An adaptive control method and system for an iron roughneck

Abstract

The application discloses a kind of adaptive control method and system of iron driller, it is related to petroleum drilling automation equipment control technical field, the method includes: obtaining the pose data, environmental data and self running state data that multi-sensor fusion perception module real-time collection;Pose and environmental data are fused by extended Kalman filter, obtain the high-precision six-degree-of-freedom pose of jaw and its speed and the joint distance identified by vision;Pose and its speed, joint distance and self running state data are input into working condition identification algorithm, to judge current operation mode;According to current operation mode, select multi-mode PID controller or model predictive controller to calculate control instruction;Control instruction is output to the drive solenoid valve of actuating mechanism to complete corresponding action.The application improves the positioning accuracy, operation efficiency and safety of iron driller in dynamic operating environment by multi-sensor fusion and adaptive control.

Description

Technical Field

[0001] This invention relates to an adaptive control method and system for iron drillers, belonging to the field of control technology for automated equipment in oil drilling. Background Technology

[0002] The drill stringer is a key piece of equipment on modern drilling platforms, used to replace manual labor in the setup and unstitching of the drill string, and is a core component in achieving automation of the drilling process. Traditional drill stringer control systems mainly rely on preset programs, remote control via control handles, or simple sensor feedback (such as limit switches, single encoders, etc.), which have many technical bottlenecks.

[0003] In recent years, multi-sensor fusion technology and adaptive control technology have been widely used in agricultural machinery, construction machinery, and robotics. However, a systematic fusion application solution has yet to be seen in the field of iron drills. Therefore, an iron drill control system that can fully utilize multi-source sensing information and adaptively adjust control strategies is needed to solve the problems existing in the current technology. Summary of the Invention

[0004] The purpose of this invention is to overcome the shortcomings of the prior art and provide an adaptive control method and system for iron drill operators. Through multi-sensor fusion and adaptive control, the positioning accuracy, work efficiency and safety of iron drill operators in dynamic working environments are improved.

[0005] To achieve the above objectives, the present invention is implemented using the following technical solution:

[0006] In a first aspect, the present invention provides an adaptive control method for a steel driller, comprising:

[0007] The system acquires real-time pose data, environmental data, and its own operating status data of the driller from the multi-sensor fusion sensing module.

[0008] The pose data and environmental data are time-synchronized, spatially aligned and deeply fused using an extended Kalman filter algorithm to obtain the high-precision six-DOF pose and velocity of the iron drill pliers jaws, as well as the visually recognized joint distance.

[0009] The high-precision six-degree-of-freedom pose and its velocity, the visually recognized joint distance, and the self-operating status data are input into the working condition recognition algorithm to determine the current working mode;

[0010] Based on the current operating mode, select a multimodal PID controller or a model predictive controller, and calculate the control command;

[0011] The control command is output to the corresponding drive solenoid valve of the actuator to complete the extension, lifting, rotation, clamping, screwing or punching action.

[0012] Furthermore, the acquisition of the pose data, environmental data, and self-operating status data of the driller collected in real time by the multi-sensor fusion sensing module includes:

[0013] To acquire pose data, specifically: acquire attitude and acceleration data through an inertial measurement unit, and acquire joint angle or displacement data through an encoder;

[0014] Acquiring environmental data specifically involves: acquiring image data of the drill pipe joint through a vision camera, and acquiring distance data and 3D point cloud data of the environment through a laser sensor;

[0015] The system acquires its own operating status data, specifically: real-time torque data of the output shaft of the punching cylinder and the rotary motor is obtained through a torque sensor; real-time pressure data of the oil circuit of the clamping cylinder and the punching cylinder is obtained through a pressure sensor; and hydraulic oil temperature data is obtained through a temperature sensor.

[0016] Furthermore, the pose data and environmental data are time-synchronized, spatially aligned, and depth-fused using an extended Kalman filter algorithm to obtain a high-precision six-DOF pose and velocity of the drill bit jaws, as well as the visually recognized joint distance, including:

[0017] The data collected by the inertial measurement unit, encoder, vision camera and laser sensor are time-stamped and transformed to obtain the aligned and transformed attitude of the inertial measurement unit, the angle of the encoder, the relative coordinates of the connector measured by the vision camera and the distance measurement of the laser sensor.

