Vision-based AGV trajectory tracking method

Abstract

The invention discloses an AGV trajectory tracking method based on vision, and relates to the technical field of heavy-load AGVs, comprising the following steps: S1, the AGV is equipped with a vision acquisition and image processing unit; s2, obtaining a tracking algorithm based on a geometric method; s3, adjusting the visual distance and the planned path to obtain the position coordinates of a front view point; s4, outputting an AGV steering instruction by using the method; s5, a set speed is added, so that the AGV vehicle can realize a trajectory tracking effect; and S6, displaying the effect picture through communication. According to the AGV trajectory tracking method based on vision, the image provided by vision is analyzed and processed to obtain the next motion point, so that the AGV trajectory tracking precision is improved, manual operation of male and female head rotation butt joint is not needed, time and labor are saved, and harm of chemical engineering places to human bodies is reduced; the method can achieve the effects that the path is planned by using environmental information, and the system is controlled by using PID control and an advanced algorithm.

Description

Technical Field

[0001] The present invention relates to the technical field of heavy-load AGVs, and in particular to a vision-based AGV trajectory tracking method. Background Art

[0002] AGV, the full name is Automated Guided Vehicle, which is an "automatic guided vehicle". It is a transport vehicle equipped with an automatic guidance device such as electromagnetic or optical, which can travel along a specified guide path and has safety protection and various transfer functions. AGV does not require a driver in industrial applications. It uses rechargeable batteries as a power source and is widely used in logistics, manufacturing and other fields to realize the automated handling and transportation of materials. AGV is divided into several categories according to its different functions. Among them, heavy-duty AGV refers to AGV used to transport heavy materials and large goods. It usually has a strong load-bearing capacity and high stability. It is suitable for application scenarios that need to transport a large amount of heavy materials within a relatively short distance, such as production lines, warehouses, etc.

[0003] In the prior art, heavy-duty AGVs generally use SLAM laser navigation, which uses the SLAM algorithm to sense the surrounding environment and obtain trajectory data to achieve AGV trajectory tracking. However, although SLAM laser navigation technology can provide more accurate positioning and map construction, its accuracy is slightly lower than that of some traditional navigation methods (such as magnetic stripe and magnetic nail navigation), and laser scanning data is easily affected by environmental factors (such as lighting, reflective surface characteristics, etc.). And because the SLAM algorithm requires real-time processing of a large amount of sensor data, as the complexity of the scene increases, the computational complexity of the algorithm will also rise sharply, resulting in system response delays or the inability to complete positioning and map construction tasks in real time.

[0004] In addition, the existing technology requires manual operation to rotate and connect the male and female heads, which is time-consuming and labor-intensive, and chemical sites have certain hazards to the human body.

[0005] Therefore, it is urgent to improve this shortcoming. The present invention studies and improves the existing technology and its shortcomings, and provides a vision-based AGV trajectory tracking method. Summary of the invention

[0006] The purpose of the present invention is to provide a vision-based AGV trajectory tracking method to solve the problems raised in the above background technology.

[0007] To achieve the above object, the present invention provides the following technical solution: a vision-based AGV trajectory tracking method, suitable for automatic docking and separation of a robotic arm to a rotary joint, specifically comprising the following steps:

[0008] S1,AGV is equipped with visual acquisition and image processing unit;

[0009] S2, based on the geometry method, obtain the tracking algorithm;

[0010] S3, obtaining the position coordinates of the foresight point by adjusting the visual distance and the planned path;

[0011] S4, using this method to output AGV steering instructions;

[0012] S5, plus the set speed, the AGV vehicle can achieve the effect of trajectory tracking;

[0013] S6. Display the effect diagram through communication.

