Method for improving double-vehicle soft linkage precision

Abstract

The invention relates to the technical field of double-vehicle soft linkage, in particular to a method for improving double-vehicle soft linkage precision. The method comprises the following steps that an industrial camera is installed in the middle of a master station AGV, a cross-shaped laser transmitter is arranged in the middle of a slave station AGV, meanwhile, a high-precision laser range finder is installed at one end of each vehicle, and a curtain with a fixed interval is arranged between the camera and the laser transmitter; the distance and angle deviation between front and rear vehicles are measured in real time through a laser range finder, the position of a cross laser line on a curtain is captured through an industrial camera, and the left-right deviation is calculated; and on the basis of a vehicle operational model, the speed needed by each Mecanum wheel is calculated, posture adjustment is carried out, and curve driving under the condition of double-vehicle linkage is completed. According to the design, the offset of the slave station AGV relative to the master station AGV is accurately controlled, it can be ensured that the optimal working distance is kept between the two vehicles, and therefore the smoother material carrying process is achieved.

Description

Technical Field

[0001] The invention relates to the technical field of double-vehicle soft linkage, and in particular to a method for improving the precision of double-vehicle soft linkage. Background Art

[0002] In actual operation, due to factors such as sensor error, uneven ground, wheel slippage, etc., the position of the slave AGV relative to the master AGV may deviate. If this deviation cannot be corrected in time, the coordination between the two vehicles will be reduced, affecting work efficiency and safety; the information exchange between AGVs depends on the wireless communication network. If there is a high delay or unstable signal, it will cause a lag in data transmission, thereby affecting the accuracy of the synchronous action between the two vehicles. For example, in a scenario that requires a quick response, communication delays may cause the AGV to be unable to adjust its position immediately, increasing the risk of a collision; there may be various dynamic obstacles or other mobile devices in the working environment, which will interfere with the normal operating path of the AGV; therefore, a method for improving the accuracy of dual-vehicle soft linkage is provided. Summary of the invention

[0003] The object of the present invention is to provide a method for improving the precision of dual-vehicle soft linkage, so as to solve the problems of actual operation position deviation, communication delay and instability and environmental interference proposed in the above-mentioned background technology.

[0004] To achieve the above object, the present invention provides a method for improving the precision of dual-vehicle soft linkage, comprising the following steps: S1. Install an industrial camera in the middle of the master AGV car, set a cross laser transmitter in the middle of the slave AGV car, and install a high-precision laser rangefinder at one end of each car, and set a curtain with a fixed distance between the camera and the laser transmitter; S2, using a laser rangefinder to measure the distance and angle deviation between the front and rear vehicles in real time, using an industrial camera to capture the position of the cross laser line on the screen, and calculating the left and right deviations; S3. Design two independent fuzzy self-tuning PID controllers to control the front and rear distance and angle deviation, and calculate the required speed of each Mecanum wheel and adjust the posture based on the vehicle operation model to complete the curve driving under the condition of dual-vehicle linkage.

[0005] As a further improvement of the technical solution, in S2, the distance and angle deviation between the front and rear vehicles are measured in real time by a laser rangefinder, including the following steps: S2.1, the laser rangefinder emits a laser pulse to the other vehicle; S2.2, the laser pulse is reflected back to the laser rangefinder by another vehicle, and the laser rangefinder receives the reflected signal; S2.3, calculate the distance between the two vehicles based on the emission time and reception time of the laser pulse; S2.4. Calculate the angle deviation using geometric methods.

[0006] As a further improvement of the technical solution, in S2, the position of the cross laser line on the screen is captured by an industrial camera to calculate the left and right deviations, including the following steps: S2.5, when two vehicles are linked, the laser line emitted by the cross laser transmitter on the slave AGV is projected onto the industrial camera screen on the master AGV; S2.6, capturing an image containing a cross laser line in real time by a camera; S2.7, preprocessing the captured image and smoothing the image using a Gaussian filter; S2.8, use the Canny edge detection algorithm to identify the laser line in the image; S2.9, find the circumscribed rectangle of the laser cross line; S2.10, calculating the Euclidean center coordinates of the circumscribed rectangle and determining the center coordinates of the image; S2.11. Calculate the left and right deviations based on the target center coordinates and the image center coordinates; S2.12. The actual offset distance between the two vehicles is calculated by using the conversion relationship between the actual distance obtained through calibration and the pixel distance.

