construct the Kalman filter for the attitude measurement device system of the rotary drilling tool; Step S2: Calculate the output estimation error and its digital characteristics of the rotary drilling tool attitude measurement device system based on the Kalman filter; Step S3: Based on the output estimation error of the rotary drilling tool attitude measurement device system, obtain the residual, evaluation function and threshold; Step S4: Based on the residuals, evaluation function, and threshold, execute the accelerometer fault detection decision; In step S1, the constructed Kalman filter is: in, express Sampling time The posterior estimate, express Sampling time The prior estimate, express The state of the rotary drilling tool attitude measurement device system at the sampling time. express The measured output of the attitude measurement device at the sampling time. for The coefficient matrix is ​​measured at the sampling time. express Kalman filter gain matrix at sampling time; when hour, and It is calculated using the following formula: in, express Sampling time The posterior estimate, express The system coefficient matrix at the sampling time, express The prior estimation error covariance matrix at the sampling time. Represents the measurement noise covariance matrix; In step S2, the output estimation error of the Kalman filter is: in, express Sampling time measurement output The estimation error; The numerical characteristics of the output estimation error are shown in the following formula: in, express Under fault-free sampling conditions The covariance matrix, express The posterior estimation error covariance matrix at the sampling time. express Sampling time fault When it happens The mean, express Sampling time fault When it happens The covariance matrix, express Sampling time fault The mean of the state estimation error when it occurs. express The List, Indicates a fault The mean, express Sampling time fault The covariance matrix of the state estimation error at the time of occurrence. Indicates a fault The covariance matrix, and These are known parameters; In step S3, the residual, evaluation function, and threshold are given by the following formulas: in, , and They represent Sampling time is used to detect faults Residuals, evaluation function and threshold, It is the column vector to be calculated. These are the parameters to be calculated.

2. The accelerometer fault detection method for the rotary drilling tool attitude measurement device as described in claim 1, characterized in that, In step S4, if If so, it is considered a fault. Occurs; if and If the condition is met, it is assumed that no fault has occurred.

Images