Pipe jacking construction automatic deviation rectifying system based on machine vision and PID (Proportion Integration Differentiation)

Abstract

The invention provides a pipe jacking construction automatic deviation rectification system based on machine vision and PID. The system comprises an image acquisition module, an image processing module, a posture calculation module, a PID control module and a deviation rectification execution module. According to the invention, the mark pattern image in the machine head is collected and processed in real time in a non-contact manner through the machine vision technology, the continuous, high-precision and automatic measurement of the three-dimensional attitude and position deviation of the pipe jacking machine head is realized, the traditional intermittent manual measurement mode is changed, and the real-time performance and accuracy of state sensing are obviously improved. A classical PID control theory is combined with modern machine vision detection, the system can automatically and rapidly calculate accurate control quantity through a PID algorithm according to the position deviation calculated in real time, a deviation correction mechanism is driven to act, the whole process from deviation detection to deviation correction execution is automatic, human intervention is greatly reduced, and the working efficiency is improved. And the intelligent level of construction is improved.

Description

Technical Field

[0001] This invention relates to the field of pipeline construction technology, and in particular to an automatic deviation correction system for pipe jacking construction based on machine vision and PID. Background Technology

[0002] In pipe jacking construction, the pipe jacking machine head is prone to deviating from the preset design axis when excavating in the soil due to uneven soil conditions, improper excavation parameters, or encountering obstacles. If this deviation is not corrected in a timely and effective manner, it will lead to errors in the pipeline laying position, and in severe cases, may cause pipeline damage, ground subsidence, or even engineering accidents. Currently, the common method of correction mainly relies on construction personnel to manually measure the machine head's posture and then manually operate the correction cylinder to adjust the machine head's direction based on experience. This method has significant shortcomings: manual measurement is lagging and cannot achieve real-time, continuous posture monitoring, making it difficult to detect minor deviations in a timely manner; manual operation relies on personal experience, resulting in poor control accuracy and stability, and is prone to over-correction or slow response, causing the machine head's trajectory to oscillate in an "S" shape, which not only affects the construction quality but also increases mechanical wear and construction risks. Summary of the Invention

[0003] In view of this, in order to solve the problems existing in the technical background, the present invention proposes an automatic deviation correction system for pipe jacking construction based on machine vision and PID. Specifically, it includes the following:

[0004] An automatic deviation correction system for pipe jacking construction based on machine vision and PID control includes an image acquisition module, an image processing module, an attitude calculation module, a PID control module, and a deviation correction execution module. The image acquisition module is fixedly installed inside the assembled pipe behind the pipe jacking machine head and is used to continuously acquire real-time images of a pre-set annular marker pattern on the inner wall of the pipe jacking machine head. The image processing module receives the real-time images and identifies the position coordinates of the annular marker pattern in the image using edge detection and feature extraction algorithms. The attitude calculation module, based on the position coordinates and the known actual size of the annular marker pattern, combined with camera calibration parameters, calculates the horizontal and vertical deviation angles α and β of the current axis of the machine head relative to the design axis, and further calculates the horizontal deviation ΔX and vertical deviation ΔZ of the current position of the machine head relative to the design axis. The PID control module uses the horizontal deviation ΔX and vertical deviation ΔZ as inputs and calculates the control quantities U_x(t) for horizontal deviation correction and U_z(t) for vertical deviation correction through independent PID controllers, respectively. The control law follows the formula:

[0005] U(t) = K_p * e(t) + K_i * ∫e(t)dt + K_d * de(t) / dt,

[0006] Where e(t) is the real-time deviation, K_p, K_i, and K_d are the pre-tuned proportional, integral, and derivative coefficients, respectively, and U(t) is the corresponding output control quantity; the correction execution module receives the control quantities U_x(t) and U_z(t) and converts them into extension and retraction commands for each set of correction cylinders in the jacking machine head, thereby changing the attitude of the machine head by adjusting the cylinder stroke to achieve automatic correction.

