Multi-source sensing and self-adaptive correction system of drilling machine equipment and self-adaptive correction method thereof

Abstract

The invention discloses a multi-source sensing and self-adaptive correction system for drilling machine equipment, which comprises a camera C1 mounted on tubular column automation equipment or a specified observation position and an ultrasonic sensor C2 mounted on the tubular column automation equipment, and is characterized in that the camera C1 is connected with a visual identification server P1; at least one executing mechanism sensor is installed on each executing mechanism of the pipe column automation equipment, the device further comprises an information collection module P2 and a self-adaptive correction module P4, the information collection module P2 is connected with each executing mechanism sensor, an ultrasonic sensor C2 and a control analysis module P3, and the control analysis module P3 is connected with an equipment motion control unit J1. According to the invention, the problem that accurate and reliable target identification is difficult to realize in the field of drilling machine equipment through single visual identification in the prior art is solved. The invention further discloses a drilling machine equipment multi-source sensing and self-adaptive correction method.

Description

Technical Field

[0001] This invention belongs to the technical field of oil drilling control systems, and relates to a multi-source sensing and adaptive correction system for drilling rig equipment. This invention also relates to a method for adaptive correction using a multi-source sensing and adaptive correction system for drilling rig equipment. Background Technology

[0002] The application of visual recognition technology in drilling equipment is constrained by numerous factors. Firstly, the significantly varying 24-hour lighting conditions at the work site create a complex and unpredictable lighting environment. Secondly, drilling operations frequently encounter adverse weather conditions such as rain, snow, and fog, significantly reducing the accuracy of visual recognition technology. Furthermore, the specifications of the drill pipes, the objects to be identified, are diverse, ranging from 27 / 28 inches to 11 inches in diameter, resulting in complex standards. Finally, the surface of the drill pipes is often covered with mud, further increasing the difficulty of identification. The combined effect of these factors makes it difficult for existing single-vision recognition technologies to achieve accurate and reliable target identification in the drilling equipment field. Therefore, traditional single-vision recognition technology clearly cannot meet the high standards and stringent requirements of target drill pipe perception in the oilfield equipment industry. There is an urgent need for a new multi-source sensing and adaptive correction system that integrates multiple sensing capabilities and possesses adaptive correction capabilities for sensing results, in order to effectively solve the problem of target string identification and position detection in drilling equipment. Summary of the Invention

[0003] The purpose of this invention is to provide a multi-source perception and adaptive correction system for drilling equipment, which solves the problem that single visual recognition in the prior art is difficult to achieve accurate and reliable target recognition in the field of drilling equipment.

[0004] Another objective of this invention is to provide a multi-source sensing and adaptive correction method for drilling rig equipment.

[0005] The technical solution adopted in this invention is a drilling rig equipment multi-source sensing and adaptive correction system, including a camera C1 installed on the automated tubing equipment or at a designated observation position, and an ultrasonic sensor C2 installed on the automated tubing equipment. The camera C1 is connected to a visual recognition server P1 via wired or wireless transmission. At least one actuator sensor is installed on each actuator of the automated tubing equipment. The system also includes an information acquisition module P2 and an adaptive correction module P4. The information acquisition module P2 is connected to each actuator sensor, the ultrasonic sensor C2, and the control analysis module P3 via wired or wireless transmission. The adaptive correction module P4 is connected to the control analysis module P3 and the drilling rig integrated control system J2 via wired or wireless transmission. The control analysis module P3 is also connected to the equipment motion control unit J1 via wired or wireless transmission.

[0006] Preferably, the camera C1 is connected to the visual recognition server P1 via a data cable or wireless network to transmit real-time image data.

[0007] Preferably, the ultrasonic sensor C2 is connected to the information acquisition module P2 via a data cable or wireless network. The ultrasonic sensor C2 is used to measure the distance information of the target in the working environment and transmit the distance information to the information acquisition module P2.

[0008] Each actuator sensor is connected to the information acquisition module P2 via a data cable or wireless network. Each actuator sensor is used to monitor the position information of the actuator and transmit it to the information acquisition module P2.

[0009] The information acquisition module P2 is connected to the control and analysis module P3 via a data cable or wireless network. The control and analysis module P3 is used to perform multi-dimensional data fusion and target recognition on the data from the ultrasonic sensor C2 and the actuator sensor, and sends the recognition results to the adaptive correction module P4 via a data cable or wireless network.

[0010] The visual recognition server P1 is also connected to the control analysis module P3 and the adaptive correction module P4 via a data cable or wireless network.

[0011] The control analysis module P3 is connected to the equipment motion control unit J1 via a data cable or wireless network, and sends the final sensing results to the equipment motion control unit J1.

[0012] The drilling rig integrated control system J2 is connected to the equipment motion control unit J1 via a data cable. The equipment motion control unit J1 is located on the automated tubing equipment and is used to drive the actions of each actuator of the automated tubing equipment to realize the motion control of the automated tubing equipment.