[0007] As a further improvement of the technical solution, in S2.8, the Gaussian filter is: ; in, Indicates at location The value of the Gaussian filter at ; Represents the position coordinates relative to the center of the filter; represents the standard deviation of the Gaussian distribution; represents the base of natural logarithms; Represents pi.

[0008] As a further improvement of the technical solution, in S2.8, the Canny edge detection algorithm is used to identify the laser line in the image, including the following steps: S2.81, use the Sobel operator to calculate the gradient strength and direction of each pixel in the image; S2.82, retain the point with the largest local gradient intensity, suppress other points, and use two thresholds to determine the edge points; S2.83, use Hough transform to detect straight lines in an image; S2.84. Filter out laser lines based on their length and position.

[0009] As a further improvement of the technical solution, in S3, two independent fuzzy self-tuning PID controllers are designed to control the front-to-back distance and the angle deviation, including the following steps: S3.1. Determine the front-rear distance control and angle deviation control targets; S3.2, set the basic structure of PID controller; S3.3, define the membership functions of input variables and output variables; S3.4, formulate fuzzy rules to describe the relationship between input variables and output variables; S3.5, converting the input variables into fuzzy sets, applying fuzzy rules for reasoning, obtaining fuzzy outputs, and defuzzifying the fuzzy outputs by using a defuzzification algorithm; S3.6, converting the fuzzy output into PID parameter value; S3.7. According to the result of fuzzy reasoning, the parameters of the PID controller are updated in real time, and the parameters are introduced into the PID control algorithm of the AGV control system to make real-time corrections to the running trajectory of the vehicle.

[0010] As a further improvement of the technical solution, in S3.5, the defuzzification algorithm is: ; in, Represents the output value after defuzzification; Indicates the quantization factor corresponding to the fuzzy output; represents the membership function value; Represents an element in the fuzzy output set.

[0011] As a further improvement of the technical solution, in S3.7, the PID control algorithm is: ; in, Indicates Control increments of moments; represents proportional gain; represents the integral gain; represents the differential gain; Indicates The error change at each moment; Indicates the moment; Indicates The error change at time.

[0012] As a further improvement of the technical solution, in S3, based on the vehicle operation model, the required speed of each Mecanum wheel is calculated and the posture is adjusted, including the following steps: S3.8, determining the relative motion relationship between the two vehicles based on the relative motion transformation matrix; S3.9. Decompose the motion of the Mecanum wheel AGV into two parts: translation and rotation. S3.10, determining the required translation speed and rotation speed of the vehicle according to the output of the PID controller; S3.11. Define the position and direction of the wheels and calculate the speed of each wheel based on the vehicle dynamics model; S3.12. Send the calculated speed of each wheel as a control instruction to the corresponding motor driver to adjust the vehicle's posture.

[0013] As a further improvement of the technical solution, in S3.8, the relative motion transformation matrix is: ; in, Indicates the relative angle between the two vehicles; Indicates the forward and backward translation speed of the following vehicle in the global coordinate system; represents the lateral translation speed of the following vehicle in the global coordinate system; Indicates the rotation speed of the following vehicle; represents the front and rear translation speed of the front vehicle in the global coordinate system; represents the lateral translation speed of the front vehicle in the global coordinate system; Indicates the rotation speed of the preceding vehicle; represents the relative lateral translation speed between the two vehicles; represents the relative lateral translation speed between the two vehicles; Represents the relative rotation speed between the two vehicles.

[0014] Compared with the prior art, the present invention has the following beneficial effects: 1. In this method of improving the precision of dual-vehicle soft linkage, by accurately controlling the offset of the slave AGV relative to the master AGV, it is possible to ensure that the two vehicles maintain an optimal working distance, thereby achieving a smoother material handling process. This can not only speed up the completion of tasks, but also reduce the adjustment time and potential collision risks caused by position deviation, thereby significantly reducing system downtime and improving overall work efficiency.

