Multi-vehicle coordinated transportation method, device and system

Abstract

The present invention discloses a multi-vehicle cooperative transportation method, device and system, the method comprising: calculating the relative position of the vehicle and the center of mass of the structure and the desired state of the virtual structure according to a three-dimensional rectangular coordinate system; constructing a target kinematic model according to the relative position of the vehicle and the center of mass of the structure; calculating the vehicle position and vehicle posture using an extended Kalman filter algorithm according to vehicle sensor data; constructing map data according to radar data; performing path planning according to a preset grid map to obtain a global path and a local path; calculating driving parameters through a target selector according to the vehicle position, vehicle posture, map data, global path and local path; and performing cooperative transportation according to the desired state of the virtual structure, the target kinematic model and the driving parameters. The present invention realizes cooperative transportation and improves cooperative accuracy, driving stability and transportation efficiency. The present invention can be widely used in the field of unmanned transportation technology.

Description

Technical Field

[0001] The present invention relates to the field of unmanned transportation technology, and in particular to a multi-vehicle cooperative transportation method, device and system. Background Art

[0002] In the context of rapid development in various fields, the demand for large-scale cargo transportation is growing. Due to the large size of oversized cargo, the traditional manual transportation method has low transportation efficiency. At the same time, during the road transportation of large-scale cargo, due to the coupling of cargo-vehicle-road-environment, there is a lack of vehicle-road information coordination and interaction, which may cause collision with obstacles, resulting in cargo falling, low vehicle driving stability, and low coordination accuracy.

[0003] In summary, the technical problems existing in the relevant technologies need to be improved. Summary of the invention

[0004] The embodiments of the present invention provide a multi-vehicle cooperative transportation method, device and system, which effectively improve the cooperative accuracy, driving stability and transportation efficiency.

[0005] On the one hand, an embodiment of the present invention provides a multi-vehicle coordinated transportation method, comprising the following steps:

[0006] According to the three-dimensional rectangular coordinate system, the relative position of the vehicle and the structure's mass center and the desired state of the virtual structure are calculated;

[0007] According to the relative position of the vehicle and the structure mass center, a target kinematic model is constructed, wherein the target kinematic model includes a single-vehicle kinematic model, a single-vehicle oblique shift kinematic model and a multi-vehicle kinematic model;

[0008] According to the vehicle sensor data, the extended Kalman filter algorithm is used to calculate the vehicle position and vehicle posture;

[0009] Construct map data based on radar data;

[0010] According to the preset grid map, path planning is performed to obtain the global path and local path;

[0011] Calculating driving parameters through a target selector according to the vehicle position, the vehicle posture, the map data, the global path and the local path, the target selector includes a behavior selector, a speed selector, and a posture corrector, and the driving parameters include a distance between the cargo and the obstacle, a positional relationship between the cargo and the obstacle, a driving speed, a vehicle position coordinate, a rotation center point offset, a slant moving distance, a slant synthetic speed and a slant steering angle;

[0012] The collaborative transportation is performed according to the desired state of the virtual structure, the target kinematic model and the driving parameters.

[0013] In some embodiments, the calculating the relative position between the vehicle and the structure's mass center and the desired state of the virtual structure according to the three-dimensional rectangular coordinate system includes:

[0014] Constructing a group of centroid position equations according to the three-dimensional rectangular coordinate system;

[0015] Connect the center of each sub-support area according to the support mode and the center of mass position equation group to obtain a stable support area;

[0016] According to the stable support area, a set of equations of three side lengths of the supporting triangle is calculated;

[0017] Calculate the perimeter of the supporting triangle according to the three-side length equation group;

[0018] Calculate the area of ​​the supporting triangle according to the perimeter;

[0019] Calculating the support angle based on the area and the centroid height;

[0020] Calculating the relative position of the vehicle and the center of mass of the structure based on the plurality of support angles;

[0021] According to the three-dimensional rectangular coordinate system, a local virtual structure dynamic model is constructed by using a virtual structure method;

[0022] Constructing a local virtual structure state vector according to the local virtual structure dynamics model;

[0023] The virtual structure desired state is constructed according to the local virtual structure state vector.