[0013] The second technical solution adopted in this invention is: a multi-source sensing and adaptive correction method for drilling rig equipment, which employs the aforementioned multi-source sensing and adaptive correction system for drilling rig equipment, specifically as follows:

[0014] Step 1: Based on the image information acquired by camera C1 and the position information acquired by the actuator sensor, determine whether the automated tubing equipment has entered the confidence interval Z1 of the perceived target position;

[0015] Step 2: When the automated tubing equipment enters the target position confidence interval Z1, the information acquisition module P2 acquires the distance information detected by the ultrasonic sensor C2. The control analysis module P3 performs weighted analysis based on the image information, position information and distance information to complete the calculation of the target perception position and sends the target perception position to the adaptive correction module P4.

[0016] Step 3: The adaptive correction module P4 performs adaptive correction on the vision recognition server P1 and the control analysis module P3 based on the target perception position and the actual position of the pipe column. After correction, steps 1-3 are repeated to ensure that the deviation between the corrected target perception position and the actual intersection position of the pipe column is less than the allowable error. The final target perception position obtained after correction is sent to the equipment motion control unit J1, and the equipment motion control unit J1 guides the equipment to run towards the target perception position.

[0017] Preferably, step 1 specifically includes:

[0018] The automated tubing equipment moves towards the target position confidence interval Z1. During this process: the actuator sensor collects the position information of the corresponding actuator and transmits the collected position information to the information acquisition module P2 and the control analysis module P3 in sequence; the camera C1 collects image information and transmits it to the vision recognition server P1. The vision recognition server P1 performs perception analysis on the image information, obtains the image recognition result, completes the visual recognition and perception of the target, and sends the visual recognition and perception result of the target to the control analysis module P3. The control analysis module P3 determines whether the automated tubing equipment has entered the target position confidence interval Z1 based on the position information and image information. If it has not entered the target position confidence interval Z1, the automated tubing equipment continues to move towards the target position confidence interval Z1 until the automated tubing equipment enters the target position confidence interval Z1.

[0019] Preferably, in step 1, the visual recognition server P1 has a built-in target visual recognition model. The target visual recognition model performs perception analysis on the image information collected by the camera C1, obtains the image recognition result, and obtains the spatial coordinate information of the target. When the target pipe obtained by the visual recognition server P1 after perception analysis is located outside the confidence region Z1, the visual recognition server P1 will determine that there is no target pipe in the current handover area and send the recognition result to the control analysis module P3. When the target pipe obtained by the visual recognition server P1 after perception analysis is located inside the confidence region Z1, the visual recognition server P1 will determine that there is a target pipe in the current handover area and send the recognition result to the control analysis module P3.

[0020] The control analysis module P3 calculates the current position of the automated tubing equipment based on the position information of the corresponding actuator collected by the actuator sensor, and determines whether the automated tubing equipment has entered the confidence interval Z1 of the perceived target position.

[0021] If both the image information and the location information indicate that the target position confidence interval Z1 has been entered, then the automated pipe column equipment is considered to have entered the target position confidence interval Z1, meaning that the target pipe column is located within the confidence interval Z1. Otherwise, the automated pipe column equipment continues to move toward the target position confidence interval Z1 until it enters the target position confidence interval Z1.

[0022] Preferably, step 2 specifically includes:

[0023] Step 2.1: When the automated tubing equipment enters the target position confidence interval Z1, the ultrasonic sensor C2 obtains the distance information between the target tubing and the ultrasonic sensor C2, and sends it sequentially to the information acquisition module P2 and the control analysis module P3.

[0024] Step 2.2: The control analysis module P3 has a built-in multi-source sensor information weighted analysis model. The control analysis module P3 performs weighted calculations on the distance information, the spatial coordinate information after the automated pipe column equipment enters the target position confidence interval Z1, and the current position information to complete the calculation of the target perception position, which is the calculation of the target pipe column position. The final target perception position is then sent to the adaptive correction module P4.

[0025] The formula for the weighted analysis model of multi-source sensor information to calculate the weighted information of distance, spatial coordinates after the automated tubing equipment enters the target position confidence interval Z1, and current position is as follows:

[0026] T(x,y)=Q1(F(p),A,B)*P(X1,Y1)+Q2(F(p),A,B)

[0027] *P(X2, Y2) formula;

[0028] Among them, the weighting coefficient Where X0 is the X-coordinate of the actual target position, X1 is the X-coordinate of the target position resolved by the ultrasonic sensor, and X2 is the X-coordinate of the target obtained by the visual recognition server P1.

[0029] Weighting coefficients Where Y0 is the Y-coordinate of the actual target position, Y1 is the Y-coordinate of the target position resolved by the ultrasonic sensor, and Y2 is the Y-coordinate of the target obtained by the visual recognition server P1.

[0030] Q1(F(p), A, B) is a visual recognition result correction function with respect to the probability distribution function F(q), the X-axis weighting coefficient A, and the Y-axis weighting coefficient B;

[0031] Q2(F(p), A, B) is a correction function for the ultrasound recognition result with respect to the probability distribution function F(q), the X-axis weighting coefficient A, and the Y-axis weighting coefficient B;

[0032] P(X1, Y1) represents the visual recognition coordinates of the target pipe column in the horizontal plane obtained by the visual recognition server P1.