[0014] Furthermore, in step S1, the visual acquisition unit (such as a camera or camera system) is responsible for capturing images of the surrounding environment and must be able to capture key information on the AGV's travel path, such as target positions, obstacles, traffic signs, etc. The image processing unit is responsible for processing the captured images and extracting useful information (such as target positions, obstacles, etc.). The processing process includes: image preprocessing (including image denoising, contrast enhancement, edge detection and other steps to improve image quality), feature extraction (extracting representative feature points or areas from the preprocessed image, such as corners, edges, textures, etc., these features will be used for subsequent target detection and recognition), target detection and recognition (locating and identifying target objects in the image based on the extracted features, such as swivel joints, cargo pallets, etc.), and obstacle detection (identifying and locating obstacles in the image, such as walls, other vehicles, pedestrians, etc., to avoid collisions with the AGV).

[0015] Furthermore, after processing the image, the image processing unit outputs the extracted useful information (such as target position, obstacle information, etc.) to the AGV control system, and the control system generates corresponding control instructions (such as turning, accelerating, decelerating, etc.) based on this information to realize autonomous navigation and trajectory tracking of the AGV.

[0016] Furthermore, in step S2, the object in the image is geometrized, and shape recognition, position measurement, etc. are performed to calculate the relative position and posture between the AGV and the target. The specific formula is as follows:

[0017]

[0018] Where L represents the distance between the front and rear axles, and δ represents the vehicle steering angle;

[0019] The tracking algorithm adjusts the motion trajectory of the AGV in real time based on the above geometric information to ensure that it can accurately track the target.

[0020] Furthermore, in step S3, after the tracking algorithm is determined, the AGV calculates the position coordinates of the next foresight point according to the current target position, its own position and the planned path.

[0021] Furthermore, the calculation process of the position coordinates of the next foresight point is as follows:

[0022] 1) Find a target point (gx,gy) so that the vehicle always moves towards the target point;

[0023] 2) Formula conclusion:

[0024]

[0025] Among them, l d Indicates the distance from the center of the vehicle's rear wheel to the tracking point (target point);

[0026] In the steering angle formula, α

[0027] The curvature of the arc, which is the reciprocal of R,

[0028] 3) Track the target point by means of an arc.

[0029] Furthermore, in step S4, after determining the position coordinates of the foresight point, a corresponding steering instruction is generated according to the position information to guide the AGV to adjust its driving direction so that it moves toward the target along a predetermined trajectory.

[0030] Furthermore, during the movement of the AGV, the path is smoothed by inserting additional points between path points or using a smoothing algorithm (such as a Bezier curve or a spline curve) to improve the stability of the AGV's travel;

[0031] Insert additional points: Insert additional points between the original path points to make the path denser and smoother; the steps are as follows:

[0032] 1) Determine the original path point set $P={p_1,p_2,...,p_n}$;

[0033] 2) Select an interpolation step size $\Delta d$ or a time interval $\Delta t$;

[0034] 3) For each pair of adjacent points $p_i$ and $p_{i+1}$, insert additional points between them at equal distances or equal time intervals;

[0035] 4) Calculate the coordinates of these insertion points, usually using the linear interpolation formula $p_{\text{new}}=p_i+\frac{\Delta d}{d_{i,i+1}}(p_{i+1}-p_i)$, where $d_{i,i+1}$ is the distance between $p_i$ and $p_{i+1}$;

[0036] Smoothing algorithm: For scenes that require higher smoothness, use Bezier curves or spline curves;

[0037] Among them, the Bezier curve is a mathematical curve defined by a starting point, an end point, and a control point. By adjusting the position of the control point, the shape and curvature of the curve can be flexibly controlled. The steps are as follows:

[0038] 1) Select the starting point, end point and several control points;

[0039] 2) Generate a smooth path using Bezier curve equations;

[0040] 3) If necessary, the position of the control points can be further optimized to achieve better smoothing effect;

[0041] Among them, the spline curve is a series of smoothly connected polynomial segments passing through given points. The most commonly used spline curve is the cubic spline curve, which is a cubic polynomial in each segment. The steps are as follows:

[0042] 1) Given a series of path points;

[0043] 2) Use numerical methods (such as interpolation or approximation) to calculate the polynomial coefficients in each segment to ensure that the curve is continuous and smooth at the path points;

[0044] 3) Connect these polynomial segments to form a smooth path.