[0024] In some embodiments, when the target kinematics model is a single-vehicle kinematics model, constructing the target kinematics model according to the relative position of the vehicle and the center of mass of the structure includes:

[0025] Constructing a velocity equation of a vehicle with a single driving wheel according to the relative position of the vehicle and the center of mass of the structure;

[0026] According to the single-drive-wheel bicycle motion speed equation, construct a left front wheel bicycle motion decomposition speed vector, a left rear wheel bicycle motion decomposition speed vector, a right front wheel bicycle motion decomposition speed vector, and a right rear wheel bicycle motion decomposition speed vector;

[0027] Performing a filling transformation on the left front wheel bicycle motion decomposition velocity vector to obtain a left front wheel bicycle motion equation group;

[0028] Performing a filling transformation on the left rear wheel bicycle motion decomposition velocity vector to obtain a left rear wheel bicycle motion equation group;

[0029] Performing a filling transformation on the right front wheel bicycle motion decomposition velocity vector to obtain a right front wheel bicycle motion equation group;

[0030] Performing a filling transformation on the right rear wheel bicycle motion decomposition velocity vector to obtain a right rear wheel bicycle motion equation group;

[0031] Constructing a center speed bicycle motion matrix according to the left front wheel bicycle motion equation group, the left rear wheel bicycle motion equation group, the right front wheel bicycle motion equation group and the right rear wheel bicycle motion equation group;

[0032] The coefficient matrix in the center speed bicycle motion matrix is ​​inverted to obtain the bicycle kinematic model.

[0033] In some embodiments, when the target kinematics model is a multi-vehicle kinematics model, constructing the target kinematics model according to the relative positions of the vehicles and the center of mass of the structure includes:

[0034] Calculate the distance from the cargo to the cargo rotation point based on the expected linear velocity and the expected angular velocity;

[0035] Decomposing the expected linear speed to obtain a group of decomposition equations of the cargo linear speed;

[0036] Constructing a vehicle linear velocity decomposition equation group according to the distance from the cargo to the cargo rotation point and the expected angular velocity;

[0037] The multi-vehicle kinematic model is constructed according to the distance from the cargo to the cargo rotation point, the relative position of the vehicle and the center of mass of the structure, the cargo linear velocity decomposition equation group and the vehicle linear velocity decomposition equation group.

[0038] In some embodiments, the method of calculating the vehicle position and vehicle posture using an extended Kalman filter algorithm based on the vehicle sensor data includes:

[0039] Using the vehicle sensor data as input to the extended Kalman filter algorithm, calculating a priori estimates of a state vector and a priori estimated covariance matrix;

[0040] Calculating a Kalman gain matrix based on the a priori estimated covariance matrix;

[0041] Calculating a posteriori estimates of state variables according to the prior estimates and the Kalman gain matrix;

[0042] Calculating a posterior estimation covariance matrix according to the prior estimation covariance matrix and the Kalman gain matrix;

[0043] The state variable posterior estimation value and the posterior estimation covariance matrix are output processed to obtain the vehicle position and the vehicle posture.

[0044] In some embodiments, constructing map data according to radar data includes:

[0045] According to the local synchronous positioning map, the global synchronous positioning map and the gravity direction, the radar data is projected into a two-dimensional plane to obtain a sub-map;

[0046] Calculate the position and posture of the radar frame in the sub-image coordinate system according to the sub-image, the rotation angle and the observation value;

[0047] Calculating a hit ratio according to the positions and poses of the sub-image and the radar frame in the sub-image coordinate system;

[0048] According to the hit ratio, construct a hit probability update equation and a miss probability update equation;

[0049] Calculating the matching probability between the scan point and the sub-graph according to the nonlinear least squares equation, the hit probability update equation and the miss probability update equation;

[0050] According to the matching probability between the scan points and the sub-map, the scan data is matched with the sub-map to obtain the map data.

[0051] In some embodiments, performing path planning according to a preset grid map to obtain a global path and a local path includes:

[0052] According to the preset grid map and the delta algorithm, determine whether there is an obstacle on the straight path from the parent node to the neighbor node;

[0053] If there are no obstacles on the straight path from the parent node to the neighbor node, the grid nodes are smoothly connected to obtain the global path;

[0054] According to the preset grid map, the local path is obtained by combining the linear speed adjusted by distance and curvature with the speed-based look-ahead using an adjusted pure pursuit algorithm and combining it with a time-scaled collision detector.