[0033] P(X2, Y2) represents the ultrasonic monitoring coordinates of the target pipe column in the horizontal plane obtained by the information acquisition module P2.

[0034] F(p) is the probability distribution function of the target pipe appearing under different (X0, Y0) coordinates within the confidence region Z1, and is dynamically updated based on the cumulative identification results.

[0035] Preferably, step 3 specifically includes:

[0036] The adaptive correction module P4 compares the perceived target position with the actual position of the tubing:

[0037] If the target sensing position is consistent with the actual position of the pipe column, the current multi-source target sensing and adaptive correction task ends, and the control analysis module P3 sends the target sensing position to the equipment motion control unit J1 to guide the equipment to move towards the target sensing position.

[0038] If the deviation between the target sensing position and the actual position of the tubing is less than or equal to the allowable error of equipment handover, the current multi-source target sensing and adaptive correction task ends. The control analysis module P3 sends the target sensing position to the equipment motion control unit J1, guides the equipment to move toward the target sensing position, and, through operator intervention, uses the adaptive correction module P4 to perform weighted correction on the visual recognition results and the ultrasonic sensor detection results.

[0039] If the deviation between the target sensing position and the actual position of the pipe column is greater than the allowable error of equipment handover, the operator intervenes and uses the adaptive correction module P4 to perform weighted correction on the visual recognition results and ultrasonic sensor detection results. After correction, steps 1-3 are repeated to make the deviation between the corrected target sensing position and the actual handover position of the pipe column less than the allowable error. The final target sensing position obtained after correction is sent to the equipment motion control unit J1, and the equipment motion control unit J1 guides the equipment to run towards the target sensing position.

[0040] Preferably, the process in step 3, where the adaptive correction module P4 performs weighted correction on the visual recognition results and the ultrasonic sensor detection results through operator intervention, specifically involves:

[0041] The operator sends a manual operation command to the equipment through the drilling rig integrated control system J2. The operation equipment motion control unit J1 runs the actuator to fine-tune the equipment so that the equipment moves to the final actual target position. The drilling rig integrated control system J2 converts the operator's operation behavior into operation commands and sends them to the adaptive correction module P4. The adaptive correction module P4 performs weighted correction on the visual recognition results and ultrasonic sensor detection results according to the final actual target position to obtain the final target perception position.

[0042] The correction process of the adaptive correction module P4 is as follows:

[0043] The adaptive correction module P4 compares the actual target position with the visually recognized target position and the target distance information detected by the ultrasonic sensor, updates the probability of the target column appearing at each point within the confidence region Z1, corrects the visual recognition results and ultrasonic sensor detection results, and obtains the final perceived position of the target column, specifically:

[0044] The adaptive correction module P4 compares the actual target position with the visually recognized target position and the target distance information detected by the ultrasonic sensor, and calculates weighting coefficients A and B, where:

[0045] Weighting coefficients Where X0 is the X-coordinate of the actual target position, X1 is the X-coordinate of the target position detected by the ultrasonic sensor, and X2 is the X-coordinate of the target position recognized by vision.

[0046] Weighting coefficients Where Y0 is the Y-coordinate of the actual target position, Y1 is the Y-coordinate of the target position detected by the ultrasonic sensor, and Y2 is the Y-coordinate of the target position recognized by vision.

[0047] Then, based on T(x, y) = Q1(F(p), A, B)*P(X1, Y1) + Q2(F(p), A, B)*P(X2, Y2), the visual recognition results and the ultrasonic sensor detection results are weighted and corrected to obtain the final target column sensing position.

[0048] Where Q1(F(p), A, B) is a visual recognition result correction function with respect to the probability distribution function F(q), the X-axis weighting coefficient A, and the Y-axis weighting coefficient B;

[0049] Q2(F(p), A, B) is a correction function for the ultrasound recognition result with respect to the probability distribution function F(q), the X-axis weighting coefficient A, and the Y-axis weighting coefficient B;

[0050] P(X1, Y1) represents the visual recognition coordinates of the target pipe column in the horizontal plane obtained by the visual recognition server P1.

[0051] P(X2, Y2) represents the ultrasonic monitoring coordinates of the target pipe column in the horizontal plane obtained by the information acquisition module P2.

[0052] F(p) is the probability distribution function of the target pipe appearing under different (X0, Y0) coordinates within the confidence region Z1, and is dynamically updated based on the cumulative identification results.

[0053] The beneficial effects of this invention are:

[0054] This invention utilizes cameras, ultrasonic sensors, and sensor information integrated into each actuator of the equipment to perform regional weighted analysis of the monitoring information, thereby achieving multi-source perception of the target. This effectively solves the problem that single visual recognition technology is insufficient to achieve accurate and reliable target recognition in the drilling equipment field.

[0055] This invention monitors operator manual adjustment commands to correct the control analysis module P3 and the visual recognition server P1, and continuously improves the reliability and accuracy of target perception through self-learning.