[0018] The attitude of the aligned and transformed inertial measurement unit, the angle of the encoder, the relative coordinates of the connector measured by the vision camera, and the distance measured by the laser sensor are used as observation vectors and input into the extended Kalman filter.

[0019] The extended Kalman filter is run at a set frequency to output the fused jaws' six-degree-of-freedom pose and velocity in the wellhead coordinate system, and to output the visually recognized joint distance.

[0020] Furthermore, the high-precision six-degree-of-freedom pose and its velocity, the visually recognized joint distance, and the user's own operating status data are input into the working condition recognition algorithm to determine the current working mode, including:

[0021] The high-precision six-degree-of-freedom pose is compared with the preset target pose to obtain the pose deviation; the torque change rate is calculated based on the data collected by the torque sensor; the pressure change rate is calculated based on the data collected by the pressure sensor.

[0022] The pose deviation, torque change rate, pressure change rate, and the visually recognized joint distance are used as inputs, and a fuzzy logic classifier is used to calculate the membership degree of each mode; the mode with the highest membership degree is taken as the current working mode.

[0023] The current operating modes include rapid approach mode, precise positioning mode, clamping mode, screw fastening mode, buckle fastening mode, and buckle reset mode.

[0024] Furthermore, based on the current operating mode, a multimodal PID controller or a model predictive controller is selected, and control commands are calculated, including:

[0025] When the current operation mode is the rapid approach mode or the precise positioning mode, the model predictive controller is activated. Based on the kinematic model of the iron drill, the control sequence of extension, lifting and rotation is optimized simultaneously in the prediction time domain. Considering the constraints of stroke, speed and acceleration, the sequential quadratic programming is used to solve the problem.

[0026] When the current working mode is clamping mode, screwing mode, fastening mode or unfastening reset mode, the multi-mode PID controller is activated. The corresponding preset PID parameter group is called according to the current working mode. When switching modes, the parameters are smoothly switched through the first-order inertial element.

[0027] Secondly, the present invention provides an adaptive control system for a steel driller, comprising:

[0028] The multi-sensor fusion sensing module is used to collect the position and posture data, environmental data, and operating status data of the iron driller in real time.

[0029] An edge computing controller, connected to the multi-sensor fusion sensing module, is used to receive the pose data and environmental data. It performs time synchronization, spatial alignment, and deep fusion through an extended Kalman filter algorithm to obtain the high-precision six-degree-of-freedom pose and velocity of the drill bit jaws, as well as the visually recognized joint distance. It also uses a working condition recognition algorithm to determine the current working mode based on the high-precision six-degree-of-freedom pose and velocity, the visually recognized joint distance, and its own operating status data.

[0030] An adaptive control module, connected to the edge computing controller, is used to receive the current operating mode, select a multimodal PID controller or a model predictive controller according to the current operating mode, and calculate control commands.

[0031] An actuator, connected to the adaptive control module, is used to receive the control commands and respond to the control commands to complete extension, lifting, rotation, clamping, screwing or punching actions.

[0032] Furthermore, the multi-sensor fusion sensing module includes:

[0033] Inertial measurement units and encoders are used to acquire pose data; vision cameras and laser sensors are used to acquire environmental data.

[0034] Torque sensor, pressure sensor and temperature sensor are used to collect data on their own operating status;

[0035] The torque sensor is installed on the output shaft of the punching cylinder and the rotary motor, and the pressure sensor is installed on the oil circuit of the clamping cylinder and the punching cylinder.

[0036] Furthermore, the extended Kalman filter running on the edge computing controller uses the attitude of the inertial measurement unit, the angle of the encoder, the relative coordinates of the joint measured by the vision camera, and the distance measured by the laser sensor as observation vectors to output the fused jaws' six-degree-of-freedom pose and velocity in the wellhead coordinate system, and outputs the visually recognized joint distance.

[0037] The working condition recognition algorithm uses a fuzzy logic classifier, taking the pose deviation, torque change rate, pressure change rate and visually recognized joint distance as inputs, and outputs the membership degree of each mode. The mode with the highest membership degree is taken as the current working mode. The working modes include rapid approach mode, precise positioning mode, clamping mode, screw fastening mode, buckle fastening mode and unscrewing reset mode.