[0007] In one embodiment of the present invention, the specific calculation method of the attitude calculation module includes: obtaining the ellipse center coordinates (u, v) and major and minor axis parameters of the ellipse on the image plane through the image processing module; correcting the image coordinates using the intrinsic parameter matrix and distortion coefficients obtained from camera calibration; solving the rotation matrix R and translation vector T of the pipe jacking machine coordinate system relative to the camera coordinate system using the PnP algorithm based on the corrected ellipse parameters and the actual world coordinates of the ellipse pattern; calculating the horizontal deflection angle α and vertical deflection angle β from the rotation matrix R; and calculating the horizontal deviation ΔX and vertical deviation ΔZ by combining the components of the translation vector T in the corresponding directions with the design axis equation.

[0008] In one embodiment of the present invention, both the horizontal PID controller and the vertical PID controller in the PID control module adopt an incremental digital PID algorithm, and the formula for calculating the increment of the control quantity is as follows:

[0009] ΔU(k) ​​= K_p * [e(k) - e(k-1)] + K_i * e(k) + K_d * [e(k) - 2e(k-1) +e(k-2)];

[0010] Where k is the current sampling time, e(k), e(k-1), and e(k-2) are the deviation values ​​of the current time, the previous time, and the two previous times, respectively, and ΔU(k) ​​is the increment of the control quantity; the controller outputs U(k) = U(k-1) + ΔU(k), and the final control quantity is limited to the effective working range of the correction cylinder through the output limiting circuit.

[0011] In one embodiment of the present invention, the image acquisition module uses an industrial camera with a global shutter, in conjunction with a high-brightness LED ring light source. The illumination direction of the ring light source is coaxial with the optical axis of the camera to ensure clear capture of the image of the ring mark pattern in the low-light environment inside the jacking pipe. The ring mark pattern is composed of high-contrast black and white concentric rings and is sprayed on a specific section of the inner wall of the jacking pipe head.

[0012] The above technical solution has the following beneficial effects:

[0013] This invention utilizes machine vision technology to non-contactly acquire and process images of marking patterns inside the jacking machine head in real time. This enables continuous, high-precision, and automated measurement of the jacking machine head's three-dimensional posture and positional deviation, changing the traditional intermittent manual measurement mode and significantly improving the real-time performance and accuracy of state perception. By combining classical PID control theory with modern machine vision inspection, the system can automatically and quickly calculate precise control quantities based on the real-time calculated positional deviation using a PID algorithm. This drives the correction mechanism to operate, achieving full automation from deviation detection to correction execution. This significantly reduces human intervention and enhances the level of intelligent construction. Attached Figure Description

[0014] Figure 1 This is a flowchart illustrating the steps of an automatic deviation correction system for pipe jacking construction based on machine vision and PID according to the present invention. Detailed Implementation

[0015] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0016] Example 1: An automatic deviation correction system for pipe jacking construction based on machine vision and PID includes an image acquisition module, an image processing module, an attitude calculation module, a PID control module, and a deviation correction execution module.

[0017] The image acquisition module is responsible for acquiring raw visual data, using an industrial camera and its matching light source fixedly installed at a specific location inside the assembled pipe behind the pipe jacking machine head. To adapt to the harsh environment inside the pipe jacking machine—dark, humid, and vibrating—an industrial camera with a global shutter is selected to avoid motion blur, and a high-brightness LED ring light source, coaxially mounted with the camera's optical axis, provides uniform and sufficient illumination. A ring-shaped marking pattern composed of high-contrast black and white concentric rings is pre-sprayed onto a specific cross-section of the inner wall of the pipe jacking machine head. This pattern serves as a target for visual recognition and measurement.

[0018] During operation, the camera, aided by a light source, continuously and stably acquires real-time images of the circular marker pattern and sends the image data to the subsequent processing unit. Its working principle lies in ensuring clear, high-contrast images of the marker pattern are obtained at any stage of tunneling through stable coaxial illumination, laying the foundation for subsequent accurate image analysis.

[0019] Example 2: Based on the previous example, after receiving a real-time image from the camera, the image processing module in this example begins digital image processing to extract key information. The module's built-in image processing algorithm first preprocesses the image, such as filtering and denoising, and enhancing contrast, to optimize image quality. An edge detection algorithm is then used to identify the outline of the circular logo pattern in the image. Since the logo is a specific concentric ring, the algorithm can further use feature extraction methods such as ellipse fitting or Hough transform to accurately identify the position of the pattern in the image's pixel coordinate system, specifically including the pixel coordinates (u, v) of the ellipse's center and the major and minor axis parameters of the ellipse. Its working principle is to convert the visual image into quantifiable digital features, accurately locking the precise position and geometric shape of the logo pattern in each frame of the image. These data form the basis for subsequent three-dimensional spatial calculations.