[0015] 2. In this method of improving the precision of dual-vehicle soft linkage, accurately grasping and controlling the offset helps to avoid accidental contact or collision between AGVs, which is crucial to ensuring the safety of personnel and the integrity of equipment. In addition, by optimizing positioning technology and strengthening communication stability, various emergencies such as environmental changes and network interference can be effectively responded to, ensuring that a high level of safety standards and operational stability can be maintained even under complex and changeable working conditions. BRIEF DESCRIPTION OF THE DRAWINGS

[0016] Figure 1 The figure is a flow chart of the overall method of the present invention. DETAILED DESCRIPTION

[0017] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0018] Example: See Figure 1 As shown, this embodiment provides a method for improving the precision of dual-vehicle soft linkage, comprising the following steps: S1. Install an industrial camera in the middle of the master AGV, set a cross laser transmitter in the middle of the slave AGV, and install a high-precision laser rangefinder at one end of each vehicle. Set a curtain with a fixed distance between the camera and the laser transmitter to ensure that the camera can accurately detect the laser target. In this embodiment, the vehicle posture acquisition AGV adopts the soft linkage mode when two vehicles are linked. Two laser rangefinders and a visual camera are set between the vehicles, which can collect the relative posture relationship between the two vehicles in real time, and can send the position relationship of the vehicles to the control system to realize the correction of the vehicle walking path; This laser distance meter has an easy-to-use graphic display. Even in an extremely tilted position, WinTec can still achieve reliable recognition of black objects. The distance meter has two output ports: analog output (0...10V / 4...20mA), which can be switched. By matching the data acquisition conversion module, high-precision measurement of the distance between two vehicles and the relative angle can be achieved.

[0019] S2, using a laser rangefinder to measure the distance and angle deviation between the front and rear vehicles in real time, using an industrial camera to capture the position of the cross laser line on the screen, and calculating the left and right deviations; In this embodiment, the distance and angle deviation between the front and rear vehicles are measured in real time by a laser rangefinder, including the following steps: S2.1, the laser rangefinder emits a laser pulse to the other vehicle; S2.2, the laser pulse is reflected back to the laser rangefinder by another vehicle, and the laser rangefinder receives the reflected signal; S2.3. Calculate the distance between the two vehicles based on the emission time and reception time of the laser pulse ,in, It's distance. is the speed of light, is the round trip time of the laser pulse; S2.4, use geometric methods to calculate the angle deviation. If two laser rangefinders measure the distance between two points respectively , and the horizontal distance between these two points is known , then the angle deviation for: .