[0055] In some embodiments, the calculating the driving parameters through a target selector according to the vehicle position, the vehicle posture, the map data, the global path and the local path includes:

[0056] When the target selector is a behavior selector, environmental obstacles are identified according to the laser radar perception information and the map data to obtain the target obstacle position, where the target obstacle position includes a static obstacle position and a dynamic obstacle position;

[0057] Calculating the relative position of the transport vehicle according to the vehicle position and the vehicle posture;

[0058] Calculating the location of the cargo based on the relative positions of the plurality of transport vehicles;

[0059] Calculating the distance between the cargo and the obstacle and the positional relationship between the cargo and the obstacle according to the target obstacle position and the cargo position;

[0060] When the target selector is a speed selector, the frame line equations are calculated based on the cargo length and cargo width;

[0061] Calculate the distance from the obstacle point to the cargo based on the frame line equation group and the coordinates of the obstacle point;

[0062] According to the global path and the local path, determining whether the distance from the obstacle point to the cargo is less than the distance of the terminating operation range;

[0063] If the distance from the obstacle point to the cargo is less than the distance within the range of termination of operation, the operation is terminated; otherwise, it is determined whether the distance from the obstacle point to the cargo is less than the distance within the range of required deceleration;

[0064] If the distance from the obstacle to the cargo is less than the range of distance that requires deceleration, the driving speed is calculated according to the deceleration ratio;

[0065] When the target selector is a posture corrector, the vehicle linear velocity and the vehicle angular velocity are calculated according to the cargo posture information and the vehicle posture information;

[0066] Calculating the vehicle position coordinates and the rotation center point offset according to the vehicle linear velocity and the vehicle angular velocity;

[0067] The tilt movement distance, the tilt synthesis speed and the tilt steering angle are calculated based on the tilt linear speed and the tilt angular speed.

[0068] On the other hand, an embodiment of the present invention further provides a coordinated transport device, including:

[0069] A connecting plate, the connecting plate is used to carry goods;

[0070] an air spring connected to the connecting plate;

[0071] A connecting shaft connected to the air spring;

[0072] A bearing seat connected to the connecting shaft and arranged on an upper layer of the device;

[0073] A laser radar, wherein the laser radar is used to monitor a first driving state;

[0074] an inertial sensor, wherein the inertial sensor is used to monitor a second driving state;

[0075] an angular velocity sensor, the angular velocity sensor being connected to the bearing seat and being used for monitoring a third driving state;

[0076] A computer host, wherein the computer host is used to execute the multi-vehicle coordinated transportation method according to the first driving state, the second driving state and the third driving state;

[0077] A display, the display being used to display driving parameters;

[0078] A voltage stabilizer, the voltage stabilizer is used to output a stable voltage;

[0079] A router is used to provide communication connection between the laser radar, the inertial sensor, the angular velocity sensor, the display, the voltage stabilizer and the computer host.

[0080] On the other hand, an embodiment of the present invention further provides a multi-vehicle coordinated transportation system, including:

[0081] Goods to be transported and at least two of the cooperative transport devices.

[0082] The beneficial effects of the present invention are as follows:

[0083] The present invention first calculates the relative position of the vehicle and the center of mass of the structure and the desired state of the virtual structure according to the three-dimensional rectangular coordinate system, and constructs a target kinematic model according to the relative position of the vehicle and the center of mass of the structure, and then calculates the vehicle position and vehicle posture according to the vehicle sensor data using an extended Kalman filter algorithm, and constructs map data according to radar data, and then performs path planning according to a preset grid map to obtain a global path and a local path, and calculates driving parameters through a target selector according to the vehicle position, vehicle posture, map data, the global path and the local path, and finally performs collaborative transportation according to the desired state of the virtual structure, the target kinematic model and the driving parameters, so that collaborative transportation can be performed by controlling vehicle parameters and cargo posture, thereby improving collaborative accuracy, driving stability and transportation efficiency.