[0056] This invention, by comprehensively utilizing visual, auditory, and ultrasonic sensors, greatly enhances the adaptability to complex external environments and effectively overcomes the negative impact of changes in lighting and severe weather on recognition accuracy, making the identification and positioning of drill pipes of various specifications and conditions more accurate.

[0057] Based on the operator's manual feedback instructions, this invention continuously corrects and optimizes the perception model, thereby continuously improving the accuracy and reliability of target perception. The adaptive capability ensures that the system can maintain an efficient and stable working state in complex and ever-changing environments, providing solid perception information support for the automated operation of drilling rigs.

[0058] This invention enables precise positioning and guidance of drill rods by drilling rigs, significantly reducing manual intervention, promoting less-manned operation during drill rod handover, and fundamentally improving the level of automation.

[0059] This invention not only improves the working efficiency of drilling equipment, but also significantly enhances its operational safety and accuracy, playing a positive role in promoting the development of the petroleum equipment field. Attached Figure Description

[0060] Figure 1 This is a schematic diagram of the multi-source sensing and adaptive correction system for drilling equipment of the present invention;

[0061] Figure 2 This is a flowchart of the multi-source sensing and adaptive correction method for drilling equipment of the present invention;

[0062] Figure 3This is the confidence partitioning map of the target area in the multi-source sensing and adaptive correction method for drilling equipment of the present invention. Detailed Implementation

[0063] The present invention will now be described in detail with reference to the accompanying drawings and specific embodiments.

[0064] Example 1

[0065] The drilling rig of this invention is equipped with a multi-source sensing and adaptive correction system, the structure of which is as follows: Figure 1 As shown, the system includes a camera C1 installed on the automated tubing string equipment or at a designated observation location, and an ultrasonic sensor C2 installed on the automated tubing string equipment. The camera C1 is connected to a vision recognition server P1 via wired or wireless transmission. Each actuator of the automated tubing string equipment is equipped with at least one actuator sensor. The system also includes an information acquisition module P2 and an adaptive correction module P4. The information acquisition module P2 is connected to each actuator sensor, the ultrasonic sensor C2, and the control analysis module P3 via wired or wireless transmission. The adaptive correction module P4 is connected to the control analysis module P3 and the drilling rig integrated control system J2 via wired or wireless transmission. The control analysis module P3 is also connected to the equipment motion control unit J1 via wired or wireless transmission.

[0066] The ultrasonic sensor C2 is connected to the information acquisition module P2 via a data cable or wireless network. The ultrasonic sensor C2 is used to measure the distance information of targets in the working environment and transmit the distance information to the information acquisition module P2.

[0067] Each actuator sensor is connected to the information acquisition module P2 via a data cable or wireless network. Each actuator sensor is used to monitor the position information of the actuator and transmit it to the information acquisition module P2.

[0068] The information acquisition module P2 is connected to the control and analysis module P3 via a data cable or wireless network. The control and analysis module P3 is used to perform multi-dimensional data fusion and target recognition on the data from the ultrasonic sensor C2 and the actuator sensor, and sends the recognition results to the adaptive correction module P4 via a data cable or wireless network.

[0069] The visual recognition server P1 is also connected to the control analysis module P3 and the adaptive correction module P4 via a data cable or wireless network.

[0070] The control analysis module P3 is connected to the equipment motion control unit J1 via a data cable or wireless network, and sends the final sensing results to the equipment motion control unit J1.

[0071] The drilling rig integrated control system J2 is connected to the equipment motion control unit J1 via a data cable. The equipment motion control unit J1 is located on the automated tubing equipment, such as the iron drill, drilling platform robot, second-level platform robot, third-level platform robot, etc. The equipment motion control unit J1 is used to drive the actions of each actuator of the automated tubing equipment to realize the motion control of the automated tubing equipment.

[0072] Example 2

[0073] Based on Example 1, camera C1 is connected to visual recognition server P1 via a data cable or wireless network to transmit real-time image data.

[0074] Example 3

[0075] Based on Example 2, assuming there are N actuators on the automated pipe column equipment, actuator sensors S1, S2, S3, ..., SN are respectively installed on each actuator of the equipment to monitor the real-time operating position and related position information of the actuators. These sensors are connected to the information acquisition module P2 via data cables or wireless networks.

[0076] Example 4

[0077] The drilling rig equipment multi-source sensing and adaptive correction method of the present invention adopts the drilling rig equipment multi-source sensing and adaptive correction system in Example 3, and its process is as follows: Figure 2 As shown, specifically:

[0078] Step 1: Based on the image information acquired by camera C1 and the position information acquired by the actuator sensor, determine whether the automated tubing equipment has entered the confidence interval Z1 of the perceived target position. The confidence interval of the perceived target area is divided as follows: Figure 3 As shown, specifically:

[0079] The automated tubing equipment moves toward the target position confidence interval Z1. During this process: the actuator sensor collects the position information of the corresponding actuator and transmits the collected position information to the information acquisition module P2 and the control analysis module P3 in sequence; the camera C1 collects image information and transmits it to the vision recognition server P1. The vision recognition server P1 performs perception analysis on the image information, obtains the image recognition result, completes the visual recognition and perception of the target, and sends the visual recognition and perception result of the target to the control analysis module P3. The control analysis module P3 determines whether the automated tubing equipment has entered the target position confidence interval Z1 based on the position information and image information. If it has not entered the target position confidence interval Z1, the automated tubing equipment continues to move toward the target position confidence interval Z1 until the automated tubing equipment enters the target position confidence interval Z1.