[0020] Considering that when two vehicles are linked, the distance and angle will change due to the accumulated walking errors, the field of view of the camera needs to be adapted within 0~2 meters. The camera with conventional automatic focal length adjustment function cannot meet the requirements. In order to solve this problem, a curtain is set between the camera and the laser transmitter. The distance between the curtain and the camera is fixed and can be set within the field of view of the camera during the debugging process. The camera only needs to check the laser target on the curtain to detect the position between the two vehicles. Furthermore, an industrial camera is used to capture the position of the cross laser line on the screen and calculate the left and right deviations, including the following steps: S2.5, when two vehicles are linked, the laser line emitted by the cross laser transmitter on the slave AGV is projected onto the industrial camera screen on the master AGV; S2.6, capturing an image containing a cross laser line in real time by a camera; S2.7, preprocessing the captured image, such as grayscale, binarization, etc., to enhance the contrast and clarity of the laser line, and smoothing the image using a Gaussian filter to reduce noise; Furthermore, the Gaussian filter smoothes the image by weighted averaging, where the weight is determined by a Gaussian distribution function. This weighted method can effectively remove or reduce random noise in the image, such as salt and pepper noise, Gaussian noise, etc., which is very important for improving the accuracy of subsequent processing steps (such as edge detection); the Gaussian filter is: ; in, Indicates at location The value of the Gaussian filter at , which is used to smooth the pixels in the image; Represents the position coordinates relative to the center of the filter; Represents the standard deviation of the Gaussian distribution. The standard deviation controls the width and smoothness of the Gaussian filter. Larger values ​​result in wider filters, which produce a stronger smoothing effect; smaller values ​​result in wider filters, which produce a stronger smoothing effect; A smaller value will produce a narrower filter with less smoothing effect; represents the base of natural logarithms; represents pi; S2.8, use the Canny edge detection algorithm to identify the laser line in the image; Furthermore, the Canny algorithm is a multi-stage edge detection method, which first smoothes the image through a Gaussian filter to reduce the influence of noise, then calculates the image gradient and applies non-maximum suppression to refine the edge, and finally, determines the final edge through double threshold processing and edge tracking. This multi-level approach enables the Canny algorithm to provide high-quality edge detection results, which is very effective for identifying slender and clear laser lines; since the Canny algorithm performs Gaussian blur processing before edge detection, it helps to remove random noise in the image, thereby improving the reliability of laser line detection, which is particularly important for applications such as laser lines that require a high signal-to-noise ratio; using the Canny edge detection algorithm to identify laser lines in an image includes the following steps: S2.81. Use the Sobel operator to calculate the gradient strength and direction of each pixel in the image (the Sobel operator is a commonly used edge detection operator that calculates the gradient of an image in the horizontal and vertical directions through convolution operations. Specifically, the Sobel operator consists of two 3x3 matrices, one for detecting horizontal changes (Gx) and the other for detecting vertical changes (Gy)); S2.82, retain the point with the largest local gradient intensity, suppress other points (check each pixel along the gradient direction, if the gradient intensity of the point is less than the gradient intensity of its neighboring points, suppress it (that is, set it to 0), so that a thinner and clearer edge can be obtained), and use two thresholds (high threshold and low threshold, the high threshold is used to ensure that only very strong edges are marked as edge points, and the low threshold is used to include some weaker but possible edge parts) to determine the edge points; S2.83. Use Hough transform to detect straight lines in an image (Hough transform is a feature extraction technique that is often used to detect simple shapes in an image, such as lines and circles. For line detection, Hough transform maps points in image space to lines in parameter space to find collinear points); S2.84, filter out laser lines according to the length and position of the straight line (from the straight lines obtained by Hough transform, filter out possible laser lines according to the length, position and other characteristics of the straight line, for example, set a minimum length threshold to exclude straight lines that are too short; or exclude straight lines that are not in the expected area according to the position of the straight line); S2.9. Find the circumscribed rectangle of the laser cross line, which contains all parts of the laser cross line; S2.10. Calculate the Euclidean center coordinates of the circumscribed rectangle , and determine the center coordinates of the image ; S2.11. Calculate the left and right deviations based on the target center coordinates and the image center coordinates. , Indicates the horizontal offset of the center of the cross laser line relative to the center of the image. , Indicates the vertical offset of the center of the cross laser line relative to the center of the image; S2.12. The actual offset distance between the two vehicles is calculated by the conversion relationship between the actual distance and the pixel distance obtained by calibration. If the conversion coefficient is , , the actual offset distance is: , , these converted deviation values and It represents the actual lateral and longitudinal offset of the slave AGV relative to the master AGV in the horizontal plane.

