A method, device, equipment and readable storage medium for monitoring drilling operation

Abstract

The application discloses a drilling operation monitoring method, device and equipment and a readable storage medium, and is applied to the technical field of oil drilling. The method comprises the following steps: acquiring real-time motion data corresponding to an operation personnel when the operation personnel performs operation based on drilling equipment, and extracting real-time motion features of the real-time motion data; comparing the real-time feature data with a standard action model, so as to identify real-time action behaviors and perform abnormal action detection; in the case that abnormal action is detected, generating a control signal and an alarm signal corresponding to the real-time action behaviors; sending the alarm signal to an alarm device to perform abnormal alarm, and sending the control signal to the drilling equipment through an on-site industrial bus, so as to control the drilling equipment to perform emergency operation. The application has the technical effect that safety accidents can be effectively prevented, so that operation safety and quality are guaranteed.

Description

Technical Field

[0001] This application relates to the field of oil drilling technology, and in particular to a drilling operation monitoring method, apparatus, equipment, and readable storage medium. Background Technology

[0002] Oil drilling is a crucial part of the oil extraction process, and the safety and efficiency of drilling operations directly affect the success or failure of the entire oilfield development. However, the oil drilling environment is complex and variable, with harsh conditions such as high temperature, high pressure, and highly corrosive gases, requiring operators to perform high-intensity and high-precision operations under these conditions.

[0003] Traditional supervision of drilling operations relies primarily on manual monitoring and experience-based judgment. This approach has several drawbacks: manual monitoring has limited reaction speed, making it prone to accidents due to operational errors or delayed responses; manual monitoring is difficult to maintain continuously, as personnel are easily fatigued, affecting the accuracy of judgments; and experience-based judgment is highly subjective, with significant differences in experience and ability among different operators leading to inconsistent operating standards and making it difficult to guarantee consistent operational quality.

[0004] Utilizing sensor technology for real-time equipment status monitoring, data analysis for fault prediction, and automatic control technology for optimizing drilling parameters, these technologies have improved the automation and intelligence of drilling operations to some extent. However, these monitoring solutions, primarily focused on equipment status monitoring and parameter optimization, still cannot guarantee the safety and quality of the operation.

[0005] In conclusion, how to effectively solve problems such as monitoring of brick well operations is a technical problem that urgently needs to be solved by those skilled in the art. Summary of the Invention

[0006] The purpose of this application is to provide a drilling operation monitoring method, apparatus, equipment, and readable storage medium to ensure the safety and quality of drilling operations.

[0007] To solve the above-mentioned technical problems, this application provides the following technical solution: A drilling operation monitoring method includes: Acquire real-time motion data of the operators performing operations based on drilling equipment, and extract the real-time motion features of the real-time motion data; The real-time feature data is compared with the standard action model to identify real-time action behaviors and detect abnormal actions; If an abnormal action is detected, a control signal and an alarm signal corresponding to the real-time action behavior are generated; The alarm signal is sent to the alarm device to trigger an abnormal alarm, and the control signal is sent to the drilling equipment via the field industrial bus to control the drilling equipment to perform emergency operations.

[0008] Preferably, comparing the real-time feature data with a standard action model to identify real-time action behaviors and perform abnormal action detection includes: The temporal features in the real-time feature data are compared with the temporal features in the standard action model to identify real-time action behavior and detect whether the action is in place in the temporal dimension. The frequency domain features in the real-time feature data are compared with the frequency domain features in the standard action model to identify real-time action behavior and detect whether the action is smooth in the frequency domain dimension. The time-frequency domain features in the real-time feature data are compared with the time-frequency domain features in the standard action model to identify real-time action behavior and detect any unexpected mutations in the time-frequency domain dimension. By merging the real-time recognition results and anomaly detection results in the time domain, frequency domain, and time-frequency domain dimensions, we obtain the action behavior recognition results and the abnormal action detection results.

[0009] Preferably, establishing the standard action model includes: Obtain standard motion data corresponding to the standard operation performed by the operator based on the drilling equipment, and extract the standard motion features of the standard motion data; Establish a mathematical model of the human skeleton and muscles that matches the drilling operation scenario; The standard motion characteristics are applied to the mathematical model by simulating human movement to generate the standard motion model.

[0010] Preferably, acquiring standard motion data corresponding to the operator's standard operation based on the drilling equipment, and extracting standard motion features from the standard motion data, includes: Acquire standard motion data corresponding to the standardized operating actions performed by the operator based on the drilling equipment; wherein the standardized operating actions are at least one of the following: pressing down the brake lever with both hands, rotating the regulating valve with one hand, bending over to lift and carry the drill string, raising both arms to connect the pipeline, and standing steadily on the monitoring device; Extract the standard motion features corresponding to the standardized operation actions from the standard motion data; Accordingly, the standard motion characteristics are applied to the mathematical model by simulating human motion to generate the standard motion model, including: By simulating human movement, the standard motion characteristics are applied to the mathematical model to generate a standard motion model corresponding to at least one of the following standardized operating actions: pressing down the brake lever with both hands, rotating and adjusting the valve with one hand, bending over to lift and carry the drilling tool, raising both arms to connect the pipeline, and standing steadily to monitor the device.

[0011] Preferably, a mathematical model of the human skeleton and muscles matching the drilling operation scenario is established, including: The skeletal segments involved in the drilling operation scenario are defined as rigid linkages, and the joints involved in the drilling operation scenario are defined as constrained joints with specific degrees of freedom. Determine the mass and centroid position of the rigid link, establish a local coordinate system for each rigid body, and use a homogeneous transformation matrix or source number to determine the rotational and translational relationships between adjacent rigid bodies. Based on the rotation of the joints in the drilling operation scenario, the rotational degree of freedom constraints of the constrained joints are determined. The passive cushioning characteristics of key human cartilage and ligaments are simulated using a spring damping system. The muscle is abstracted into active contraction elements, series elastic elements, and parallel elastic elements using a muscle force model, and the virtual muscle origin is supported on a rigid skeleton model to obtain the digital model; wherein, the rigid skeleton model includes the rigid link and the constraint joint.

[0012] Preferably, the standard motion characteristics are applied to the mathematical model by simulating human motion to generate the standard motion model, including: Based on the motion trajectory and acceleration in the standard motion characteristics, and combined with the mathematical model, the internal torque and interaction force borne by the joint at an instant are deduced. By performing statistical averaging and line extraction on multiple internal torques and multiple interaction forces, a range of joint torques that corresponds to standardized operating actions, is reasonably subjected to force, conforms to ergonomics, and can be safely completed is obtained. The standard motion model is determined using the joint torque range.

[0013] Preferably, the control signal is sent to the drilling equipment via a fieldbus to control the drilling equipment to perform emergency operations, including: The control signals are sent to the programmable logic controller or hydraulic actuator at the bottom of the drilling rig via the field industrial bus to drive the relay to close or the valve to lock, and force the winch or large clamp equipment to decelerate, lock the action or stop in an emergency, so as to completely prevent safety accidents at the physical level.

[0014] A drilling operation monitoring device, comprising: The real-time motion acquisition module is used to acquire real-time motion data of the operator when performing operations based on drilling equipment, and to extract the real-time motion features of the real-time motion data. The action comparison module is used to compare the real-time feature data with the standard action model to identify real-time action behavior and detect abnormal actions; The control signal generation module is used to generate control signals and alarm signals corresponding to the real-time action behavior when an abnormal action is detected. The anomaly handling module is used to send the alarm signal to the alarm device for anomaly alarm, and to send the control signal to the drilling equipment through the field industrial bus to control the drilling equipment to perform emergency operations.

[0015] An electronic device, comprising: Memory, used to store computer programs; A processor is used to implement the steps of the above-described drilling operation monitoring method when executing the computer program.