[0080] The visual recognition server P1 has a built-in target visual recognition model. The target visual recognition model performs perception and analysis on the image information collected by the camera C1, obtains the image recognition result, and obtains the spatial coordinate information of the target. When the target pipe obtained by the visual recognition server P1 after perception and analysis is outside the confidence region Z1, the visual recognition server P1 will determine that there is no target pipe in the handover area and send the recognition result to the control and analysis module P3. When the target pipe obtained by the visual recognition server P1 after perception and analysis is inside the confidence region Z1, the visual recognition server P1 will determine that there is a target pipe in the current handover area and send the recognition result to the control and analysis module P3.

[0081] The control analysis module P3 calculates the current position of the automated tubing equipment based on the position information of the corresponding actuator collected by the actuator sensor, and determines whether the automated tubing equipment has entered the confidence interval Z1 of the perceived target position.

[0082] When both image information and location information indicate that the target position confidence interval Z1 has been entered, the automated pipe column equipment is considered to have entered the target position confidence interval Z1, meaning that the target pipe column is located inside the confidence interval Z1. Otherwise, the automated pipe column equipment continues to move toward the target position confidence interval Z1 until it enters the target position confidence interval Z1.

[0083] Step 2: When the automated tubing equipment enters the target position confidence interval Z1, the information acquisition module P2 acquires the distance information detected by the ultrasonic sensor C2. The control and analysis module P3 performs weighted analysis based on the image information, position information, and distance information to calculate the target sensing position and sends the target sensing position to the adaptive correction module P4; specifically:

[0084] Step 2.1: When the automated tubing equipment enters the target position confidence interval Z1, the ultrasonic sensor C2 obtains the distance information between the target tubing and the ultrasonic sensor C2, and sends it sequentially to the information acquisition module P2 and the control analysis module P3.

[0085] Step 2.2: The control analysis module P3 has a built-in multi-source sensor information weighted analysis model. The control analysis module P3 performs weighted calculations on the distance information, the spatial coordinates of the automated pipe column equipment after entering the target position confidence interval Z1, and the current position information to complete the calculation of the target sensing position, which is the target pipe column position. The final target sensing position is then sent to the adaptive correction module P4. The formula for the multi-source sensor information weighted analysis model to perform weighted calculations on the distance information, the spatial coordinates of the automated pipe column equipment after entering the target position confidence interval Z1, and the current position information is as follows:

[0086] T(x,y)=Q1(F(p),A,B)*P(X1,Y1)+Q2(F(p),A,B)*P(X2,

[0087] Y2) formula;

[0088] Among them, the weighting coefficient Where X0 is the X-coordinate of the actual target position, X1 is the X-coordinate of the target position resolved by the ultrasonic sensor, and X2 is the X-coordinate of the target obtained by the visual recognition server P1.

[0089] Weighting coefficients Where Y0 is the Y-coordinate of the actual target position, Y1 is the Y-coordinate of the target position resolved by the ultrasonic sensor, and Y2 is the Y-coordinate of the target obtained by the visual recognition server P1.

[0090] Q1(F(p), A, B) is a visual recognition result correction function with respect to the probability distribution function F(q), the X-axis weighting coefficient A, and the Y-axis weighting coefficient B;

[0091] Q2(F(p), A, B) is a correction function for the ultrasound recognition result with respect to the probability distribution function F(q), the X-axis weighting coefficient A, and the Y-axis weighting coefficient B;

[0092] P(X1, Y1) represents the visual recognition coordinates of the target pipe column in the horizontal plane obtained by the visual recognition server P1.

[0093] P(X2, Y2) represents the ultrasonic monitoring coordinates of the target pipe column in the horizontal plane obtained by the information acquisition module P2.

[0094] F(p) is the probability distribution function of the target pipe appearing under different (X0, Y0) coordinates within the confidence region Z1, and is dynamically updated based on the cumulative identification results.

[0095] Step 3: The adaptive correction module P4 adaptively corrects the vision recognition server P1 and the control analysis module P3 based on the target perception position and the actual position of the pipe column. After correction, steps 1-3 are repeated to ensure that the deviation between the corrected target perception position and the actual pipe column junction position is less than the allowable error. The final target perception position obtained after correction is sent to the equipment motion control unit J1. The equipment motion control unit J1 guides the equipment to move towards the target perception position. Specifically:

[0096] The adaptive correction module P4 compares the perceived target position with the actual position of the tubing:

[0097] If the target sensing position is consistent with the actual position of the pipe column, the current multi-source target sensing and adaptive correction task ends, and the control analysis module P3 sends the target sensing position to the equipment motion control unit J1 to guide the equipment to move towards the target sensing position.