[0045] Furthermore, in step S5, the AGV is given a set travel speed, and combined with factors such as the current state of the AGV (such as acceleration and deceleration limits), the curvature of the path, and the target position, the AGV is made to move smoothly along the predetermined trajectory to achieve accurate trajectory tracking.

[0046] Furthermore, in step S6, real-time images, trajectory information, AGV status and other data are sent to a monitoring terminal or display screen through a communication interface, so that the operator can understand the operating status of the system in real time and make necessary adjustments.

[0047] The present invention provides a vision-based AGV trajectory tracking method, which has the following beneficial effects:

[0048] The present invention analyzes and processes the image provided by vision to obtain the next movement point, thereby improving the AGV trajectory tracking accuracy, and does not require manual operation of the male and female heads for rotation and docking, saving time and effort, and reducing the harm to the human body in chemical sites; the present invention can achieve the following effects: using environmental information to plan the path, using PID control, and realizing the control system with advanced algorithms; classifying and identifying the image, using computer vision, obtaining the surrounding environment, and improving positioning accuracy and safety performance.

[0049] 1. Control the vehicle's steering to make it execute a planned global path;

[0050] 2. Minimize the distance and heading error between the vehicle and the path;

[0051] 3. Limit the absolute value and rate of change of the steering angle to ensure smooth driving of the vehicle. BRIEF DESCRIPTION OF THE DRAWINGS

[0052] Figure 1 A schematic diagram of a flow chart of a vision-based AGV trajectory tracking method of the present invention;

[0053] Figure 2 A schematic diagram of the geometric algorithm principle of a vision-based AGV trajectory tracking method of the present invention;

[0054] Figure 3 A geometric schematic diagram of a tracking controller of a vision-based AGV trajectory tracking method of the present invention. DETAILED DESCRIPTION

[0055] The following embodiments of the present invention are described in further detail in conjunction with the accompanying drawings and examples. The following examples are used to illustrate the present invention, but are not intended to limit the scope of the present invention.

[0056] like Figure 1-Figure 3 As shown, a vision-based AGV trajectory tracking method is suitable for automatic docking and separation of a robotic arm to a rotary joint, and specifically includes the following steps:

[0057] S1,AGV is equipped with visual acquisition and image processing unit;

[0058] Capture images of the surrounding environment through a visual acquisition unit (such as a camera or camera system) to capture key information on the AGV's driving path, such as target location, obstacles, traffic signs, etc.

[0059] The captured image is processed by the image processing unit to extract useful information (such as target location, obstacles, etc.). The processing process includes: image preprocessing (including image denoising, contrast enhancement, edge detection and other steps to improve image quality), feature extraction (extracting representative feature points or areas from the preprocessed image, such as corners, edges, textures, etc., these features will be used for subsequent target detection and recognition), target detection and recognition (locating and identifying target objects in the image based on the extracted features, such as rotary joints, cargo pallets, etc.), obstacle detection (identifying and locating obstacles in the image, such as walls, other vehicles, pedestrians, etc., to avoid collisions with AGVs);

[0060] After the image processing unit processes the image, it outputs the extracted useful information (such as target location, obstacle information, etc.) to the AGV control system. The control system generates corresponding control instructions (such as turning, acceleration, deceleration, etc.) based on this information to achieve autonomous navigation and trajectory tracking of the AGV.