[0016] A readable storage medium storing a computer program, which, when executed by a processor, implements the steps of the drilling operation monitoring method described above.

[0017] The method provided in this application embodiment acquires real-time motion data corresponding to the operation of the drilling equipment by the operator, and extracts real-time motion features from the real-time motion data; compares the real-time feature data with the standard motion model to identify real-time motion behavior and detect abnormal motion; in the case of detecting abnormal motion, generates control signals and alarm signals corresponding to the real-time motion behavior; sends the alarm signal to the alarm device to trigger an abnormal alarm, and sends the control signal to the drilling equipment through the field industrial bus to control the drilling equipment to perform emergency operations.

[0018] To ensure the safety and quality of operations, real-time motion data of the operators is acquired while they are working on the drilling equipment. Then, feature extraction is performed on this real-time motion data to obtain real-time motion characteristics. By comparing the real-time feature data with a standard motion model, the current real-time motion behavior can be identified and abnormal motions can be detected. Then, in the event of an anomaly detection, control and alarm signals corresponding to the real-time motion behavior are automatically generated. The alarm signal is sent to the alarm device to trigger an anomaly alarm, and the control signal is sent to the drilling equipment via the fieldbus to control the drilling equipment for emergency operation. In other words, this application monitors drilling operations not by focusing on the parameters and characteristics of the equipment itself, but by focusing on the actions and behaviors of the operators. By pre-setting a standard motion model, motion recognition and abnormal motion detection can be achieved. In the event of an anomaly, alarm and control signals can be generated. The alarm signal enables real-time alerting, and the control signal allows direct control of the drilling equipment for emergency operation, bypassing manual intervention, thereby preventing safety accidents and ensuring operational safety and quality.

[0019] Accordingly, embodiments of this application also provide drilling operation monitoring devices, equipment, and readable storage media corresponding to the above-described drilling operation monitoring method, which have the aforementioned technical effects, and will not be elaborated further here. Attached Figure Description

[0020] To more clearly illustrate the technical solutions in the embodiments or related technologies of this application, the accompanying drawings used in the description of the embodiments or related technologies will be briefly introduced below. Obviously, the accompanying drawings described below are only some embodiments of this application. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0021] Figure 1 This is a flowchart illustrating the implementation of a drilling operation monitoring method according to an embodiment of this application. Figure 2 This is a schematic diagram of the structure of a drilling operation monitoring device according to an embodiment of this application; Figure 3 This is a schematic diagram of the structure of an electronic device according to an embodiment of this application; Figure 4 This is a schematic diagram of the specific structure of an electronic device in an embodiment of this application. Detailed Implementation

[0022] To enable those skilled in the art to better understand the present application, the present application will be further described in detail below with reference to the accompanying drawings and specific embodiments. Obviously, the described embodiments are merely some embodiments of the present application, and not all embodiments. Based on the embodiments in this application, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of this application.

[0023] Please refer to Figure 1 , Figure 1 This is a flowchart of a drilling operation monitoring method according to an embodiment of this application. The method includes the following steps.

[0024] S101. Obtain real-time motion data of the operators when they are working on the drilling equipment, and extract the real-time motion features of the real-time motion data.

[0025] In this embodiment, motion data is collected using various sensors installed on the drilling equipment or the operator. Specifically, accelerometers, gyroscopes, magnetometers, and position sensors can be installed at key locations on the drilling equipment and the operator to collect motion data in real time. This sensor data is transmitted wirelessly to a data acquisition terminal for processing. Key locations refer to the operator's major joints (such as wrists, elbows, and knees) and the moving actuators of the drilling equipment. The specific installation locations and sensor types can be determined based on the motion characteristics to be captured. For example, gyroscopes can be used to monitor joint rotation, and accelerometers can be used to monitor arm movements or equipment vibrations.

[0026] Among them, drilling equipment refers to drilling-related equipment that requires human operation. For example, in the scenario of tripping in and out of the drilling rig, the hydraulic tongs are the corresponding drilling equipment; in the scenario of core control of the drilling rig, the winch brake lever is the corresponding drilling equipment; in the scenario of material transfer on the drilling site, heavy tools such as slips and short sections are the corresponding drilling equipment; and in the scenario of manifold operations, the pipeline being operated is the corresponding drilling equipment.

[0027] After acquiring real-time motion data, the time-series data collected by sensors can be processed through data cleaning, normalization, and data synchronization to ensure data reliability.

[0028] For example, the real-time motion data acquisition formula is: D(t)={a(t), ω(t), m(t), p(t)}, where D(t) is the motion dataset acquired at time t, including acceleration a(t), angular velocity ω(t), magnetic field data m(t), and position data p(t).

[0029] It should be noted that to accurately recreate the skeletal posture of the human body in three-dimensional space (e.g., whether the operator's arm is pointing directly in front of the drilling platform or 30 degrees to the left), magnetic field data is required to eliminate yaw drift of the gyroscope. With magnetic field data, the system can accurately know the absolute orientation of each joint point in three-dimensional space over a long period of time, thus ensuring extremely high accuracy in comparing standard motion models.

[0030] To facilitate action recognition and detection of abnormal actions, features can be extracted from real-time motion data to obtain more representative real-time feature data.

[0031] In one specific embodiment of this application, to make action recognition and abnormal action detection more accurate, real-time motion features are extracted from real-time motion data. Features can be extracted from different domains to obtain time-domain features, frequency-domain features, and time-frequency-domain features. Specifically, oil drilling sites are filled with machine vibration and noise. If only time-domain features are used, it is easy to misjudge normal machine vibrations as human violations. By adding frequency and time-frequency domain features, the system can act like an experienced safety officer, not only checking whether the action is performed correctly (time domain), but also whether the action is smooth (frequency domain), and whether there are any unexpected sudden changes (time-frequency domain), thereby making the most accurate action recognition and warning.

[0032] S102. Compare real-time feature data with standard action models to identify real-time action behaviors and detect abnormal actions.

[0033] In this embodiment, a standard motion model can be pre-constructed. This model is derived by modeling the standard motion data collected from workers performing standard operations. Different actions correspond to different standard motion models. That is, standard motion models for multiple standard actions can be pre-constructed according to actual conditions. Specifically, a large amount of standard motion data from operators can be collected, and then these high-dimensional time-series data can be processed by feature space dimensionality reduction and density peak optimization based on kinematic manifolds. Specifically, the system maps the extracted multi-dimensional features in the time domain, frequency domain, etc., to discrete coordinate points in a multi-dimensional state space. Using adaptive density search logic, it autonomously calculates the center of the coordinate cluster with the highest local density in the feature space (i.e., representing the envelope of the most typical and stable standard motion trajectory), thereby extracting the core dynamic benchmark parameters of common motion patterns, thus obtaining the standard motion model.

[0034] For example, oil drilling site operators vary in height, arm span, and force application habits, and the site is accompanied by strong mechanical vibrations and noise. If we collect the braking actions of 100 workers, the data will present 100 curves with similar shapes but varying details. By employing density-based unsupervised optimization (the underlying logic of clustering), the system can automatically ignore outliers caused by individual differences or accidental vibrations, autonomously converge, and extract a core force application habit and posture trend (mathematical expectation value) common to the vast majority of experienced workers. This makes the established action standard model both universally applicable and noise-resistant.

[0035] After obtaining real-time feature data, the real-time feature data can be compared with the standard action model to achieve real-time action behavior recognition and abnormal action detection.