[0098] If the deviation between the target sensing position and the actual position of the tubing is less than or equal to the allowable error of equipment handover, the current multi-source target sensing and adaptive correction task ends. The control analysis module P3 sends the target sensing position to the equipment motion control unit J1, guides the equipment to move toward the target sensing position, and, through operator intervention, uses the adaptive correction module P4 to perform weighted correction on the visual recognition results and the ultrasonic sensor detection results.

[0099] If the deviation between the target sensing position and the actual position of the pipe column is greater than the allowable error of equipment handover, the operator intervenes and uses the adaptive correction module P4 to perform weighted correction on the visual recognition results and ultrasonic sensor detection results. After correction, steps 1-3 are repeated to make the deviation between the corrected target sensing position and the actual handover position of the pipe column less than the allowable error. The final target sensing position obtained after correction is sent to the equipment motion control unit J1, and the equipment motion control unit J1 guides the equipment to run towards the target sensing position.

[0100] Step 3, where the adaptive correction module P4 performs weighted correction on the visual recognition results and ultrasonic sensor detection results through operator intervention, is specifically as follows:

[0101] The operator sends a manual operation command to the equipment through the drilling rig integrated control system J2. The operation equipment motion control unit J1 runs the actuator to fine-tune the equipment so that the equipment moves to the final actual target position. The drilling rig integrated control system J2 converts the operator's operation behavior into operation commands and sends them to the adaptive correction module P4. The adaptive correction module P4 performs weighted correction on the visual recognition results and ultrasonic sensor detection results based on the final actual target position.

[0102] The correction process of the adaptive correction module P4 is as follows:

[0103] The adaptive correction module P4 compares the actual target position with the visually recognized target position and the target distance information detected by the ultrasonic sensor, updates the probability of the target column appearing at each point within the confidence region Z1, and performs weighted correction on the visual recognition results and ultrasonic sensor detection results to obtain the final perceived position of the target column, specifically:

[0104] The adaptive correction module P4 compares the actual target position with the visually recognized target position and the target distance information detected by the ultrasonic sensor, and calculates weighting coefficients A and B, where:

[0105] Weighting coefficients Where X0 is the X-coordinate of the actual target position, X1 is the X-coordinate of the target position detected by the ultrasonic sensor, and X2 is the X-coordinate of the target position recognized by vision.

[0106] Weighting coefficients Where Y0 is the Y-coordinate of the actual target position, Y1 is the Y-coordinate of the target position detected by the ultrasonic sensor, and Y2 is the Y-coordinate of the target position recognized by vision.

[0107] Then, based on T(x, y) = Q1(F(p), A, B)*P(X1, Y1) + Q2(F(p), A, B)*P(X2, Y2), the visual recognition results and the ultrasonic sensor detection results are weighted and corrected to obtain the final target column sensing position.

[0108] Where Q1(F(p), A, B) is a visual recognition result correction function with respect to the probability distribution function F(q), the X-axis weighting coefficient A, and the Y-axis weighting coefficient B;

[0109] Q2(F(p), A, B) is a correction function for the ultrasound recognition result with respect to the probability distribution function F(q), the X-axis weighting coefficient A, and the Y-axis weighting coefficient B;

[0110] P(X1, Y1) represents the visual recognition coordinates of the target pipe column in the horizontal plane obtained by the visual recognition server P1.

[0111] P(X2, Y2) represents the ultrasonic monitoring coordinates of the target pipe column in the horizontal plane obtained by the information acquisition module P2.

[0112] F(p) is the probability distribution function of the target pipe appearing under different (X0, Y0) coordinates within the confidence region Z1, and is dynamically updated based on the cumulative identification results.

Claims

1. A drilling rig equipment multi-source perception and adaptive correction system, characterized in that, The camera C1 installed on the pipe column automation equipment or the designated observation position is connected with the visual recognition server P1 through wired or wireless transmission, at least one actuator sensor is installed on each actuator of the pipe column automation equipment, and the information acquisition module P2, the adaptive correction module P4 are further included, the information acquisition module P2 is connected with each actuator sensor, the ultrasonic sensor C2 and the control analysis module P3 through wired or wireless transmission, the adaptive correction module P4 is connected with the control analysis module P3 and the drilling rig integrated control system J2 through wired or wireless transmission, and the control analysis module P3 is further connected with the equipment motion control unit J1 through wired or wireless transmission.

2. The drilling rig equipped multi-source perception and adaptive correction system of claim 1, wherein, The camera C1 is connected with the visual recognition server P1 through a data line or a wireless network, and real-time image data is transmitted.

3. The drilling rig equipped multi-source perception and adaptive correction system of claim 2, wherein, The ultrasonic sensor C2 is connected with the information acquisition module P2 through a data line or a wireless network, the ultrasonic sensor C2 is used for measuring distance information of a target in a working environment and transmitting the distance information to the information acquisition module P2. Each actuator sensor is connected with the information acquisition module P2 through a data line or a wireless network, and each actuator sensor is used for monitoring actuator position information and transmitting the actuator position information to the information acquisition module P2.