[0038] Furthermore, the adaptive control module includes a multimodal PID controller and a model predictive controller. The model predictive controller is activated when the current working mode is a rapid approach mode or a precise positioning mode. It is used to simultaneously optimize the control sequence of extension, lifting, and rotation in the prediction time domain based on the kinematic model of the iron drill, and solves the problem using sequential quadratic programming. The multimodal PID controller is activated when the current working mode is a clamping mode, a screw-on mode, a locking mode, or a release and reset mode. It is used to call the corresponding preset PID parameter group according to the current working mode and achieve smooth parameter switching through a first-order inertial element.

[0039] Furthermore, it also includes a display, which is connected to the edge computing controller via an Ethernet bus to receive real-time status and graphically display the current operating mode, the real-time three-dimensional pose of the jaws in the wellhead coordinate system, drill pipe count, torque-rotation curve, and system health status.

[0040] Compared with the prior art, the beneficial effects achieved by the present invention are as follows:

[0041] This invention provides an adaptive control method and system for a steel drill bit. It uses a multi-sensor fusion sensing module to collect pose data, environmental data, and its own operating status data in real time. The pose and environmental data are fused using an extended Kalman filter algorithm to obtain the high-precision six-degree-of-freedom pose and velocity of the drill bit's jaws, as well as the visually recognized joint distance. Combined with its own operating status data such as torque and pressure, the system inputs this data into a working condition recognition algorithm to determine the current operating mode. Based on the operating mode, it adaptively selects a multi-modal PID controller or a model predictive controller to output control commands. This achieves high-precision positioning, real-time operating mode recognition, and adaptive control of the steel drill bit in complex dynamic operating environments, significantly improving the efficiency, stability, and safety of automated hook-and-unhook operations. Attached Figure Description

[0042] Figure 1 This is a schematic diagram of the system architecture according to some embodiments of this disclosure;

[0043] Figure 2 These are schematic diagrams of electrical principles according to some embodiments of this disclosure;

[0044] Figure 3 This is a schematic diagram of an extended data fusion process according to some embodiments of this disclosure;

[0045] Figure 4 This is a schematic flowchart of an adaptive control method according to some embodiments of the present disclosure;

[0046] Figure 5 This is a schematic diagram of a working modal state machine model according to some embodiments of this disclosure;

[0047] In the diagram: 1, Multi-sensor fusion sensing module; 2, Edge computing controller; 3, Adaptive control module; 4, Actuator; 5, Display; 101, Inertial measurement unit; 102, Encoder; 103, Vision camera; 104, Laser sensor; 105, Torque sensor; 106, Pressure sensor; 107, Temperature sensor; 301, Multimodal PID controller; 302, Model predictive controller; 401, Telescopic drive solenoid valve; 402, Lifting drive solenoid valve; 403, Rotary drive solenoid valve; 404, Clamping drive solenoid valve; 405, Rotary drive solenoid valve; 406, Punching drive solenoid valve.

[0048] M0 is the rapid approach mode, M1 is the precise positioning mode, M2 is the clamping mode, M3 is the screw-on mode, M4 is the buckle-on mode, and M5 is the buckle-off and reset mode. Detailed Implementation

[0049] The present invention will be further described below with reference to the accompanying drawings. The following embodiments are only used to more clearly illustrate the technical solution of the present invention, and should not be used to limit the scope of protection of the present invention.

[0050] Example 1: An adaptive control method for a drill rig includes: acquiring the drill rig's pose data, environmental data, and its own operating status data in real time, collected by a multi-sensor fusion sensing module. For example... Figure 1 and Figure 2 As shown, the multi-sensor fusion sensing module 1 includes an inertial measurement unit 101, an encoder 102, a vision camera 103, a laser sensor 104, a torque sensor 105, a pressure sensor 106, and a temperature sensor 107.

[0051] In this embodiment, acquiring pose data specifically involves: acquiring attitude and acceleration data through an inertial measurement unit, and acquiring joint angle or displacement data through an encoder; acquiring environmental data specifically involves: acquiring image data of the drill pipe joint through a vision camera, and acquiring distance data and environmental 3D point cloud data through a laser sensor; acquiring its own operating status data specifically involves: acquiring real-time torque data of the output shaft of the punching cylinder and the rotary motor through a torque sensor, acquiring real-time pressure data of the clamping cylinder and the punching cylinder oil circuit through a pressure sensor, and acquiring hydraulic oil temperature data through a temperature sensor.