[0036] In one specific embodiment of this application, comparing real-time feature data with a standard action model to identify real-time action behavior and perform abnormal action detection includes: By comparing the temporal features in real-time feature data with the temporal features in the standard action model, real-time action behavior can be identified and whether the action is in place in the temporal dimension. By comparing the frequency domain features in real-time feature data with the frequency domain features in the standard action model, real-time action behavior can be identified and the smoothness of the action can be detected in the frequency domain dimension. By comparing the time-frequency domain features in real-time feature data with the time-frequency domain features in the standard action model, real-time action behavior can be identified in the time-frequency domain dimension and the presence of unexpected mutations can be detected. By merging the real-time recognition results and anomaly detection results in the time domain, frequency domain, and time-frequency domain dimensions, we obtain the action behavior recognition results and the abnormal action detection results.

[0037] In other words, in this embodiment, when comparing real-time feature data with the standard action model, comparisons can be performed separately based on different domains, thereby achieving action behavior recognition and detection in different domain dimensions. Finally, the recognition results of the three dimensions and the abnormal action detection results are integrated to obtain the action behavior recognition results and the abnormal action detection results.

[0038] Specifically, for action recognition, a voting system can be used. For example, if all three results point to the same action, that action is output; if two results are the same, the same result is output. For abnormal action detection results, all abnormal actions can be output to avoid omissions.

[0039] In one specific embodiment of this application, establishing a standard action model includes: Acquire standard motion data corresponding to the standard operations performed by operators based on drilling equipment, and extract standard motion features from the standard motion data; Establish a mathematical model of the human skeleton and muscles that matches the drilling operation scenario; By simulating human movement, standard motion characteristics are applied to a mathematical model to generate a standard motion model.

[0040] This includes acquiring standard motion data corresponding to the standard operations performed by the operators based on the drilling equipment, and extracting the standard motion features of the standard motion data, including: Acquire standard motion data corresponding to the standardized operating actions performed by the operators based on the drilling equipment; wherein the standardized operating actions are at least one of the following: pressing down the brake lever with both hands, rotating the regulating valve with one hand, bending over to lift and carry the drill string, raising both arms to connect the pipeline, and standing steadily on the monitoring device; Extract standard motion features corresponding to standardized operational movements from standard motion data; Accordingly, standard motion characteristics are applied to a mathematical model by simulating human movement to generate a standard motion model, including: By simulating human movement, standard motion characteristics are applied to mathematical models to generate standard motion models corresponding to at least one of the following standardized operating actions: pressing down the brake lever with both hands, rotating and adjusting the valve with one hand, bending over to lift and carry the drilling tool, raising both arms to connect the pipeline, and standing steadily to monitor the instrument.

[0041] Among these, establishing a mathematical model of the human skeleton and muscles that matches the drilling operation scenario includes: The skeletal segments involved in drilling operations are defined as rigid links, and the joints involved in drilling operations are defined as constrained joints with specific degrees of freedom. Determine the mass and center of mass of the rigid link, and establish a local coordinate system for each rigid body. Use homogeneous transformation matrix or source number to determine the rotational and translational relationships between adjacent rigid bodies. Based on the rotation of joints in drilling operations, determine the rotational degree of freedom constraints of the joints. The passive cushioning characteristics of key human cartilage and ligaments are simulated using a spring damping system. By using a muscle exertion model, muscles are abstracted into active contraction elements, series elastic elements, and parallel elastic elements, and the virtual muscle origin is supported on a rigid skeleton model to obtain a digital model; the rigid skeleton model includes rigid links and constrained joints.

[0042] This includes applying standard motion characteristics to a mathematical model by simulating human movement to generate a standard motion model, including: Based on the motion trajectory and acceleration in the standard motion characteristics, and combined with mathematical models, the internal torque and interaction force that the joint experiences instantaneously are deduced. By statistically averaging and extracting the retention lines of multiple internal torques and multiple interaction forces, the range of joint torques corresponding to standardized operating actions that are reasonably subjected to force, conform to ergonomics, and can be safely completed is obtained; The standard motion model is determined by utilizing the joint torque range.

[0043] For ease of description, the establishment of the standard motion model will be explained in detail below.

[0044] First, accelerometers, gyroscopes, magnetometers, and position sensors are installed on key parts of drilling equipment and personnel to collect standard motion data in real time. Key parts of the drilling equipment include winch brake handles, hydraulic tongs operating handles, slip handles, and regulating valves. Key parts of the personnel include areas of exertion, important joints, and limbs. It is important to note that the types and installation locations of the sensors used to collect standard motion data and those used to collect real-time motion data can be identical to facilitate subsequent action identification and anomaly detection.

[0045] This sensor data can be transmitted wirelessly to a data acquisition terminal for processing. The specific implementation steps include: installing sensors at key locations on the drilling equipment and personnel, and transmitting standard motion data from each sensor to the data acquisition terminal in real time via a wireless transmission module.

[0046] Next, the collected standard motion data undergoes preprocessing, including data cleaning, normalization, and data synchronization. Data cleaning uses filtering algorithms to remove noise and outliers; data normalization standardizes the data from each sensor to eliminate the influence of different units of measurement; and data synchronization employs a time synchronization algorithm to ensure the consistency of data from multiple sensors. The specific implementation steps include: using filtering algorithms to remove noise and outliers, normalizing the data, and using time synchronization algorithms for data synchronization. The data cleaning and normalization formulas are as follows: in This is the cleaned data; Filter is the filtering algorithm. These are the normalized data, where μ and σ are the mean and standard deviation of the data, respectively.

[0047] Multiple features are extracted from preprocessed standard motion data, including time-domain features, frequency-domain features, and time-frequency-domain features. Time-domain features are obtained by calculating the mean, standard deviation, and root mean square (RMS). Frequency-domain features, such as the dominant frequency and spectral energy, are extracted using Fourier transform. Time-frequency-domain features are extracted using methods such as wavelet transform to analyze the changes in data at different times and frequencies. The specific implementation steps include: calculating time-domain features such as the mean, standard deviation, and RMS; performing Fourier transform to extract frequency-domain features such as the dominant frequency and spectral energy; and using wavelet transform to extract time-frequency-domain features. The feature extraction formula is as follows: , , - ; in, It is a time-domain feature, which is obtained by analyzing the normalized data. Calculate the mean ( ), standard deviation ) and root mean square ( ), to quantify the basic physical manifestation of actions in the time domain; These are frequency domain features, used to apply the Fast Fourier Transform to the normalized data. The timing signal is converted to the frequency domain to extract the main frequency component and spectral energy of the movement, which is used to evaluate the periodicity, speed and smoothness of human movements. These are time-frequency domain features, which are used to utilize wavelet transform ( By jointly expanding motion data in both time and frequency dimensions, it is possible to accurately locate and extract local high-frequency oscillations or unexpected force abrupt changes during the action process.

[0048] By using bionics technology to simulate human movement patterns, standard action models are created. That is, a large amount of standard movement data of operators is collected, and then statistical analysis is performed on this standard movement data (for example, using clustering algorithms (such as K-means algorithm) or machine learning methods such as principal component analysis (PCA) to cluster and reduce the dimensionality of a large amount of collected normal movement feature data (time domain, frequency domain, etc.)) to extract common action patterns (both hands pressing down the brake lever to control the drilling rig), one hand rotating the regulating valve, bending over to lift and carry the drilling tool, both arms raising to connect the pipeline, and standing steadily to monitor the instrument, etc.).

[0049] Next, a mathematical model of the human skeleton and muscles is established to simulate human movement. The extracted motion patterns are then applied to the mathematical model to generate a standard motion model. Specific implementation steps include: establishing a mathematical model of the human skeleton and muscles: based on multi-rigid-body dynamics, the human body is abstracted into a three-dimensional skeletal topology composed of rigid links and constrained joints, and a spring-damping system or Hill's muscle model is introduced to simulate muscle force generation mechanisms; simulating human movement: based on the Newton-Euler dynamics equations, the extracted motion feature parameters are input into the model to calculate the forces and motion trajectories of each joint, thereby generating a standard motion model. The bionic model formula is as follows: ; in, It is a standard action model, while Biomechanics is a bionic mathematical model. These are motion data from normal operation.