[0061] S2, based on the geometry method, obtain the tracking algorithm;

[0062] The objects in the image are geometricized to perform shape recognition, position measurement, etc., so as to calculate the relative position and posture between the AGV and the target. Figure 2 , the specific formula is as follows:

[0063]

[0064] Where L represents the distance between the front and rear axles, and δ represents the vehicle steering angle;

[0065] The tracking algorithm adjusts the motion trajectory of the AGV in real time based on the above geometric information to ensure that it can accurately track the target;

[0066] S3, obtaining the position coordinates of the foresight point by adjusting the visual distance and the planned path;

[0067] After determining the tracking algorithm, the AGV calculates the position coordinates of the next foresight point based on the current target position, its own position, and the planned path. Figure 3 , the specific calculation process is as follows:

[0068] 1) Find a target point (gx,gy) so that the vehicle always moves towards the target point;

[0069] 2) Formula conclusion:

[0070]

[0071] Among them, l d Indicates the distance from the center of the vehicle's rear wheel to the tracking point (target point);

[0072] In the steering angle formula, α

[0073] The curvature of the arc, which is the reciprocal of R,

[0074] 3) Track the target point by means of an arc;

[0075] S4, using this method to output AGV steering instructions;

[0076] After determining the position coordinates of the foresight point, the corresponding steering instructions are generated according to this position information to guide the AGV to adjust its driving direction so that it moves toward the target along the predetermined trajectory; and during the movement of the AGV, the path is smoothed by inserting additional points between the path points or using smoothing algorithms (such as Bezier curves and spline curves) to improve the stability of the AGV's driving;

[0077] Insert additional points: Insert additional points between the original path points to make the path denser and smoother; the steps are as follows:

[0078] 1) Determine the original path point set $P={p_1,p_2,...,p_n}$;

[0079] 2) Select an interpolation step size $\Delta d$ or a time interval $\Delta t$;

[0080] 3) For each pair of adjacent points $p_i$ and $p_{i+1}$, insert additional points between them at equal distances or equal time intervals;

[0081] 4) Calculate the coordinates of these insertion points, usually using the linear interpolation formula $p_{\text{new}}=p_i+\frac{\Delta d}{d_{i,i+1}}(p_{i+1}-p_i)$, where $d_{i,i+1}$ is the distance between $p_i$ and $p_{i+1}$;

[0082] Smoothing algorithm: For scenes that require higher smoothness, use Bezier curves or spline curves;

[0083] Among them, the Bezier curve is a mathematical curve defined by a starting point, an end point, and a control point. By adjusting the position of the control point, the shape and curvature of the curve can be flexibly controlled. The steps are as follows:

[0084] 1) Select the starting point, end point and several control points;

[0085] 2) Generate a smooth path using Bezier curve equations;

[0086] 3) If necessary, the position of the control points can be further optimized to achieve better smoothing effect;

[0087] Among them, the spline curve is a series of smoothly connected polynomial segments passing through given points. The most commonly used spline curve is the cubic spline curve, which is a cubic polynomial in each segment. The steps are as follows:

[0088] 1) Given a series of path points;

[0089] 2) Use numerical methods (such as interpolation or approximation) to calculate the polynomial coefficients in each segment to ensure that the curve is continuous and smooth at the path points;

[0090] 3) Connect these polynomial segments to form a smooth path;

[0091] S5, plus the set speed, the AGV vehicle can achieve the effect of trajectory tracking;

[0092] Assign the AGV a set driving speed, and combine the AGV's current state (such as acceleration, deceleration limit), path curvature, and target position to make the AGV move smoothly along the predetermined trajectory and achieve accurate trajectory tracking;

[0093] S6. Display the effect diagram through communication to facilitate monitoring and debugging;

[0094] Through the communication interface, real-time images, trajectory information, AGV status and other data are sent to the monitoring terminal or display screen, so that the operator can understand the operating status of the system in real time and make necessary adjustments.

[0095] The embodiments of the present invention are given for the purpose of illustration and description, and are not intended to be exhaustive or to limit the invention to the disclosed forms. Many modifications and variations will be apparent to those of ordinary skill in the art. The embodiments are selected and described in order to better illustrate the principles and practical applications of the present invention and to enable those of ordinary skill in the art to understand the present invention and thereby design various embodiments with various modifications suitable for specific uses.