[0021] S3. Design two independent fuzzy self-tuning PID controllers to control the front-to-rear distance and angle deviation, and calculate the required speed of each Mecanum wheel and adjust the posture based on the vehicle operation model to complete the curve driving under the condition of dual-vehicle linkage; Due to the front and rear distance deviation during the AGV movement With angle deviation There are two deviations. Therefore, two different fuzzy self-tuning PID controllers are needed to control the two errors respectively to achieve the purpose of precise control. The whole algorithm combines the fuzzy algorithm with the PID algorithm and uses the fuzzy algorithm to calculate the PID algorithm in real time. , , parameter( , , The PID controller can adjust the PID parameters dynamically according to the real-time response of the system, so as to more accurately control the position and posture of the AGV, which helps to reduce the position error and angle error and improve the overall control accuracy. The fuzzy logic can handle the nonlinear and uncertain system characteristics, so that the controller can maintain good performance under different working conditions. By adjusting the PID parameters through fuzzy rules, it can better cope with external interferences such as environmental changes and load changes. The traditional PID controller needs to manually adjust the Kp (proportional gain), Ki (integral gain) and Kd (differential gain) parameters, which is usually a time-consuming and complicated process. The fuzzy self-tuning PID controller can automatically adjust these parameters through fuzzy rules, which greatly simplifies the parameter adjustment process and improves the parameter adjustment efficiency. In this embodiment, two independent fuzzy self-tuning PID controllers are designed to control the front-to-back distance and the angle deviation, including the following steps: S3.1. Determine the front-to-rear distance control (the control target is to keep the distance between the front and rear vehicles close to the set value) and the angle deviation control target (the control target is to keep the angle deviation between the front and rear vehicles close to the set value); S3.2, set up the basic structure of the PID controller (define the proportional, integral and derivative terms); S3.3. Define the membership functions of input and output variables (triangular function graph); S3.4, formulate fuzzy rules to describe the relationship between input variables and output variables; S3.5, convert the input variables into fuzzy sets, apply fuzzy rules for reasoning (fuzzy reasoning is to use fuzzy input to compare with fuzzy rule base to get the optimal output value), get fuzzy output, and defuzzify the fuzzy output through defuzzification algorithm. Fuzzy controller needs to input fuzzy signal, so first of all, two deviations , , Deviation Change Rate And the output , , Perform fuzzy processing; First, calculate the quantization factors of each quantity, where The basic domain is , The basic domain is ; The basic domain is , The basic domain is . Assume that the domain of all input quantities on the fuzzy set is , that is, the domain {NB, NM, NS, ZO, PS, PM, PB}, which are respectively expressed as negative large, negative medium, negative small, zero, positive small, positive medium, and positive large. Therefore, the quantization factors of each quantity are: Based on experience, we can know the domain of the output quantity, and similarly, we can get the quantization factor of the output quantity; Furthermore, the defuzzification algorithm is: ; in, Represents the output value after defuzzification, that is, the accurate PID parameter value; Indicates the quantization factor corresponding to the fuzzy output; Represents the membership function value, representing an output variable The degree of membership; represents an element in the fuzzy output set, indicating possible PID parameter values; S3.6, converting the fuzzy output into PID parameter value; S3.7, based on the results of fuzzy reasoning, update the parameters of the PID controller in real time ( , , ), and introduce the parameters into the PID control algorithm of the AGV control system to make real-time corrections to the vehicle's running trajectory; Furthermore, the PID control algorithm is: ; ; in, Indicates Control increments of moments; Represents the proportional gain, which controls the effect of the current error on the control output; Represents the integral gain, which controls the impact of the cumulative error on the control output; Represents the differential gain, the effect of the rate of change of the control error on the control output; Indicates The error change at each moment; Indicates the moment; Indicates The error change at each moment; Indicates The error at the moment, that is, the difference between the set value and the actual value; Indicates Error in time; in, , , and The value of determines the deviation correction effect of the AGV vehicle. , . They correspond to the real-time difference in position and distance between the two AGVs.