[0050] Specifically, the core logic can be constructed based on a three-dimensional skeletal model of multi-rigid-body dynamics: the human body is extremely complex, making direct force calculations impossible. The core logic of multi-rigid-body dynamics simplifies this complexity by treating the torso, thighs, and lower legs as rigid, non-deformable links, and the elbows and knees as constrained joints with specific degrees of freedom (such as hinges and ball joints). By constructing kinematic chains, the topological skeleton of the human body in three-dimensional space is established.

[0051] Implementation steps: Identify the key skeletal segments involved in the drilling operation (such as the spine, upper arm, forearm, thigh, and lower leg), and define the mass (m) and center of mass position of each rigid body. Establish a local coordinate system for each rigid body, and use homogeneous transformation matrices or quaternions to define the rotational and translational relationships between adjacent rigid bodies.

[0052] Define rotational degree of freedom constraints for each joint (e.g., the knee joint can only flex and extend in the sagittal plane, with 1 degree of freedom; the shoulder joint can rotate in three dimensions, with 3 degrees of freedom). Corresponding data structure / output: Topological structure data: Set of rigid body geometric parameters. , The length of the link. For quality, (For inertia tensor). Kinematic state data: joint angle vectors and three-dimensional spatial coordinate matrix ,in, Indicates time The system state vector formed by the angles of various joints in the human body; These represent the first to the second elements in the skeletal topology. The actual rotation angle or relative flexion-extension value of a constrained joint (such as wrist, elbow, knee, etc.) at the current moment; This indicates the total number of dimensions of the constrained joints in the human skeleton model during drilling operations; superscript. This represents the matrix transpose operation, which transposes the horizontally arranged angular elements into column vectors so that they can be directly substituted into subsequent Newton-Euler equations for matrix operations.

[0053] The algorithm employs a core logic incorporating a spring-damping system and the Hill muscle model: skeletons alone cannot explain force generation and cushioning. The spring-damping system simulates the passive cushioning characteristics of human joint cartilage and ligaments; while the Hill muscle model is the most classic empirical model of muscle force generation in biomechanics, abstracting the muscle into three parallel / series components: active contractile element (Ce, representing muscle fiber force generation), series elastic element (SEE, representing tendon), and parallel elastic element (Pee, representing the passive tensile force of connective tissue). This algorithm aims to calculate the actual internal tension output by the muscle to maintain a certain posture or movement. Implementation steps: Attach the virtual muscle origin to the rigid skeleton model established in step one.

[0054] The current muscle length is calculated using the extracted sensor data (velocity, displacement). and contraction speed .

[0055] Input the muscle activation parameter a(t), and calculate the dynamic tension of a single muscle group at this moment using the Hill equation.

[0056] Corresponding database structure: Muscle force data: Total muscle tension Joint impedance data: joint stiffness coefficient K and damping coefficient C (reflecting the degree of stiffness or relaxation in movement).

[0057] The core logic of generating standard motion models based on Newton-Euler dynamics equations: Newton-Euler equations are classical mechanics equations describing the translation of rigid bodies (Newton's laws). ) and rotation (Euler's equations) The core equation of motion capture is inverse dynamics. In motion capture, this primarily utilizes inverse dynamics: given the motion trajectory and acceleration (measured by sensors), it inversely calculates the internal torque and interaction forces experienced by each joint at that instant. The implementation steps are as follows: The standard acceleration a(t), angular velocity ω(t), and angular acceleration a(t) exhibited by a normal operator are input into the equation as known variables. Starting from the peripheral end (such as the contact force generated at the brake lever), each link is recursively calculated towards the body center (torso coordinate system), calculating the force (F) and torque (t) on each joint. ) Statistical averaging and envelope extraction are performed on standard motion data from a large number of operators to find the joint torque range that is most reasonable in terms of force distribution, most ergonomic, and allows for safe operation. The corresponding data structure / output (i.e., the final standard motion model) is then generated. Standard dynamic boundary (envelope matrix): The core output is the standard torque timing boundary of each key joint. ,in, : Indicates the torque of a standard critical joint. : Indicates time At this instant, the minimum safe torque boundary (lower edge of the envelope) required to complete the standard action is determined. If the actual measured torque is lower than this boundary, the system can determine that the force is insufficient, the grip is weak, or the operation is perfunctory. Indicates time At this instant, the maximum safe torque boundary (upper edge of the envelope) for the operator's joint to exert force or bear reaction force. If the actual measured torque exceeds this upper limit, the system can determine that the force is excessive, the posture is distorted, or the operation is too violent. Such situations can easily cause acute lumbar sprains, joint dislocations, or cause drilling equipment to go out of control due to sudden changes in force.

[0058] Standard kinematic feature set: a set of reference values ​​for the three-dimensional feature vectors of compliant movements. In order to enable real-time detection of stress on worker joints If the angle exceeds the safety boundary matrix calculated using the Newton-Euler equations, the system can determine that the force f is illegal or dangerous, thus detecting abnormal action.

[0059] After constructing a standard motion model, the real-time motion data collected in real time and the extracted real-time feature data can be compared with the standard motion model to identify the motion behavior.

[0060] Specifically, by calculating the matching degree, specific actions can be identified, and these actions are compared with standard action models in a pre-set action library (e.g., standard two-handed pushing and pulling of a large tong in drilling operations, bending over to handle non-standard forces, and a set of standard and abnormal actions such as smoothly pressing down the brake lever in brake control scenarios) to confirm the action category. The specific implementation steps include: performing matching analysis between real-time feature data and standard action models, calculating the matching degree, identifying specific actions, and comparing the identified actions with a pre-set action library to confirm the action category.

[0061] The matching analysis formula is: Match( ; where Match is the matching analysis function, It is real-time collected motion data. It is a standard action model.

[0062] For example, real-time motion data is in the form of matrix / vector expressions, and the motion feature data is extracted in real time. The system presents these features as multidimensional feature vectors (in sequence form). Specific data items include time-domain, frequency-domain, and time-frequency-domain features within a time window. The formula is expressed as: ,in, arrive These represent different feature extraction items, such as: root mean square value of hand acceleration, angle of waist tilt, dominant frequency energy of wrist movement, joint angular velocity, etc.

[0063] Different action expressions in the standard action model It is an action library containing multiple preset action templates, each with a specific action model. (k is the action category number) is derived from the standard feature center vector of this action. Feature weight matrix (representing the degree of importance of different features in the judgment of this action) constitutes the composition, i.e. .

[0064] Specific example: Action 1 (Smoothly press down the brake lever with both hands) The weight matrix assigns significant weights to the spatial overlap of the hands and the smoothness of the downward angular velocity. Its standard motion model is expressed as: ,in The eigenvalues ​​in the model define low-frequency smooth changes without instantaneous acceleration abrupt changes.

[0065] Action 2 (Standard Two-Handed Push-Pull Pliers) The weighting emphasizes the symmetry of force distribution and the average lateral displacement of the left and right arms. Its standard movement model is expressed as: This limits the consistency of force exerted by both hands.

[0066] Action 3 (Bending over and lifting non-standard loads) (Abnormal movements): Weighting is primarily assigned to the lumbar lordosis angle and knee flexion angle. The standard movement model is expressed as: When the input feature vector shows that the waist increases while the knee angle remains unchanged, it closely matches the abnormal model.