[0084] Other features and advantages of the present invention will be described in the following description, and partly become apparent from the description, or understood by practicing the present invention. The purpose and other advantages of the present invention can be realized and obtained by the structures particularly pointed out in the description and the drawings. BRIEF DESCRIPTION OF THE DRAWINGS

[0085] In order to more clearly illustrate the technical solutions in the embodiments of the present application, the drawings required for use in the description of the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present application. For ordinary technicians in this field, other drawings can be obtained based on these drawings without creative work.

[0086] Figure 1 This is a flow chart of a multi-vehicle coordinated transportation method according to an embodiment of the present invention;

[0087] Figure 2 A schematic diagram of a three-dimensional coordinate system of a structure according to an embodiment of the present invention;

[0088] Figure 3 A schematic diagram of a multi-vehicle kinematic modeling according to an embodiment of the present invention;

[0089] Figure 4 This is a schematic diagram of a single vehicle local posture correction during multi-vehicle cooperative transportation according to an embodiment of the present invention;

[0090] Figure 5 A schematic diagram of an overall process of multi-vehicle cooperative transportation according to an embodiment of the present invention;

[0091] Figure 6 This is a schematic diagram of the structure of a coordinated transport device according to an embodiment of the present invention;

[0092] Figure 7 A schematic diagram of an unmanned transport platform framework structure according to an embodiment of the present invention;

[0093] Figure 8 This is a schematic diagram of the structure of a multi-vehicle cooperative transportation system according to an embodiment of the present invention;

[0094] Fig. 9 It is a schematic top view of the relative position relationship between cargo and a collaborative transport device according to an embodiment of the present invention. DETAILED DESCRIPTION

[0095] In order to make the purpose, technical solutions and advantages of the present application clearer, the present application is further described in detail below in conjunction with the accompanying drawings and examples. It should be understood that the specific embodiments described herein are only used to explain the present application and are not intended to limit the present application. When the following description refers to the drawings, unless otherwise indicated, the same numbers in different drawings represent the same or similar elements. The embodiments described in the following exemplary embodiments do not represent all embodiments consistent with the embodiments of the present application. They are only examples of devices and methods consistent with some aspects of the embodiments of the present application as detailed in the attached claims.

[0096] It is understood that the terms "first", "second", etc. used in this application can be used to describe various concepts in this article, but unless otherwise specified, these concepts are not limited by these terms. These terms are only used to distinguish one concept from another concept. For example, without departing from the scope of the embodiment of the present application, the first information may also be referred to as the second information, and similarly, the second information may also be referred to as the first information. Depending on the context, the words "if" and "if" as used herein can be interpreted as "at the time of" or "when" or "in response to determination".

[0097] The terms "at least one", "multiple", "each", "any", etc. used in this application, at least one includes one, two or more, multiple includes two or more, each refers to each of the corresponding multiple, and any refers to any one of the multiple.

[0098] Unless otherwise defined, all technical and scientific terms used herein have the same meaning as those commonly understood by those skilled in the art to which this application belongs. The terms used herein are only for the purpose of describing the embodiments of this application and are not intended to limit this application.

[0099] Before describing the embodiments of the present application in detail, some nouns and terms involved in the embodiments of the present application are first described. The nouns and terms involved in the embodiments of the present application are subject to the following explanations.

[0100] Collaborative transportation: refers to the use of multiple transport vehicles or transport robots to transport a large piece of cargo at the same time. During the transportation process, multiple transport vehicles are required to travel in coordination, and the driving parameters of each transport vehicle, including speed, direction, acceleration, rotation angle, angular velocity, etc., need to be controlled to keep the cargo stable.

[0101] In the relevant technology, under the environment of rapid development in various fields, the demand for large-scale cargo transportation is growing. At the same time, this challenge not only needs to meet the requirements of efficient and safe transportation, but also needs to comply with the various regulations of the transportation system. Therefore, the present invention aims to find and implement an innovative and effective solution to the problems existing in the industry. Due to the large size of oversized cargo, it even requires the cooperation of multiple vehicles for transportation. The traditional manual transportation method has low transportation efficiency, low coordination and synchronization accuracy, and slow speed. At the same time, the road transportation of large cargo is affected by the coupling of cargo-vehicle-road-environment. There is a lack of vehicle-road information coordination and interaction during transportation, which may collide with obstacles and cause cargo to fall. The vehicle has low driving stability, low coordination accuracy, low transportation efficiency, and cannot ensure safety during transportation. For example, the existing "adaptive unmanned off-road platform" has a fixed structure and limited steering, and cannot flexibly adapt to different transport targets. At the same time, it cannot autonomously realize convoy transportation, which increases transportation costs.