[0022] Furthermore, by accurately calculating the speed of each Mecanum wheel, high-precision control of the position and direction of the AGV can be achieved, which helps ensure that the AGV can accurately travel along the predetermined path and reduce position deviation; the Mecanum wheel allows the AGV to move in all directions on a plane, including forward, backward, sideways, and rotation. The use of the operation model can more effectively plan these complex movement patterns and improve the flexibility and maneuverability of the AGV; real-time calculation based on the operation model can quickly respond to changes in the external environment or changes in task requirements. For example, when encountering an obstacle, the direction or speed of the AGV can be quickly changed by adjusting the speed of each wheel to avoid a collision; Based on the vehicle dynamics model, the required speed of each Mecanum wheel is calculated and the posture is adjusted, including the following steps: S3.8, determining the relative motion relationship between the two vehicles based on the relative motion transformation matrix; Furthermore, the relative motion transformation matrix is: ; in, Indicates the relative angle between the two vehicles; Indicates the forward and backward translation speed of the following vehicle in the global coordinate system; represents the lateral translation speed of the following vehicle in the global coordinate system; Indicates the rotation speed (angular velocity) of the following vehicle; represents the front and rear translation speed of the front vehicle in the global coordinate system; represents the lateral translation speed of the front vehicle in the global coordinate system; Indicates the rotation speed (angular velocity) of the front vehicle; represents the relative lateral translation speed between the two vehicles, which is the lateral translation speed of the rear vehicle relative to the front vehicle; represents the relative lateral translation speed between the two vehicles, which is the lateral translation speed of the rear vehicle relative to the front vehicle; represents the relative rotation speed between the two vehicles, which is the rotation speed of the rear vehicle relative to the front vehicle; S3.9. Decompose the motion of the Mecanum wheel AGV into two parts: translation and rotation. S3.10, according to the output of the PID controller, determine the required translation speed and rotation speed of the vehicle, each Mecanum wheel has a tilt angle , the roller deviates from the horizontal plane at this angle, and the vehicle's motion includes three degrees of freedom: forward and backward translation speed , lateral translation speed , and the rotation speed , the center of the vehicle’s reference coordinate system is located at its center of gravity; Since the roller of the Mecanum wheel has a For an inclination angle of degrees, the speed of the roller can be expressed as: ; ; in, is the linear speed of the wheel; For the Mecanum wheels of the eight wheels of the front vehicle, the motion of the vehicle can be expressed in the following matrix form: ; For the Mecanum wheels of the eight wheels of the rear vehicle, the motion of the vehicle can be expressed in the following matrix form: ; S3.11. Define the position and direction of the wheels and calculate the speed of each wheel based on the vehicle dynamics model (a vehicle dynamics model refers to a mathematical model that describes the motion characteristics of an AGV. These models are used to analyze and predict the behavior of an AGV under given input conditions, such as speed, acceleration, and steering). S3.12. The calculated speed of each wheel is sent as a control command to the corresponding motor driver to adjust the vehicle's posture. During the dual-vehicle linkage operation, the master AGV vehicle is responsible for sending control commands and moving along the set speed and direction according to the operation commands. The slave AGV vehicle measures the posture of the dual vehicles through the visual system and laser rangefinder, and realizes real-time linkage and correction through the internal correction algorithm. When the distance exceeds the set value during the linkage movement of the vehicle body, it can stop automatically.

[0023] The above shows and describes the basic principles, main features and advantages of the present invention. Those skilled in the art should understand that the present invention is not limited by the above embodiments, and the above embodiments and descriptions are only preferred examples of the present invention, and are not intended to limit the present invention. Without departing from the spirit and scope of the present invention, the present invention may have various changes and improvements, and these changes and improvements all fall within the scope of the present invention to be protected.

Claims

1. A method for improving the precision of dual-vehicle soft linkage, characterized in that: The following steps are involved: S1. Install an industrial camera in the middle of the master AGV car, set a cross laser transmitter in the middle of the slave AGV car, and install a high-precision laser rangefinder at one end of each car, and set a curtain with a fixed distance between the camera and the laser transmitter; S2, using a laser rangefinder to measure the distance and angle deviation between the front and rear vehicles in real time, using an industrial camera to capture the position of the cross laser line on the screen, and calculating the left and right deviations; S3. Design two independent fuzzy self-tuning PID controllers to control the front and rear distance and angle deviation, and calculate the required speed of each Mecanum wheel and adjust the posture based on the vehicle operation model to complete the curve driving under the condition of dual-vehicle linkage.

2. The method for improving the soft linkage accuracy of two vehicles according to claim 1, characterized in that: In S2, the distance and angle deviation between the front and rear vehicles are measured in real time by a laser rangefinder, including the following steps: S2.1, the laser rangefinder emits a laser pulse to the other vehicle; S2.2, the laser pulse is reflected back to the laser rangefinder by another vehicle, and the laser rangefinder receives the reflected signal; S2.3, calculate the distance between the two vehicles based on the emission time and reception time of the laser pulse; S2.

4. Calculate the angle deviation using geometric methods.

3. The method for improving the soft linkage accuracy of two vehicles according to claim 2, characterized in that: In S2, the position of the cross laser line on the screen is captured by an industrial camera, and the left and right deviations are calculated, which includes the following steps: S2.5, when two vehicles are linked, the laser line emitted by the cross laser transmitter on the slave AGV is projected onto the industrial camera screen on the master AGV; S2.6, capturing an image containing a cross laser line in real time by a camera; S2.7, preprocessing the captured image and smoothing the image using a Gaussian filter; S2.8, use the Canny edge detection algorithm to identify the laser line in the image; S2.9, find the circumscribed rectangle of the laser cross line; S2.10, calculating the Euclidean center coordinates of the circumscribed rectangle and determining the center coordinates of the image; S2.