[0067] The specific expression of the Match analysis function: The Match function uses weighted Euclidean distance or dynamic time warping algorithm to calculate the similarity distance between the real-time feature vector and the standard model. Taking weighted Euclidean distance as an example, the smaller the distance, the higher the matching score. The matching distance formula is as follows: The matching score formula is as follows: The final identification result category The action category with the highest score that exceeds the set safety threshold of 0: .in, Represents the real-time motion feature vector and the first The feature difference distance between standard action models (calculated using weighted Euclidean distance; the smaller the value, the smaller the difference). Indicates real-time motion data and the first The matching score of a standard action model. The score is calculated by the inverse of the distance and normalized to the range (0,1]. The smaller the distance, the closer the score is to 1 (i.e., the higher the matching degree). This indicates the specific action category result that the system ultimately determines and outputs after comparison.

[0068] This represents the total number of dimensions (i.e., the total number of feature terms) of the extracted multidimensional motion features. Represents the eigenvector of the th element. Feature Dimension Index . Represents real-time motion feature vectors The first in Specific feature values ​​(e.g., the first) The value is the real-time wrist acceleration. Indicates the first In the standard action model, the first The standard reference value of a feature (i.e., the feature center point of that dimension). Indicates the first In the standard action model, the first The weight coefficient corresponding to each feature. The larger this value, the more significant the weight coefficient. Feature in determining the first The more important each action is.

[0069] Indicates time Real-time extraction of multidimensional motion feature vector matrix of operators. Indicates the first in the preset action library A standard action model.

[0070] This indicates that during the traversal of all... After (all action models), find the function value within the parentheses (i.e.) When it reaches its maximum, the corresponding model number . This represents the system's preset safety matching threshold. Only when the maximum matching score exceeds this threshold will the system be certain that the action has been recognized, thereby filtering out invalid random movements or interfering noise.

[0071] S103. In the event of an abnormal action, generate control signals and alarm signals corresponding to the real-time action behavior.

[0072] When abnormal actions are detected, control signals and alarm signals corresponding to the real-time actions can be generated.

[0073] Among them, the control signal is the signal that controls the drilling equipment corresponding to the real-time action; the alarm signal is the signal that alarms the personnel, monitoring platform or alarm device corresponding to the real-time action.

[0074] For example, when the identification result (such as the matching score or joint force parameters) crosses the preset safety red line or matches the preset high-risk abnormal action model, the system will immediately trigger an interrupt response and automatically compile and generate a digital control instruction package in a specific format as the corresponding control signal and alarm signal.

[0075] S104. Send the alarm signal to the alarm device to trigger an abnormal alarm, and send the control signal to the drilling equipment via the field industrial bus to control the drilling equipment to perform emergency operations.

[0076] In other words, the alarm signal is transmitted to the alarm device to trigger an alarm for an abnormality.

[0077] Control signals are sent to the corresponding drilling equipment to enable emergency control of the drilling equipment.

[0078] In other words, a dual-link approach can be used to send control and alarm signals in parallel: on the one hand, the signal is instantly transmitted to the smart wearable device worn by the operator (triggering vibration or buzzer alarm) and the on-site sound and light warning system via low-latency wireless radio frequency communication (such as 5G or industrial Wi-Fi) to remind the operator to make adjustments; on the other hand, for high-risk violations at the active safety intervention level (e.g., detecting the driller's one-handed brake slip), the control signal will bypass the manual confirmation process and be directly sent to the drilling equipment with millisecond-level low latency via the field industrial bus (such as EtherCAT or Modbus protocol) to control the drilling equipment.

[0079] In one specific embodiment of this application, a control signal is sent to the drilling equipment via a fieldbus to control the drilling equipment to perform emergency operations, including: Control signals are sent to the programmable logic controller or hydraulic actuator at the bottom of the drilling rig via a fieldbus to drive relays to close or valves to lock, forcing the winch or tongs to decelerate, lock actions, or stop in an emergency, thus completely preventing safety accidents at the physical level.

[0080] Specifically, data can be sent directly to the programmable logic controller (PLC) or hydraulic actuator at the bottom of the drilling rig with millisecond-level low latency via fieldbus (such as EtherCAT or Modbus protocol), driving relays to close or valves to lock, forcing related winches, tongs and other equipment to decelerate, lock actions or stop in an emergency, thereby completely preventing the occurrence of safety accidents at the physical level.

[0081] The method provided in this application embodiment acquires real-time motion data corresponding to the operation of the drilling equipment by the operator, and extracts real-time motion features from the real-time motion data; compares the real-time feature data with the standard motion model to identify real-time motion behavior and detect abnormal motion; in the case of detecting abnormal motion, generates control signals and alarm signals corresponding to the real-time motion behavior; sends the alarm signal to the alarm device to trigger an abnormal alarm, and sends the control signal to the drilling equipment through the field industrial bus to control the drilling equipment to perform emergency operations.

[0082] To ensure the safety and quality of operations, real-time motion data of the operators is acquired while they are working on the drilling equipment. Then, feature extraction is performed on this real-time motion data to obtain real-time motion characteristics. By comparing the real-time feature data with a standard motion model, the current real-time motion behavior can be identified and abnormal motions can be detected. Then, in the event of an anomaly detection, control and alarm signals corresponding to the real-time motion behavior are automatically generated. The alarm signal is sent to the alarm device to trigger an anomaly alarm, and the control signal is sent to the drilling equipment via the fieldbus to control the drilling equipment for emergency operation. In other words, this application monitors drilling operations not by focusing on the parameters and characteristics of the equipment itself, but by focusing on the actions and behaviors of the operators. By pre-setting a standard motion model, motion recognition and abnormal motion detection can be achieved. In the event of an anomaly, alarm and control signals can be generated. The alarm signal enables real-time alerting, and the control signal allows direct control of the drilling equipment for emergency operation, bypassing manual intervention, thereby preventing safety accidents and ensuring operational safety and quality.

[0083] Specifically, the drilling operation monitoring method provided in this application embodiment can achieve the following beneficial technical effects: (1) Utilize time-domain, frequency-domain, and time-frequency-domain feature extraction to comprehensively analyze motion data. These feature extraction techniques can capture multi-dimensional information from the data, providing a richer and more accurate feature set, and improving the accuracy of action behavior recognition. This is more comprehensive and efficient than using only a single feature extraction technique in related methods.

[0084] (2) Using bionic technology to simulate human movement patterns and create standard action models. By collecting a large amount of movement data from normal operators, a mathematical model of the human skeleton and muscles is established and applied to the generation of standard action models. This bionic technology can accurately simulate and understand human movement, providing a scientific basis for action behavior recognition, and is more biologically reasonable than a simple mathematical model.

[0085] (3) Multivariate comprehensive analysis: Environmental parameters are monitored in real time through a sensor network and combined with motion data for comprehensive analysis. Multivariate analysis is used to evaluate the impact of environmental parameters on motion behavior. This comprehensive analysis can more accurately reflect the complex situation in the actual operating environment and is more comprehensive and accurate than the approach of ignoring environmental factors in related methods.

[0086] (4) Based on the identification results, the drilling operation is monitored in real time, and corresponding control signals are generated. By comparing the actions and operating procedures in real time, control signals are generated immediately when abnormal behavior is detected, and operators are alerted or equipment is adjusted. Real-time monitoring and intelligent control mechanisms can effectively prevent accidents and improve operational safety, which is more timely and effective than post-event analysis and manual intervention.

[0087] (5) By combining advanced technologies such as bionics, the Internet of Things, and blockchain, a comprehensive solution for motion data acquisition, processing, analysis, and control is provided. Compared with traditional single-technology applications, this invention achieves a higher level of intelligence and comprehensive performance through the integration of multiple cutting-edge technologies.

[0088] It should be noted that, based on the above embodiments, the embodiments of this application also provide corresponding improvement schemes. In the preferred / improved embodiments, the same or corresponding steps as in the above embodiments can be referred to each other, and the corresponding beneficial effects can also be referred to each other; however, these will not be elaborated upon in the preferred / improved embodiments herein.