Claims

1. A vision-based AGV trajectory tracking method, characterized in that: It is suitable for the automatic docking and separation of the robot arm and the rotary joint, and specifically includes the following steps: S1,AGV is equipped with visual acquisition and image processing unit; S2, based on the geometry method, obtain the tracking algorithm; S3, obtaining the position coordinates of the foresight point by adjusting the visual distance and the planned path; S4, using this method to output AGV steering instructions; S5, plus the set speed, the AGV vehicle can achieve the effect of trajectory tracking; S6. Display the effect diagram through communication.

2. The vision-based AGV trajectory tracking method according to claim 1, characterized in that: In step S1, the visual acquisition unit is responsible for capturing images of the surrounding environment and must be able to capture key information on the AGV's driving path. The image processing unit is responsible for processing the captured images and extracting useful information. The processing process includes: image preprocessing, feature extraction, target detection and recognition, and obstacle detection.

3. The vision-based AGV trajectory tracking method according to claim 2, characterized in that: After processing the image, the image processing unit outputs the extracted useful information to the control system of the AGV, and the control system generates corresponding control instructions according to the information to realize autonomous navigation and trajectory tracking of the AGV.

4. The vision-based AGV trajectory tracking method according to claim 1, characterized in that: In step S2, the object in the image is geometrized, and shape recognition and position measurement are performed to calculate the relative position and posture between the AGV and the target. The specific formula is as follows: Where L represents the distance between the front and rear axles, and δ represents the vehicle steering angle; The tracking algorithm adjusts the motion trajectory of the AGV in real time based on the above geometric information to ensure that it can accurately track the target.

5. The vision-based AGV trajectory tracking method according to claim 1, characterized in that: In step S3, after the tracking algorithm is determined, the AGV calculates the position coordinates of the next foresight point according to the current target position, its own position and the planned path.

6. The vision-based AGV trajectory tracking method according to claim 5, characterized in that: The calculation process of the position coordinates of the next foresight point is as follows: 1) Find a target point (gx,gy) so that the vehicle always moves towards the target point; 2) Formula conclusion: Among them, l d Indicates the distance from the center of the vehicle's rear wheel to the tracking point; In the steering angle formula, α The curvature of the arc, which is the reciprocal of R, 3) Track the target point by means of an arc.

7. The vision-based AGV trajectory tracking method according to claim 1, characterized in that: In the step S4, after the position coordinates of the foresight point are determined, a corresponding steering instruction is generated based on this position information to guide the AGV to adjust its driving direction so that it moves toward the target along a predetermined trajectory. In the step S4, after the position coordinates of the foresight point are determined, a corresponding steering instruction is generated based on this position information to guide the AGV to adjust its driving direction so that it moves toward the target along a predetermined trajectory.

8. The vision-based AGV trajectory tracking method according to claim 7, characterized in that: During the movement of the AGV, the path is smoothed by inserting additional points between the path points or using a smoothing algorithm to improve the stability of the AGV's travel; Insert extra points: insert extra points between the original path points to make the path denser and smoother; Smoothing algorithm: For scenes that require higher smoothness, use Bezier curves or spline curves; a Bezier curve is a mathematical curve defined by a start point, an end point, and control points. By adjusting the position of the control points, the shape and curvature of the curve can be flexibly controlled; a spline curve is a smooth connection of a series of polynomial segments passing through given points.

9. The vision-based AGV trajectory tracking method according to claim 1, characterized in that: In step S5, a set travel speed is given to the AGV, and factors such as the current state of the AGV, the curvature of the path, and the target position are combined to enable the AGV to move smoothly along the predetermined trajectory to achieve accurate trajectory tracking.

10. The vision-based AGV trajectory tracking method according to claim 1, characterized in that: In step S6, data such as real-time images, trajectory information, and AGV status are sent to a monitoring terminal or display screen through a communication interface, so that operators can understand the operating status of the system in real time and make necessary adjustments.

Images