[0052] The pose data and environmental data are time-synchronized, spatially aligned, and deeply fused using an extended Kalman filter algorithm to obtain a high-precision six-DOF pose and velocity of the drill bit's jaws, as well as the visually recognized joint distance. For example... Figure 3 As shown, this embodiment specifically employs the following method: The attitude and acceleration data of the inertial measurement unit (IMU), the joint angle or displacement data of the encoder, the drill pipe joint image data from the vision camera, the distance data from the laser sensor, and the environmental 3D point cloud data are time-stamped and coordinate-transformed to obtain the aligned and transformed attitude of the IMU, the angle of the encoder, the relative coordinates of the joint measured by the vision camera, and the distance measured by the laser sensor. The aligned and transformed attitude of the IMU, the angle of the encoder, the relative coordinates of the joint measured by the vision camera, and the distance measured by the laser sensor are used as observation vectors and input into an extended Kalman filter. The extended Kalman filter is run at a frequency of 100Hz, outputting the fused six-degree-of-freedom pose and velocity of the jaws in the wellhead coordinate system, and outputting the visually recognized joint distance.

[0053] The high-precision six-degree-of-freedom pose and its velocity, the visually recognized joint distance, and the self-operating status data are input into the working condition recognition algorithm to determine the current working mode. In this embodiment, the specific steps are as follows: the high-precision six-degree-of-freedom pose is compared and calculated with a preset target pose to obtain the pose deviation; the torque change rate is calculated based on the torque sensor data in the self-operating status data; the pressure change rate is calculated based on the pressure sensor data in the self-operating status data; the pose deviation, torque change rate, pressure change rate, and the visually recognized joint distance are used as inputs, and a fuzzy logic classifier is used to calculate the membership degree of each mode; the mode with the highest membership degree is taken as the current working mode.

[0054] like Figure 5 As shown, the current operating modes include rapid approach mode M0, precise positioning mode M1, clamping mode M2, screw fastening mode M3, buckle fastening mode M4, and unscrew reset mode M5. The temperature sensor data in the self-operating status data is used for system health monitoring and does not participate in operating condition identification. In this embodiment, the state transition conditions are as follows: When the distance to the drill bit joint recognized by vision is greater than the set value and the jaw speed deviation is greater than the set value, the rapid approach mode is maintained; when the distance is less than 200mm and the speed deviation is less than the set value, the rapid approach mode is switched to the precise positioning mode; when the jaws are aligned with the center of the joint and the height difference is less than the set value, the precise positioning mode is switched to the clamping mode; at this time, the clamping drive solenoid valve 404 outputs, and the pressure sensor 107 reports that the pressure rises to the set value and then enters the swivel mode; in the swivel mode, the swivel drive solenoid valve 405 is activated, the rotation speed sensor reports the rotation speed, and when the number of swivel turns reaches the preset value and the torque sensor 106 detects a step increase in torque, the upper swivel mode is entered; in the upper swivel mode, the punch drive solenoid valve 406 is activated, and the upper swivel is completed to the set torque in a torque closed-loop manner, and then the unswivel reset mode is entered, the front vise is released, and the power head retracts.

[0055] Based on the current operating mode, select a multimodal PID controller or a model predictive controller, and calculate the control commands. For example... Figure 4 As shown, in this embodiment, specifically: when the current operating mode is a rapid approach mode or a precise positioning mode, a model predictive controller is activated. Based on the kinematic model of the iron drill, the control sequences for extension, lifting, and rotation are simultaneously optimized in the prediction time domain, considering travel, velocity, and acceleration constraints, and solved using sequential quadratic programming; when the current operating mode is a clamping mode, a screw-locking mode, an upper-locking mode, or a lower-locking reset mode, a multi-modal PID controller is activated. The corresponding preset PID parameter set is called according to the current operating mode, and smooth parameter switching is achieved through a first-order inertial element during mode switching. In this embodiment, the preset PID parameter set is obtained through offline genetic algorithm combined with hardware-in-the-loop simulation optimization.