[0089] In one specific embodiment of this application, a sensor network can be used to monitor environmental parameters in real time and perform comprehensive analysis in conjunction with motion data; the comprehensive analysis results are then transmitted to a central control system for processing and storage.

[0090] In other words, environmental parameters are monitored in real time through a sensor network and comprehensively analyzed in conjunction with motion data. Environmental sensors monitor environmental parameters such as temperature, humidity, air pressure, and concentration of harmful gases. This environmental parameter data is used as a feature for auxiliary optimization and is comprehensively analyzed along with motion data. Multivariate analysis methods are used to assess the impact of environmental parameters on behavioral patterns, thereby dynamically adjusting the threshold for identifying abnormal movements and the alarm level. For example, when a high-temperature and high-humidity environment is detected, if the system detects a decrease in the amplitude or speed of the operator's movements, it comprehensively determines that this is an environment-induced fatigue state; when the system detects that the concentration of harmful gases exceeds the standard and detects a person falling to the ground, it comprehensively determines that this is a poisoning emergency.

[0091] The specific implementation steps include: installing environmental sensors, monitoring environmental parameters in real time, comprehensively analyzing environmental parameter data and motion data, and evaluating the impact of environmental parameters on motion behavior through multivariate analysis methods.

[0092] The environmental monitoring and analysis process involves the system distributing high-precision environmental sensor nodes across key areas of the drilling platform. These sensors continuously and cyclically sample environmental parameters such as temperature, humidity, air pressure, mechanical vibration and noise, and the concentration of harmful gases at a set frequency. In the comprehensive analysis phase, the system introduces a multi-source data fusion algorithm to construct a data combination function. This function uses real-time environmental parameters as context variables to dynamically adjust the confidence weights of motion characteristic data or directly correct the threshold for judging abnormal actions. For example, when environmental sensors detect a sustained high-temperature and high-humidity environment, the data combination function automatically lowers the threshold for judging fatigue. If it detects a slight decrease in the operator's movement amplitude or a slowing response speed, the system comprehensively judges this as environmentally induced early signs of fatigue and issues an early warning. Similarly, when the concentration of harmful gases (such as hydrogen sulfide) approaches a critical value, and the spatial posture characteristics show a sudden drop in the operator's center of gravity (such as falling), the data combination function outputs the highest-level comprehensive danger signal, directly judging it as a poisoning emergency and triggering a global coordinated rescue operation.

[0093] Comprehensive analysis formula: ;in, It is a comprehensive analysis result. It is a data combination function. These are environmental parameter data.

[0094] Among them, the environmental parameter E(t) in the comprehensive analysis formula specifically refers to the external physical variables that are collected in real time at the drilling site and may directly affect the accuracy of sensors or the action status of operators, such as the high-frequency mechanical vibration frequency of the drilling platform, extreme temperature and humidity, and background data such as deck wind force and slipperiness.

[0095] Among them, the data combination function Combine is a multi-source information fusion algorithm (such as adaptive Kalman filtering or dynamic weight allocation mechanism that introduces environmental factors). Its core function is to dynamically adjust the normalized motion data according to the intensity of environmental parameters. The confidence weights for each dimension (e.g., when severe platform vibration is detected, the algorithm automatically reduces the weight of severely disturbed acceleration features, or cancels out environmental background noise); the comprehensive analysis results output after fusion processing by this function. In essence, it is a high-precision multi-dimensional feature data vector or matrix after environmental context compensation and denoising. It accurately restores the operator's absolutely pure real motion state after removing complex environmental interference, thus serving as the most reliable data base input to the subsequent standard action model for accurate comparison and judgment.

[0096] Finally, the comprehensive analysis results are transmitted in real time to the cloud-based central control system via IoT technology for processing and storage. The central control system further processes the analysis results to generate reports and logs (reports mainly include statistics on operator behavior, analysis of abnormal action frequency, and overall operational safety assessment results; logs mainly refer to business logs related to action recognition, detailing the time of specific actions, the identified behavior category, the matching degree, and the entire process of triggered warnings and control signals). Blockchain technology is used to securely store and manage the data, ensuring its integrity and immutability. Specific implementation steps include: transmitting the comprehensive analysis results in real time to the cloud-based central control system via IoT technology; further processing the analysis results in the central control system to generate reports and logs; and using blockchain technology to securely store and manage the data, ensuring its integrity and immutability.

[0097] The specific transmission and storage process is as follows: The edge computing gateway on-site first encrypts and encapsulates the integrated analysis results, triggered control command status, and raw data of key time slices according to standardized IoT communication protocols (such as MQTT or OPC UA), constructing a structured data payload with timestamps. This payload is then securely uploaded to the cloud-based central control system via a dedicated network or 5G communication base station. In the central control system, the data parsing engine unpacks the received payload and automatically generates visual reports and business logs. (The reports mainly include statistics on operator behavior, analysis of abnormal action frequency, and overall operational safety assessment results; the logs record in detail the time of specific actions, the identified behavior category, the matching degree, and the entire process of triggering warnings and equipment intervention.) To ensure the legal validity and data credibility of traceability, the system further extracts data fingerprints from the generated key logs and reports using a hash algorithm (such as SHA-256) and records them on the blockchain as smart contract transactions. The distributed ledger and multi-node consensus mechanism of blockchain technology are used for solidified storage, ensuring that all action identification records and equipment intervention data have extremely high transparency, traceability, and absolute immutability.

[0098] In other words, blockchain technology is used to securely store and manage data, ensuring its integrity and immutability. Comprehensive analysis results are transmitted in real-time to a cloud-based central control system for processing and storage via IoT technology. This data security management method is more secure and reliable than traditional centralized data storage and management methods, and offers greater transparency and trustworthiness.

[0099] To facilitate a better understanding and implementation of the drilling operation monitoring method provided in this application by those skilled in the art, the following detailed explanation of the drilling operation monitoring method is provided with specific scenario examples.

[0100] For core drilling rig control scenarios, both hands must smoothly press down on the brake levers: the winch brake levers control the suspension and descent of the hundreds-ton drill string. Extracting this standard action and establishing a standard action model aims to confirm in real time that the driller must have both hands in position and apply force smoothly. This is to strictly prevent catastrophic accidents such as drill slippage or severe derrick vibration caused by one hand slipping off the lever (slipping brake) or sudden changes in force (abrupt braking and sudden release).

[0101] In drilling operations, the standard two-handed push-pull maneuver involves hydraulic tongs with significant rotational inertia when uncoupling on the tubing string. Extracting this standard movement and establishing a model aims to ensure the operator's chassis is stable and that both hands work in coordination. This prevents careless single-handed operation from causing the tongs to suddenly rebound and lose engagement, directly resulting in severe mechanical crush injuries to the operator's chest or arms.

[0102] In deck material transfer scenarios, non-standard bending and force handling (abnormal movement interception): Heavy tools such as slips and short sections are frequently moved at drilling sites. The purpose of extracting and monitoring these standard movements and establishing a standard movement model is to capture abnormal characteristics such as insufficient knee squatting and excessive forward tilting of the lumbar spine, so as to correct the worker's incorrect force exertion posture in a timely manner and prevent the most common high-risk acute lumbar sprains or chronic occupational diseases among drilling workers from the source.

[0103] In a manifold operation scenario, where personnel raise both arms to connect pipelines, extracting this action and establishing a standard motion model is primarily for monitoring the muscle strain limits of personnel working in upper spaces. By analyzing high-frequency tremor signals in the time-frequency domain characteristics, the system can determine whether the personnel's shoulders have reached the fatigue threshold, thus issuing timely shift change warnings before secondary accidents such as tools slipping from their hands and falling from height occur.