[0102] To solve the above problems, the goal of the present invention is to develop and build an unmanned transport platform with a four-wheel drive system to realize multi-vehicle coordinated transportation of large goods, which solves the current problems of huge labor costs, low economic benefits, and poor adaptability to work scenarios in the transportation of large goods. It can carry out coordinated transportation by controlling vehicle parameters and cargo posture, thereby improving coordinated accuracy, driving stability and transportation efficiency, and improving the efficiency, accuracy and safety of large-scale transportation fleet formations.

[0103] A multi-vehicle collaborative transportation method provided in an embodiment of the present application relates to the field of unmanned transportation technology. A multi-vehicle collaborative transportation method provided in an embodiment of the present application can be applied to a terminal, can be applied to a server, or can be software running in a terminal or a server. In some embodiments, the terminal can be a smart phone, a tablet computer, a laptop computer, a desktop computer, a smart speaker, a smart watch, and a vehicle-mounted terminal, etc., but is not limited to this; the server side can be configured as an independent physical server, or can be configured as a server cluster or distributed system composed of multiple physical servers, and can also be configured as a cloud server that provides basic cloud computing services such as cloud services, cloud databases, cloud computing, cloud functions, cloud storage, network services, cloud communications, middleware services, domain name services, security services, CDN, and big data and artificial intelligence platforms. The server can also be a node server in a blockchain network; the software can be an application that implements a multi-vehicle collaborative transportation method, etc., but is not limited to the above forms.

[0104] The present application can be used in many general or special computer system environments or configurations. For example: personal computers, server computers, handheld or portable devices, tablet devices, multiprocessor systems, microprocessor-based systems, set-top boxes, programmable consumer electronics, network PCs, minicomputers, mainframe computers, distributed computing environments including any of the above systems or devices, etc. The present application can be described in the general context of computer-executable instructions executed by a computer, such as program modules. Generally, program modules include routines, programs, objects, components, data structures, etc. that perform specific tasks or implement specific abstract data types. The present application can also be practiced in distributed computing environments, in which tasks are performed by remote processing devices connected through a communication network. In a distributed computing environment, program modules can be located in local and remote computer storage media including storage devices.

[0105] The following is a detailed explanation of the embodiments of the present application in conjunction with the accompanying drawings:

[0106] Figure 1 is an optional flow chart of a multi-vehicle cooperative transportation method provided in an embodiment of the present application. Figure 1The method may include but is not limited to steps S101 to S107.

[0107] Step S101, calculating the relative position of the vehicle and the center of mass of the structure and the desired state of the virtual structure according to the three-dimensional rectangular coordinate system;

[0108] Step S102: construct a target kinematic model according to the relative position of the vehicle and the structure mass center, wherein the target kinematic model includes a single-vehicle kinematic model, a single-vehicle oblique shift kinematic model, and a multi-vehicle kinematic model;

[0109] Step S103, calculating the vehicle position and vehicle posture using an extended Kalman filter algorithm based on the vehicle sensor data;

[0110] Step S104: constructing map data according to radar data;

[0111] Step S105: Perform path planning according to the preset grid map to obtain a global path and a local path;

[0112] Step S106, according to the vehicle position, vehicle posture, map data, global path and local path, the driving parameters are calculated by the target selector, the target selector includes a behavior selector, a speed selector, and a posture corrector, and the driving parameters include the distance between the cargo and the obstacle, the position relationship between the cargo and the obstacle, the driving speed, the vehicle position coordinates, the rotation center point offset, the oblique movement distance, the oblique movement synthesis speed and the oblique movement steering angle;

[0113] Step S107: performing collaborative transportation according to the desired state of the virtual structure, the target kinematic model and the driving parameters.

[0114] Steps S101 to S107 shown in the embodiment of the present application realize collaborative transportation and improve collaborative accuracy, driving stability and transportation efficiency.