[0056] The control commands are output to the corresponding drive solenoid valves of the actuator to complete extension, lifting, rotation, clamping, locking, or punching actions. For example... Figure 2 As shown, the actuator 4 includes a telescopic drive solenoid valve 401, a lifting drive solenoid valve 402, a rotation drive solenoid valve 403, a clamping drive solenoid valve 404, a rotary drive solenoid valve 405, and a snap-fit ​​drive solenoid valve 406. In this embodiment, the telescopic control command is output to the telescopic drive solenoid valve, the lifting control command is output to the lifting drive solenoid valve, the rotation control command is output to the rotation drive solenoid valve, the clamping control command is output to the clamping drive solenoid valve, and the actual clamping pressure fed back by the pressure sensor forms a closed-loop control; the rotary control command is output to the rotary drive solenoid valve, and the actual rotary torque fed back by the torque sensor forms a closed-loop control; the snap-fit ​​control command is output to the snap-fit ​​drive solenoid valve, and the actual snap-fit ​​torque fed back by the torque sensor forms a closed-loop control.

[0057] Example 2: This example provides an adaptive control system for a drill rig, used to implement the method described in Example 1, such as... Figure 1 As shown, the system includes a multi-sensor fusion sensing module 1, an edge computing controller 2, an adaptive control module 3, an actuator 4, and a display 5. The multi-sensor fusion sensing module 1 is used to collect the pose data, environmental data, and its own operating status data of the drill bit in real time. The edge computing controller 2 is connected to the multi-sensor fusion sensing module 1 and is used to receive the pose data and environmental data. It performs time synchronization, spatial alignment, and deep fusion using an extended Kalman filter algorithm to obtain the high-precision six-degree-of-freedom pose and velocity of the drill bit's jaws, as well as the visually recognized joint distance. Based on the high-precision six-degree-of-freedom pose and velocity, the visually recognized joint distance, and its own operating status data, it determines the current operating mode using a working condition recognition algorithm. The adaptive control module 3 is connected to the edge computing controller 2 and is used to receive the current operating mode. Based on the current operating mode, it selects a multi-modal PID controller 301 or a model prediction controller 302 and calculates control commands. The actuator 4 is connected to the adaptive control module 3 and is used to receive the control commands and respond to them to complete extension, lifting, rotation, clamping, screwing, or punching actions. The display 5 is connected to the edge computing controller 2 via an Ethernet bus to receive real-time status and display the current operating mode, the real-time three-dimensional pose of the jaws in the wellhead coordinate system, drill pipe count, torque-rotation curve and system health status in a graphical manner.

[0058] like Figure 1 and Figure 2As shown, the multi-sensor fusion sensing module 1 includes: an inertial measurement unit 101 and an encoder 102 for acquiring pose data; a vision camera 103 and a laser sensor 104 for acquiring environmental data; and a torque sensor 105, a pressure sensor 106, and a temperature sensor 107 for acquiring its own operating status data. The torque sensor 105 is installed on the output shaft of the punching cylinder and the rotary motor, and the pressure sensor 106 is installed on the clamping cylinder and the hydraulic circuit of the punching cylinder.

[0059] The extended Kalman filter running on the edge computing controller 2 uses the attitude of the inertial measurement unit, the angle of the encoder, the relative coordinates of the joint measured by the vision camera, and the distance measured by the laser sensor as observation vectors. It outputs the fused six-degree-of-freedom pose and velocity of the jaws in the wellhead coordinate system, and also outputs the visually recognized joint distance. The working condition recognition algorithm uses a fuzzy logic classifier, taking the pose deviation, torque change rate, pressure change rate, and visually recognized joint distance as inputs. It outputs the membership degree of each mode, and takes the mode with the highest membership degree as the current working mode. The working modes include rapid approach mode, precise positioning mode, clamping mode, screwing mode, snapping mode, and unscrewing reset mode.

[0060] The adaptive control module 3 includes a multimodal PID controller 301 and a model predictive controller 302. The model predictive controller 302 is activated when the current operating mode is a rapid approach mode or a precise positioning mode. It is used to simultaneously optimize the control sequence of extension, lifting, and rotation in the prediction time domain based on the kinematic model of a drill rig, using sequential quadratic programming for solution. The multimodal PID controller 301 is activated when the current operating mode is a clamping mode, a screw-on mode, a locking mode, or a release / reset mode. It is used to call the corresponding preset PID parameter group according to the current operating mode and achieve smooth parameter switching through a first-order inertial element.