[0104] In other words, in practical applications, different standard action models can be established according to the actual needs of the scenario, thereby achieving millisecond-level microscopic perception of personnel's illegal exertion of force, dangerous posture, and fatigue operation at the drilling site, upgrading passive post-accident investigation to proactive pre-accident physical prevention.

[0105] Corresponding to the above method embodiments, this application also provides a drilling operation monitoring device. The drilling operation monitoring device described below can be referred to in correspondence with the drilling operation monitoring method described above.

[0106] See Figure 2 As shown, the device includes the following modules: The real-time motion acquisition module 101 is used to acquire real-time motion data corresponding to the operation of the drilling equipment by the operator, and to extract the real-time motion features of the real-time motion data. The action comparison module 102 is used to compare real-time feature data with standard action models to identify real-time action behaviors and detect abnormal actions. The control signal generation module 103 is used to generate control signals and alarm signals corresponding to real-time action behavior when abnormal actions are detected. The anomaly handling module 104 is used to send alarm signals to the alarm device for anomaly alarm, and to send control signals to the drilling equipment through the field industrial bus to control the drilling equipment to perform emergency operations.

[0107] Using the apparatus provided in this application embodiment, real-time motion data corresponding to the operator's work based on drilling equipment is acquired, and real-time motion features of the real-time motion data are extracted; the real-time feature data is compared with a standard motion model to identify real-time motion behavior and detect abnormal motion; in the case of detecting abnormal motion, control signals and alarm signals corresponding to the real-time motion behavior are generated; the alarm signal is sent to the alarm device to trigger an abnormal alarm, and the control signal is sent to the drilling equipment through the field industrial bus to control the drilling equipment to perform emergency operations.

[0108] To ensure the safety and quality of operations, real-time motion data of the operators is acquired while they are working on the drilling equipment. Then, feature extraction is performed on this real-time motion data to obtain real-time motion characteristics. By comparing the real-time feature data with a standard motion model, the current real-time motion behavior can be identified and abnormal motions can be detected. Then, in the event of an anomaly detection, control and alarm signals corresponding to the real-time motion behavior are automatically generated. The alarm signal is sent to the alarm device to trigger an anomaly alarm, and the control signal is sent to the drilling equipment via the fieldbus to control the drilling equipment for emergency operation. In other words, this application monitors drilling operations not by focusing on the parameters and characteristics of the equipment itself, but by focusing on the actions and behaviors of the operators. By pre-setting a standard motion model, motion recognition and abnormal motion detection can be achieved. In the event of an anomaly, alarm and control signals can be generated. The alarm signal enables real-time alerting, and the control signal allows direct control of the drilling equipment for emergency operation, bypassing manual intervention, thereby preventing safety accidents and ensuring operational safety and quality.

[0109] In one specific embodiment of this application, the action comparison module is specifically used to compare the temporal features in the real-time feature data with the temporal features in the standard action model, so as to identify real-time action behavior and detect whether the action is in place in the temporal dimension. By comparing the frequency domain features in real-time feature data with the frequency domain features in the standard action model, real-time action behavior can be identified and the smoothness of the action can be detected in the frequency domain dimension. By comparing the time-frequency domain features in real-time feature data with the time-frequency domain features in the standard action model, real-time action behavior can be identified in the time-frequency domain dimension and the presence of unexpected mutations can be detected. By merging the real-time recognition results and anomaly detection results in the time domain, frequency domain, and time-frequency domain dimensions, we obtain the action behavior recognition results and the abnormal action detection results.

[0110] In one specific embodiment of this application, the model building module is used to build a standard motion model, including: acquiring standard motion data corresponding to the operator performing standard operations based on drilling equipment, and extracting standard motion features from the standard motion data; Establish a mathematical model of the human skeleton and muscles that matches the drilling operation scenario; By simulating human movement, standard motion characteristics are applied to a mathematical model to generate a standard motion model.

[0111] In one specific embodiment of this application, the model building module is specifically used to acquire standard motion data corresponding to the standardized operating actions performed by the operator based on the drilling equipment; wherein, the standardized operating actions are at least one of the following: pressing down the brake lever with both hands, rotating the regulating valve with one hand, bending over to lift and carry the drilling tool, raising both arms to connect the pipeline, and standing steadily on the monitoring device. Extract standard motion features corresponding to standardized operational movements from standard motion data; Accordingly, standard motion characteristics are applied to a mathematical model by simulating human movement to generate a standard motion model, including: By simulating human movement, standard motion characteristics are applied to mathematical models to generate standard motion models corresponding to at least one of the following standardized operating actions: pressing down the brake lever with both hands, rotating and adjusting the valve with one hand, bending over to lift and carry the drilling tool, raising both arms to connect the pipeline, and standing steadily to monitor the instrument.

[0112] In one specific embodiment of this application, the model building module is specifically used to determine the skeletal segments involved in the drilling operation scenario as rigid connecting rods, and the joints involved in the drilling operation scenario as constrained joints with specific degrees of freedom. Determine the mass and center of mass of the rigid link, and establish a local coordinate system for each rigid body. Use homogeneous transformation matrix or source number to determine the rotational and translational relationships between adjacent rigid bodies. Based on the rotation of joints in drilling operations, determine the rotational degree of freedom constraints of the joints. The passive cushioning characteristics of key human cartilage and ligaments are simulated using a spring damping system. By using a muscle exertion model, muscles are abstracted into active contraction elements, series elastic elements, and parallel elastic elements, and the virtual muscle origin is supported on a rigid skeleton model to obtain a digital model; the rigid skeleton model includes rigid links and constrained joints.

[0113] In one specific embodiment of this application, the model building module is specifically used to infer the internal torque and interaction force borne by the joint instantaneously based on the motion trajectory and acceleration in the standard motion characteristics and combined with the mathematical model; By statistically averaging and extracting the retention lines of multiple internal torques and multiple interaction forces, the range of joint torques corresponding to standardized operating actions that are reasonably subjected to force, conform to ergonomics, and can be safely completed is obtained; The standard motion model is determined by utilizing the joint torque range.

[0114] In one specific embodiment of this application, the exception handling module is specifically used to send control signals to the programmable logic controller or hydraulic actuator at the bottom of the drilling rig via a field industrial bus, so as to drive the relay to close or the valve to self-lock, and force the winch or large clamp equipment to perform deceleration, action locking or emergency stop, so as to completely prevent safety accidents at the physical level.

[0115] Corresponding to the above method embodiments, this application also provides an electronic device. The electronic device described below can be referred to in correspondence with the drilling operation monitoring method described above.

[0116] See Figure 3 As shown, the electronic device includes: Memory 332 is used to store computer programs; The processor 322 is used to execute a computer program to implement the steps of the drilling operation monitoring method described in the above method embodiment.

[0117] For details, please refer to Figure 4 , Figure 4 This is a schematic diagram of the specific structure of an electronic device provided in this embodiment. The electronic device can vary significantly due to differences in configuration or performance. It may include one or more central processing units (CPUs) (e.g., one or more processors) and a memory 332. The memory 332 stores one or more computer programs 342 or data 344. The memory 332 can be temporary or permanent storage. The program stored in the memory 332 may include one or more modules (not shown in the diagram), each module may include a series of instruction operations on the data processing device. Furthermore, the processor 322 may be configured to communicate with the memory 332 and execute the series of instruction operations stored in the memory 332 on the electronic device 301.

[0118] Electronic device 301 may also include one or more power supplies 326, one or more wired or wireless network interfaces 350, one or more input / output interfaces 358, and / or one or more operating systems 341.

[0119] The steps in the drilling operation monitoring method described above can be implemented by the structure of electronic equipment.

[0120] Corresponding to the above method embodiments, this application also provides a readable storage medium. The readable storage medium described below can be referred to in conjunction with the drilling operation monitoring method described above.