11. Calculate the left and right deviations based on the target center coordinates and the image center coordinates; S2.

12. The actual offset distance between the two vehicles is calculated by using the conversion relationship between the actual distance obtained through calibration and the pixel distance.

4. The method for improving the soft linkage accuracy of two vehicles according to claim 3 is characterized in that: In S2.8, the Gaussian filter is: ； in, Indicates at location The value of the Gaussian filter at ; Represents the position coordinates relative to the center of the filter; represents the standard deviation of the Gaussian distribution; represents the base of natural logarithms; Represents pi.

5. The method for improving the soft linkage accuracy of two vehicles according to claim 4, characterized in that: In S2.8, the Canny edge detection algorithm is used to identify the laser line in the image, including the following steps: S2.81, use the Sobel operator to calculate the gradient strength and direction of each pixel in the image; S2.82, retain the point with the largest local gradient intensity, suppress other points, and use two thresholds to determine the edge points; S2.83, use Hough transform to detect straight lines in an image; S2.

84. Filter out laser lines based on their length and position.

6. The method for improving the soft linkage accuracy of two vehicles according to claim 5, characterized in that: In S3, two independent fuzzy self-tuning PID controllers are designed to control the front-to-back distance and the angle deviation, including the following steps: S3.

1. Determine the front-rear distance control and angle deviation control targets; S3.2, set the basic structure of PID controller; S3.3, define the membership functions of input variables and output variables; S3.4, formulate fuzzy rules to describe the relationship between input variables and output variables; S3.5, converting the input variables into fuzzy sets, applying fuzzy rules for reasoning, obtaining fuzzy outputs, and defuzzifying the fuzzy outputs by using a defuzzification algorithm; S3.6, converting the fuzzy output into PID parameter value; S3.

7. According to the result of fuzzy reasoning, the parameters of the PID controller are updated in real time, and the parameters are introduced into the PID control algorithm of the AGV control system to make real-time corrections to the running trajectory of the vehicle.

7. The method for improving the soft linkage accuracy of two vehicles according to claim 6, characterized in that: In S3.5, the defuzzification algorithm is: ； in, represents the output value after defuzzification; Indicates the quantization factor corresponding to the fuzzy output; represents the membership function value; Represents an element in the fuzzy output set.

8. The method for improving the soft linkage accuracy of two vehicles according to claim 7, characterized in that: In S3.7, the PID control algorithm is: ； in, Indicates Control increments of moments; represents proportional gain; represents the integral gain; represents the differential gain; Indicates The error change at each moment; Indicates the moment; Indicates The error change at time.

9. The method for improving the precision of dual-vehicle soft linkage according to claim 8, characterized in that: In S3, based on the vehicle operation model, the required speed of each Mecanum wheel is calculated and the posture is adjusted, including the following steps: S3.8, determining the relative motion relationship between the two vehicles based on the relative motion transformation matrix; S3.

9. Decompose the motion of the Mecanum wheel AGV into two parts: translation and rotation. S3.10, determining the required translation speed and rotation speed of the vehicle according to the output of the PID controller; S3.

11. Define the position and direction of the wheels and calculate the speed of each wheel according to the vehicle dynamics model; S3.

12. Send the calculated speed of each wheel as a control instruction to the corresponding motor driver to adjust the vehicle's posture.

10. The method for improving the soft linkage accuracy of two vehicles according to claim 9, characterized in that: In S3.8, the relative motion transformation matrix is: ； in, Indicates the relative angle between the two vehicles; It represents the forward and backward translation speed of the following vehicle in the global coordinate system; represents the lateral translation speed of the following vehicle in the global coordinate system; Indicates the rotation speed of the following vehicle; represents the front and rear translation speed of the front vehicle in the global coordinate system; represents the lateral translation speed of the front vehicle in the global coordinate system; Indicates the rotation speed of the preceding vehicle; represents the relative lateral translation speed between the two vehicles; represents the relative lateral translation speed between the two vehicles; Represents the relative rotation speed between the two vehicles.

Images