[0061] Through the collaborative work of the above methods and systems, a complete intelligent closed loop of perception-decision-control-execution is formed, which can autonomously complete the entire process of automated operation from identifying the target drill rod, planning the motion path, precise alignment, reliable clamping to precise application of torque. This significantly improves the positioning accuracy, environmental adaptability, work efficiency and safety of drillers in complex and dynamic working environments.

[0062] The above description is only a preferred embodiment of the present invention. It should be noted that for those skilled in the art, several improvements and modifications can be made without departing from the technical principles of the present invention, and these improvements and modifications should also be considered within the scope of protection of the present invention.

Claims

1. An adaptive control method for a steel drill operator, characterized in that, include: The system acquires real-time pose data, environmental data, and its own operating status data of the driller from the multi-sensor fusion sensing module. The pose data and environmental data are time-synchronized, spatially aligned and deeply fused using an extended Kalman filter algorithm to obtain the high-precision six-DOF pose and velocity of the iron drill pliers jaws, as well as the visually recognized joint distance. The high-precision six-degree-of-freedom pose and its velocity, the visually recognized joint distance, and the self-operating status data are input into the working condition recognition algorithm to determine the current working mode; Based on the current operating mode, select a multimodal PID controller or a model predictive controller, and calculate the control command; The control command is output to the corresponding drive solenoid valve of the actuator to complete the extension, lifting, rotation, clamping, screwing or punching action.

2. The adaptive control method for the iron driller according to claim 1, characterized in that, The acquisition of the multi-sensor fusion sensing module, which collects the driller's pose data, environmental data, and its own operating status data in real time, includes: To acquire pose data, specifically: acquire attitude and acceleration data through an inertial measurement unit, and acquire joint angle or displacement data through an encoder; Acquiring environmental data specifically involves: acquiring image data of the drill pipe joint through a vision camera, and acquiring distance data and 3D point cloud data of the environment through a laser sensor; The system acquires its own operating status data, specifically: real-time torque data of the output shaft of the punching cylinder and the rotary motor is obtained through a torque sensor; real-time pressure data of the oil circuit of the clamping cylinder and the punching cylinder is obtained through a pressure sensor; and hydraulic oil temperature data is obtained through a temperature sensor.

3. The adaptive control method for the iron driller according to claim 2, characterized in that, The pose data and environmental data are time-synchronized, spatially aligned, and depth-fused using an extended Kalman filter algorithm to obtain a high-precision six-DOF pose and velocity of the drill bit jaws, as well as the visually recognized joint distance, including: The data collected by the inertial measurement unit, encoder, vision camera and laser sensor are time-stamped and transformed to obtain the aligned and transformed attitude of the inertial measurement unit, the angle of the encoder, the relative coordinates of the connector measured by the vision camera and the distance measurement of the laser sensor. The attitude of the aligned and transformed inertial measurement unit, the angle of the encoder, the relative coordinates of the connector measured by the vision camera, and the distance measured by the laser sensor are used as observation vectors and input into the extended Kalman filter. The extended Kalman filter is run at a set frequency to output the fused jaws' six-degree-of-freedom pose and velocity in the wellhead coordinate system, and to output the visually recognized joint distance.

4. The adaptive control method for the iron driller according to claim 2, characterized in that, The high-precision six-degree-of-freedom pose and its velocity, the visually recognized joint distance, and the user's own operating status data are input into the working condition recognition algorithm to determine the current working mode, including: The high-precision six-degree-of-freedom pose is compared with the preset target pose to obtain the pose deviation; the torque change rate is calculated based on the data collected by the torque sensor; the pressure change rate is calculated based on the data collected by the pressure sensor. The pose deviation, torque change rate, pressure change rate, and the visually recognized joint distance are used as inputs, and a fuzzy logic classifier is used to calculate the membership degree of each mode; the mode with the highest membership degree is taken as the current working mode. The current operating modes include rapid approach mode, precise positioning mode, clamping mode, screw fastening mode, buckle fastening mode, and buckle reset mode.