[0121] A readable storage medium storing a computer program, which, when executed by a processor, implements the steps of the drilling operation monitoring method described in the above method embodiments.

[0122] The readable storage medium can specifically be a USB flash drive, external hard drive, read-only memory (ROM), random access memory (RAM), magnetic disk, or optical disk, or any other readable storage medium capable of storing program code.

[0123] The various embodiments in this specification are described in a progressive manner, with each embodiment focusing on its differences from other embodiments. Similar or identical parts between embodiments can be referred to interchangeably. For the apparatus disclosed in the embodiments, since it corresponds to the method disclosed in the embodiments, the description is relatively simple; relevant parts can be referred to the method section.

[0124] Those skilled in the art will further recognize that the units and algorithm steps of the various examples described in conjunction with the embodiments disclosed herein can be implemented in electronic hardware, computer software, or a combination of both. To clearly illustrate the interchangeability of hardware and software, the components and steps of the various examples have been generally described in terms of functionality in the foregoing description. Whether these functions are implemented in hardware or software depends on the specific application and design constraints of the technical solution. Those skilled in the art can use different methods to implement the described functions for each specific application, but such implementation should not be considered beyond the scope of this application.

[0125] The steps of the methods or algorithms described in conjunction with the embodiments disclosed herein can be implemented directly by hardware, a software module executed by a processor, or a combination of both. The software module can be located in random access memory (RAM), main memory, read-only memory (ROM), electrically programmable ROM, electrically erasable programmable ROM, registers, hard disk, removable disk, CD-ROM, or any other form of storage medium known in the art.

[0126] Finally, it should be noted that in this document, relationships such as "first" and "second" are used merely to distinguish one entity or operation from another, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Furthermore, the terms "include," "contain," or any other variations are intended to cover non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such a process, method, article, or apparatus.

[0127] This document uses specific examples to illustrate the principles and implementation methods of this application. The descriptions of the above embodiments are only for the purpose of helping to understand the methods and core ideas of this application. At the same time, for those skilled in the art, there will be changes in the specific implementation methods and application scope based on the ideas of this application. Therefore, the content of this specification should not be construed as a limitation of this application.

Claims

1. A method for monitoring drilling operations, characterized in that, include: Acquire real-time motion data of the operators performing operations based on drilling equipment, and extract the real-time motion features of the real-time motion data; The real-time feature data is compared with the standard action model to identify real-time action behaviors and detect abnormal actions; If an abnormal action is detected, a control signal and an alarm signal corresponding to the real-time action behavior are generated; The alarm signal is sent to the alarm device to trigger an abnormal alarm, and the control signal is sent to the drilling equipment via the field industrial bus to control the drilling equipment to perform emergency operations.

2. The method according to claim 1, characterized in that, Comparing the real-time feature data with a standard action model to identify real-time action behaviors and perform abnormal action detection includes: The temporal features in the real-time feature data are compared with the temporal features in the standard action model to identify real-time action behavior and detect whether the action is in place in the temporal dimension. The frequency domain features in the real-time feature data are compared with the frequency domain features in the standard action model to identify real-time action behavior and detect whether the action is smooth in the frequency domain dimension. The time-frequency domain features in the real-time feature data are compared with the time-frequency domain features in the standard action model to identify real-time action behavior and detect any unexpected mutations in the time-frequency domain dimension. By merging the real-time recognition results and anomaly detection results in the time domain, frequency domain, and time-frequency domain dimensions, we obtain the action behavior recognition results and the abnormal action detection results.

3. The method according to claim 1, characterized in that, Establishing the standard action model includes: Obtain standard motion data corresponding to the standard operation performed by the operator based on the drilling equipment, and extract the standard motion features of the standard motion data; Establish a mathematical model of the human skeleton and muscles that matches the drilling operation scenario; The standard motion characteristics are applied to the mathematical model by simulating human movement to generate the standard motion model.

4. The method according to claim 3, characterized in that, Acquire standard motion data corresponding to the standard operation performed by the operator based on the drilling equipment, and extract standard motion features from the standard motion data, including: Acquire standard motion data corresponding to the standardized operating actions performed by the operator based on the drilling equipment; wherein the standardized operating actions are at least one of the following: pressing down the brake lever with both hands, rotating the regulating valve with one hand, bending over to lift and carry the drill string, raising both arms to connect the pipeline, and standing steadily on the monitoring device; Extract the standard motion features corresponding to the standardized operation actions from the standard motion data; Accordingly, the standard motion characteristics are applied to the mathematical model by simulating human motion to generate the standard motion model, including: By simulating human movement, the standard motion characteristics are applied to the mathematical model to generate a standard motion model corresponding to at least one of the following standardized operating actions: pressing down the brake lever with both hands, rotating and adjusting the valve with one hand, bending over to lift and carry the drilling tool, raising both arms to connect the pipeline, and standing steadily to monitor the device.

5. The method according to claim 3, characterized in that, Establish mathematical models of the human skeleton and muscles that match drilling operation scenarios, including: The skeletal segments involved in the drilling operation scenario are defined as rigid linkages, and the joints involved in the drilling operation scenario are defined as constrained joints with specific degrees of freedom. Determine the mass and centroid position of the rigid link, establish a local coordinate system for each rigid body, and use a homogeneous transformation matrix or source number to determine the rotational and translational relationships between adjacent rigid bodies. Based on the rotation of the joints in the drilling operation scenario, the rotational degree of freedom constraints of the constrained joints are determined. The passive cushioning characteristics of key human cartilage and ligaments are simulated using a spring damping system. The muscle is abstracted into active contraction elements, series elastic elements, and parallel elastic elements using a muscle force model, and the virtual muscle origin is supported on a rigid skeleton model to obtain the digital model; wherein, the rigid skeleton model includes the rigid link and the constraint joint.

6. The method according to claim 3, characterized in that, The standard motion model is generated by applying the standard motion features to the mathematical model through simulation of human motion, including: Based on the motion trajectory and acceleration in the standard motion characteristics, and combined with the mathematical model, the internal torque and interaction force borne by the joint at an instant are deduced. By performing statistical averaging and line extraction on multiple internal torques and multiple interaction forces, a range of joint torques that corresponds to standardized operating actions, is reasonably subjected to force, conforms to ergonomics, and can be safely completed is obtained. The standard motion model is determined using the joint torque range.

7. The method according to any one of claims 1 to 6, characterized in that, The control signal is sent to the drilling equipment via a fieldbus to control the drilling equipment to perform emergency operations, including: The control signals are sent to the programmable logic controller or hydraulic actuator at the bottom of the drilling rig via the field industrial bus to drive the relay to close or the valve to lock, and force the winch or large clamp equipment to decelerate, lock the action or stop in an emergency, so as to completely prevent safety accidents at the physical level.

8. A drilling operation monitoring device, characterized in that, include: The real-time motion acquisition module is used to acquire real-time motion data of the operator when performing operations based on drilling equipment, and to extract the real-time motion features of the real-time motion data. The action comparison module is used to compare the real-time feature data with the standard action model to identify real-time action behavior and detect abnormal actions; The control signal generation module is used to generate control signals and alarm signals corresponding to the real-time action behavior when an abnormal action is detected. The anomaly handling module is used to send the alarm signal to the alarm device for anomaly alarm, and to send the control signal to the drilling equipment through the field industrial bus to control the drilling equipment to perform emergency operations.

9. An electronic device, characterized in that, include: Memory, used to store computer programs; A processor, configured to implement the steps of the drilling operation monitoring method as described in any one of claims 1 to 7 when executing the computer program.

10. A readable storage medium, characterized in that, The readable storage medium stores a computer program that, when executed by a processor, implements the steps of the drilling operation monitoring method as described in any one of claims 1 to 7.

Images