[0020] The core task of the attitude calculation module is to convert the two-dimensional pixel coordinates in the image into the attitude and position deviation of the pipe jacking machine head in three-dimensional space. This module first calls the camera intrinsic parameter matrix and distortion coefficients obtained through prior calibration to correct the ellipse center coordinates and parameters obtained by the image processing module, eliminating errors caused by lens distortion. Based on the corrected ellipse parameters and the known size and position of the ring-shaped marker pattern in the real-world coordinate system, the perspective n-point (PnP) algorithm is used to solve for the rotation matrix R and translation vector T of the pipe jacking machine head coordinate system relative to the camera coordinate system. From the rotation matrix R, the horizontal and vertical deviation angles α and β of the current axis direction of the machine head relative to the camera coordinate system can be calculated. Simultaneously, the translation vector T reflects the position of the machine head's center point relative to the camera. Combined with the preset design axis spatial equation, the horizontal deviation ΔX and vertical deviation ΔZ of the current position of the machine head relative to the design axis can be further calculated. Its working principle utilizes the camera geometry model and spatial transformation principle in computer vision, and through monocular vision measurement technology, achieves non-contact, real-time calculation of attitude and position from two-dimensional images to three-dimensional space.

[0021] The PID control module is responsible for generating control commands based on the deviation, receiving real-time horizontal deviation ΔX and vertical deviation ΔZ from the attitude calculation module. Two independent but structurally identical PID controllers are set up for the two mutually perpendicular control directions, horizontal and vertical. Each controller takes the real-time deviation e(t) in the current direction as input and follows the classic PID control law:

[0022] U(t) = K_p * e(t) + K_i * ∫e(t)dt + K_d * de(t) / dt;

[0023] Where K_p, Ki, and K_d are proportional, integral, and derivative coefficients pre-tuned based on the dynamic characteristics of the pipe jacking machine and the formation response. The proportional term addresses the current error, the integral term eliminates accumulated steady-state error, and the derivative term predicts the error change trend to suppress oscillations. In this embodiment, to facilitate digital system implementation and improve reliability, the controller adopts an incremental digital PID algorithm. Its control increment calculation formula is as follows:

[0024] ΔU(k) ​​= K_p * [e(k) - e(k-1)] + K_i * e(k) + K_d * [e(k) - 2e(k-1) +e(k-2)];

[0025] Where k represents the current sampling time. The current control output U(k) equals the output U(k-1) of the previous time plus the increment ΔU(k) ​​calculated this time. In addition, the module includes an output limiting circuit to restrict the final calculated horizontal control quantity U_x(t) and vertical control quantity U_z(t) within the actual effective stroke range of the correction cylinder, preventing actuator saturation. Its working principle utilizes feedback control theory to weightedly synthesize the position deviation and its historical changes, generating a control signal that is both responsive and stable in real time, providing precise guidance for the correction action.

[0026] The correction execution module translates control commands into physical actions. This module receives control quantities U_x(t) and U_z(t) from the PID control module. These control quantities are digital signals and need to be converted into analog or switching commands that can drive the hydraulic system. The controller or actuator within the module generates corresponding electrical signals based on the magnitude and direction of the control quantities to control the opening of the electro-hydraulic proportional valve or servo valve. Changes in valve opening adjust the flow and direction of hydraulic oil to the correction cylinders arranged in groups within the pipe jacking head, thereby precisely controlling the extension and retraction stroke of the cylinders. For example, during horizontal correction, the horizontal turning of the jacking head is achieved by adjusting the difference in extension and retraction between the left and right groups of cylinders; during vertical correction, the pitching is achieved by adjusting the difference in extension and retraction between the upper and lower groups of cylinders. By changing the cylinder stroke, the final effect is on the jacking head, causing it to deflect slightly around the hinge point, thereby changing the tunneling direction and gradually eliminating the deviation from the design axis. Its working principle is to convert the digital commands of the control system into direct and reliable mechanical adjustments to the posture of the pipe jacking machine head through a hydraulic transmission mechanism, thereby completing the final execution link of closed-loop control and achieving the purpose of automatic deviation correction.