5. The adaptive control method for the iron driller according to claim 1, characterized in that, Based on the current operating mode, select a multimodal PID controller or a model predictive controller, and calculate control commands, including: When the current operation mode is the rapid approach mode or the precise positioning mode, the model predictive controller is activated. Based on the kinematic model of the iron drill, the control sequence of extension, lifting and rotation is optimized simultaneously in the prediction time domain. Considering the constraints of stroke, speed and acceleration, the sequential quadratic programming is used to solve the problem. When the current working mode is clamping mode, screwing mode, fastening mode or unfastening reset mode, the multi-mode PID controller is activated. The corresponding preset PID parameter group is called according to the current working mode. When switching modes, the parameters are smoothly switched through the first-order inertial element.

6. An adaptive control system for a drill rig, characterized in that, include: The multi-sensor fusion sensing module (1) is used to collect the pose data, environmental data and self-operation status data of the iron driller in real time; The edge computing controller (2) is connected to the multi-sensor fusion sensing module (1) and is used to receive the pose data and environmental data. It performs time synchronization, spatial alignment and deep fusion through the extended Kalman filter algorithm to obtain the high-precision six-degree-of-freedom pose and velocity of the iron drill pliers jaw, as well as the visually recognized joint distance. It also uses the working condition recognition algorithm to determine the current working mode based on the high-precision six-degree-of-freedom pose and velocity, the visually recognized joint distance and its own operating status data. An adaptive control module (3) is connected to the edge computing controller (2) and is used to receive the current working mode and select a multimodal PID controller (301) or a model prediction controller (302) according to the current working mode, and calculate control commands; The actuator (4) is connected to the adaptive control module (3) and is used to receive the control command and respond to the control command to complete the telescopic, lifting, rotating, clamping, screwing or punching action.

7. The adaptive control system for the iron driller according to claim 6, characterized in that, The multi-sensor fusion sensing module (1) includes: An inertial measurement unit (101) and an encoder (102) are used to acquire pose data; a vision camera (103) and a laser sensor (104) are used to acquire environmental data; A torque sensor (105), a pressure sensor (106), and a temperature sensor (107) are used to collect their own operating status data; The torque sensor (105) is installed on the output shaft of the punching cylinder and the rotary motor, and the pressure sensor (106) is installed in the oil circuit of the clamping cylinder and the punching cylinder.

8. The adaptive control system for the iron driller according to claim 6, characterized in that, The extended Kalman filter running by the edge computing controller (2) uses the attitude of the inertial measurement unit, the angle of the encoder, the relative coordinates of the joint measured by the vision camera and the distance measured by the laser sensor as the observation vector, outputs the fused jaws in the six-degree-of-freedom pose and velocity in the wellhead coordinate system, and outputs the visually recognized joint distance. The working condition recognition algorithm uses a fuzzy logic classifier, taking the pose deviation, torque change rate, pressure change rate and visually recognized joint distance as inputs, and outputs the membership degree of each mode. The mode with the highest membership degree is taken as the current working mode. The working modes include rapid approach mode, precise positioning mode, clamping mode, screw fastening mode, buckle fastening mode and unscrewing reset mode.

9. The adaptive control system for the iron driller according to claim 6, characterized in that, The adaptive control module (3) includes a multimodal PID controller (301) and a model predictive controller (302). The model predictive controller (302) is activated when the current working mode is a rapid approach mode or a precise positioning mode. It is used to simultaneously optimize the control sequence of extension, lifting, and rotation in the prediction time domain based on the kinematic model of the iron drill, and solves the problem by using a sequence quadratic programming. The multimodal PID controller (301) is activated when the current working mode is a clamping mode, a screw-on mode, an upper-locking mode, or a lower-locking reset mode. It is used to call the corresponding preset PID parameter group according to the current working mode and achieve smooth parameter switching through a first-order inertial link.

10. The adaptive control system for the iron driller according to claim 6, characterized in that, It also includes a display (5), which is connected to the edge computing controller (2) via an Ethernet bus to receive real-time status and display the current working mode, the real-time three-dimensional pose of the jaws in the wellhead coordinate system, drill pipe count, torque-rotation curve and system health status in a graphical manner.

Images