4. The drilling rig equipped multi-source perception and adaptive correction system of claim 3, wherein, The information acquisition module P2 is connected with the control analysis module P3 through a data line or a wireless network, the control analysis module P3 is used for multi-dimensional data fusion and target recognition on data of the ultrasonic sensor C2 and the actuator sensor, and the recognition result is sent to the adaptive correction module P4 through a data line or a wireless network; The visual recognition server P1 is further connected with the control analysis module P3 and the adaptive correction module P4 through a data line or a wireless network; The control analysis module P3 is connected with the equipment motion control unit J1 through a data line or a wireless network, and the final perception result is sent to the equipment motion control unit J1; The drilling rig integrated control system J2 is connected with the equipment motion control unit J1 through a data line, the equipment motion control unit J1 is located on the pipe column automation equipment, the equipment motion control unit J1 is used for driving each actuator of the pipe column automation equipment to act, and motion control of the pipe column automation equipment is realized.

5. A method for rig equipment multi-source perception and adaptive correction, characterized in that, The drilling rig equipment multi-source perception and adaptive correction system of claim 4 is specifically: Step 1, whether the pipe column automation equipment enters a perception target position confidence interval Z1 is judged according to image information acquired by the camera C1 and position information acquired by the actuator sensor; Step 2, when the pipe column automation equipment enters the perception target position confidence interval Z1, the information acquisition module P2 acquires distance information detected by the ultrasonic sensor C2, the control analysis module P3 performs weighted analysis according to the image information, the position information and the distance information, completes calculation of a target perception position, and sends the target perception position to the adaptive correction module P4; Step 3, the adaptive correction module P4 corrects the visual recognition server P1 and the control analysis module P3 according to the target perception position and the actual position of the pipe column, and repeats steps 1-3 after correction, so that the deviation between the target perception position after correction and the actual handover position of the pipe column is less than the allowable error, and the final target perception position obtained by correction is sent to the equipment motion control unit J1, and the equipment motion control unit J1 guides the equipment to run to the target perception position.

6. The method of claim 5, wherein, The step 1 is specifically: The pipe column automation equipment moves towards the perception target position confidence interval Z1, and in the process: the actuator sensor collects position information of the corresponding actuator, and the collected position information is sequentially transmitted to the information collection module P2 and the control analysis module P3; the camera C1 collects image information and transmits it to the visual recognition server P1, the visual recognition server P1 performs perception analysis on the image information, obtains the image recognition result, completes the visual recognition and perception of the target, and sends the visual recognition and perception result of the target to the control analysis module P3, the control analysis module P3 judges whether the pipe column automation equipment enters the perception target position confidence interval Z1 according to the position information and the image information, if not, the pipe column automation equipment continues to move towards the perception target position confidence interval Z1, until the pipe column automation equipment enters the perception target position confidence interval Z1.

7. The method of claim 6, wherein, The visual recognition server P1 in the step 1 is built-in with a target visual recognition model, which performs perception analysis on the image information collected by the camera C1, obtains the image recognition result, and obtains the spatial coordinate information of the target, when the target pipe column obtained by the visual recognition server P1 after perception analysis is located outside the confidence area Z1, the visual recognition server P1 will determine that there is no target pipe column in the handover area this time, and send the recognition result to the control analysis module P3; when the visual recognition server P1 performs perception analysis, the target pipe column obtained is located inside the confidence area Z1, the visual recognition server P1 will determine that there is a target pipe column in the current handover area, and send the recognition result to the control analysis module P3; The control analysis module P3 calculates the current position of the pipe column automation equipment according to the position information of the corresponding actuator collected by the actuator sensor, and judges whether the pipe column automation equipment enters the perception target position confidence interval Z1; When the image information and the position information are both judged to enter the perception target position confidence interval Z1, it is considered that the pipe column automation equipment enters the perception target position confidence interval Z1, that is, the target pipe column is located inside the confidence area Z1, otherwise the pipe column automation equipment continues to move towards the perception target position confidence interval Z1, until the pipe column automation equipment enters the perception target position confidence interval Z1.