[0027] The basic principles and main features of the present invention have been described above. Those skilled in the art should understand that the present invention is not limited to the above embodiments. The embodiments and descriptions in the specification are only illustrative of the principles of the present invention. Various changes and modifications can be made to the present invention without departing from the spirit and scope of the present invention. All such changes and modifications fall within the scope of the present invention as claimed. The scope of protection of the invention is defined by the appended claims and their equivalents.

Claims

1. An automatic deviation correction system for pipe jacking construction based on machine vision and PID, characterized in that, The system includes an image acquisition module, an image processing module, an attitude calculation module, a PID control module, and a correction execution module. The image acquisition module is fixedly installed inside the assembled pipe behind the pipe jacking machine head and is used to continuously acquire real-time images of a pre-set annular marker pattern on the inner wall of the pipe jacking machine head. The image processing module receives the real-time images and identifies the position coordinates of the annular marker pattern in the image using edge detection and feature extraction algorithms. The attitude calculation module, based on the position coordinates and the known actual size of the annular marker pattern, combined with camera calibration parameters, calculates the horizontal and vertical deviation angles α and β of the current axis of the machine head relative to the design axis, and further calculates the horizontal deviation ΔX and vertical deviation ΔZ of the current position of the machine head relative to the design axis. The PID control module uses the horizontal deviation ΔX and vertical deviation ΔZ as inputs and calculates the control quantities U_x(t) for horizontal correction and U_z(t) for vertical correction through independent PID controllers, respectively. The control law follows the formula: U(t) = K_p * e(t) + K_i * ∫e(t)dt + K_d * de(t) / dt, Where e(t) is the real-time deviation, K_p, K_i, and K_d are the pre-tuned proportional, integral, and derivative coefficients, respectively, and U(t) is the corresponding output control quantity; the correction execution module receives the control quantities U_x(t) and U_z(t) and converts them into extension and retraction commands for each set of correction cylinders in the jacking machine head, thereby changing the attitude of the machine head by adjusting the cylinder stroke to achieve automatic correction.

2. The automatic deviation correction system for pipe jacking construction based on machine vision and PID as described in claim 1, characterized in that, The specific calculation method of the attitude calculation module includes: obtaining the ellipse center coordinates (u, v) and major and minor axis parameters of the ellipse on the image plane through the image processing module; correcting the image coordinates using the intrinsic parameter matrix and distortion coefficients obtained from camera calibration; solving the rotation matrix R and translation vector T of the pipe jacking machine head coordinate system relative to the camera coordinate system using the PnP algorithm based on the corrected ellipse parameters and the actual world coordinates of the ellipse pattern; calculating the horizontal deflection angle α and vertical deflection angle β from the rotation matrix R; and calculating the horizontal deviation ΔX and vertical deviation ΔZ from the components of the translation vector T in the corresponding directions combined with the design axis equation.

3. The automatic deviation correction system for pipe jacking construction based on machine vision and PID as described in claim 1, characterized in that, Both the horizontal and vertical PID controllers in the PID control module employ incremental digital PID algorithms, and the formula for calculating the control increment is as follows: ΔU(k) ​​= K_p * [e(k) - e(k-1)] + K_i * e(k) + K_d * [e(k) - 2e(k-1) + e(k-2)]; Where k is the current sampling time, e(k), e(k-1), and e(k-2) are the deviation values ​​of the current time, the previous time, and the two previous times, respectively, and ΔU(k) ​​is the increment of the control quantity; the controller outputs U(k) = U(k-1) + ΔU(k), and the final control quantity is limited to the effective working range of the correction cylinder through the output limiting circuit.

4. The automatic deviation correction system for pipe jacking construction based on machine vision and PID as described in claim 1, characterized in that, The image acquisition module uses an industrial camera with a global shutter, combined with a high-brightness LED ring light source. The illumination direction of the ring light source is coaxial with the camera's optical axis to ensure clear capture of the image of the ring mark pattern in the low-light environment inside the jacking pipe. The ring mark pattern is composed of high-contrast black and white concentric rings, sprayed on a specific section of the inner wall of the jacking pipe head.

Images