8. The method of claim 7, wherein, The step 2 is specifically: Step 2.1, when the pipe column automation equipment enters the target position confidence interval Z1, the ultrasonic sensor C2 obtains the distance information between the target pipe column and the ultrasonic sensor C2, and sequentially sends it to the information collection module P2 and the control analysis module P3; Step 2.2: The control analysis module P3 has a built-in multi-source sensor information weighted analysis model. The control analysis module P3 performs weighted calculations on the distance information, the spatial coordinate information after the automated pipe column equipment enters the target position confidence interval Z1, and the current position information to complete the calculation of the target perception position, which is the calculation of the target pipe column position. The final target perception position is then sent to the adaptive correction module P4. The formula for the weighted analysis model of multi-source sensor information to calculate the weighted distance information, the spatial coordinate information of the automated pipeline equipment after entering the target position confidence interval Z1, and the current position information is as follows: T(x,y)=Q1(F(p),A,B)*P(X1,Y1)+Q2(F(p),A,B)*P(X2, Y2) formula; wherein the weighting coefficients wherein X0 is the X-coordinate of the actual target position, X1 is the X-coordinate of the target position resolved by the ultrasonic sensor, and X2 is the X-coordinate of the target obtained by the visual recognition server P1. Weighting coefficient Wherein, Y0 is the Y direction coordinate of the actual target position, Y1 is the target position Y direction coordinate analyzed by the ultrasonic sensor, and Y2 is the target Y direction coordinate obtained by the visual identification server P1. Q1(F(p), A, B) is a visual recognition result correction function with respect to the probability distribution function F(q), the X-axis weighting coefficient A, and the Y-axis weighting coefficient B; Q2(F(p), A, B) is a correction function for the ultrasound recognition result with respect to the probability distribution function F(q), the X-axis weighting coefficient A, and the Y-axis weighting coefficient B; P(X1, Y1) represents the visual recognition coordinates of the target pipe column in the horizontal plane obtained by the visual recognition server P1. P(X2, Y2) represents the ultrasonic monitoring coordinates of the target pipe column in the horizontal plane obtained by the information acquisition module P2. F(p) is the probability distribution function of the target pipe appearing under different (X0, Y0) coordinates within the confidence region Z1, and is dynamically updated based on the cumulative identification results.

9. The method of claim 8, wherein, Step 3 specifically involves: The adaptive correction module P4 compares the perceived target position with the actual position of the tubing: If the target sensing position is consistent with the actual position of the pipe column, the current multi-source target sensing and adaptive correction task ends, and the control analysis module P3 sends the target sensing position to the equipment motion control unit J1 to guide the equipment to move towards the target sensing position. If the deviation between the target sensing position and the actual position of the tubing is less than or equal to the allowable error of equipment handover, the current multi-source target sensing and adaptive correction task ends. The control analysis module P3 sends the target sensing position to the equipment motion control unit J1, guides the equipment to move toward the target sensing position, and, through operator intervention, uses the adaptive correction module P4 to perform weighted correction on the visual recognition results and the ultrasonic sensor detection results. If the deviation between the target sensing position and the actual position of the pipe column is greater than the allowable error of equipment handover, the operator intervenes and uses the adaptive correction module P4 to perform weighted correction on the visual recognition results and ultrasonic sensor detection results. After correction, steps 1-3 are repeated to make the deviation between the corrected target sensing position and the actual handover position of the pipe column less than the allowable error. The final target sensing position obtained after correction is sent to the equipment motion control unit J1, and the equipment motion control unit J1 guides the equipment to run towards the target sensing position.

10. The method of claim 9, wherein, In step 3, the process of using the adaptive correction module P4 to perform weighted correction on the visual recognition results and ultrasonic sensor detection results through operator intervention is as follows: The operator sends a manual operation command to the equipment through the drilling rig integrated control system J2. The operation equipment motion control unit J1 runs the actuator to fine-tune the equipment so that the equipment moves to the final actual target position. The drilling rig integrated control system J2 converts the operator's operation behavior into operation commands and sends them to the adaptive correction module P4. The adaptive correction module P4 performs weighted correction on the visual recognition results and ultrasonic sensor detection results according to the final actual target position to obtain the final target perception position. The correction process of the adaptive correction module P4 is as follows: The adaptive correction module P4 compares the actual target position with the visually recognized target position and the target distance information detected by the ultrasonic sensor, updates the probability of the target column appearing at each point within the confidence region Z1, corrects the visual recognition results and ultrasonic sensor detection results, and obtains the final perceived position of the target column, specifically: The adaptive correction module P4 compares the actual target position with the visually recognized target position and the target distance information detected by the ultrasonic sensor, and calculates weighting coefficients A and B, where: Weighting coefficient Wherein, X0 is the X direction coordinate of the actual target position, X1 is the X direction coordinate of the target position detected by the ultrasonic sensor, and X2 is the X direction coordinate of the target position identified by the vision. Weighting coefficient wherein Y0 is the Y-direction coordinate of the actual target position, Y1 is the Y-direction coordinate of the target position detected by the ultrasonic sensor, and Y2 is the Y-direction coordinate of the target position identified by the vision recognition. Then, based on T(x, y) = Q1(F(p), A, B)*P(X1, Y1) + Q2(F(p), A, B)*P(X2, Y2), the visual recognition results and the ultrasonic sensor detection results are weighted and corrected to obtain the final target column sensing position. Where Q1(F(p), A, B) is a visual recognition result correction function with respect to the probability distribution function F(q), the X-axis weighting coefficient A, and the Y-axis weighting coefficient B; Q2(F(p), A, B) is a correction function for the ultrasound recognition result with respect to the probability distribution function F(q), the X-axis weighting coefficient A, and the Y-axis weighting coefficient B; P(X1, Y1) represents the visual recognition coordinates of the target pipe column in the horizontal plane obtained by the visual recognition server P1. P(X2, Y2) represents the ultrasonic monitoring coordinates of the target pipe column in the horizontal plane obtained by the information acquisition module P2. F(p) is the probability distribution function of the target pipe appearing under different (X0, Y0) coordinates within the confidence region Z1, and is dynamically updated based on the cumulative identification results.

Images