[0115] In some embodiments, in step S101, the relative position between the vehicle and the structure's mass center and the desired state of the virtual structure are calculated according to the three-dimensional rectangular coordinate system, which may include but is not limited to the following steps:

[0116] According to the three-dimensional rectangular coordinate system, the center of mass position equations are constructed;

[0117] According to the support mode and the center of mass position equation group, the center of each sub-support area is connected to obtain the stable support area;

[0118] According to the stable support area, calculate the three-side length equations of the supporting triangle;

[0119] Calculate the perimeter of the supporting triangle based on the three side length equations;

[0120] Based on the perimeter, calculate the area of ​​the supporting triangle;

[0121] Based on the area and centroid height, calculate the support angle;

[0122] Calculate the relative position of the vehicle and the structure's mass center based on multiple support angles;

[0123] According to the three-dimensional rectangular coordinate system, the local virtual structure dynamic model is constructed by using the virtual structure method;

[0124] According to the local virtual structure dynamics model, a local virtual structure state vector is constructed;

[0125] According to the local virtual structure state vector, the virtual structure expected state is constructed.

[0126] In some embodiments, a three-dimensional rectangular coordinate system may be first established, and a group of mass center position equations may be constructed based on the three-dimensional rectangular coordinate system, wherein the expression of the group of mass center position equations is: In the formula, is the center of mass position of the cargo, ρ(x, y, z) is the three-dimensional density function, M is the total weight of the cargo, and (x, y, z) is the three-dimensional coordinate point. Then, according to the support mode and the center of mass position equation group, the center of each sub-support area is connected, and four points are used to support the cargo. The line connecting the centers of every two support areas is the tipping line, and the area surrounded by the tipping line is the stable support area. Furthermore, the stability angle can be used to determine the risk of the center of mass of the structure deviating from the stable support area. Then establish a three-dimensional coordinate system for the structure, such as Figure 2 As shown, the ABCD quadrilateral (or triangle) is the stable support area, G' is G(COG), and the projection on the ABCD quadrilateral can calculate the coordinates of each point G(X, Y, Z), G'(X, Y), A(XA, YA), B(XB, YB), C(XC, YC), and D(XD, YD). Given A(XA, YA) and B(XB, YB), the three-side length equations of the supporting triangle can be calculated based on the stable support area. The calculation formula of the three-side length equations of the supporting triangle is: Where, L AB , L AG` , L BG` is the equation system of the three sides of the supporting triangle, (X A , Y A ) is the coordinate of point A, (X B , Y B ) is the coordinate of point B, (X G` , Y G` ) is the coordinate of point G`. Then, according to the three-side length equation group, the perimeter of the supporting triangle is calculated, where the calculation formula for the perimeter of the supporting triangle is: L AG`B =L AB+L AG` +L BG` , where L AG`B is the perimeter of the supporting triangle. According to the perimeter and Heron's theorem, the area of ​​the supporting triangle is calculated, where the calculation formula for the area of ​​the supporting triangle is: In the formula, S ΔAG`B is the area of ​​the supporting triangle, and p is 1 / 2 of the perimeter. Then, the support angle is calculated based on the area and the centroid height. The calculation formula for the support angle is: Where β is the support angle, L2=2S ΔAG`B / L AB , H=Z+h0, H is the height of the center of mass, Z is the height of the center of mass of the transport structure from the vehicle plate, and h0 is the lifting height of the vehicle plate. Similarly, the values ​​of α, γ, and η can be calculated, and based on multiple support angles, the relative position of the vehicle and the center of mass of the structure, that is, the position of each vehicle relative to the center of mass of the transport structure, can be calculated through the values ​​of α, β, γ, and η. In some embodiments, a virtual structure method can be used to define a local (i.e., local) formation virtual structure for each module unmanned vehicle, called a local virtual structure. The formation control algorithm gradually converges all local virtual structures through communication and coordination between robots. According to the three-dimensional rectangular coordinate system, the virtual structure method is used to construct the local virtual structure dynamics model of the i-th robot, where the expression of the local virtual structure dynamics model is: In the formula, Indicates the speed of the vehicle at a certain moment, describing its moving speed on the plane. represents the rate of change of vehicle speed, i.e. acceleration, represents the rate at which the vehicle rotates, describing its angular velocity, represents the rate of change of the vehicle's angular velocity, i.e., angular acceleration, For speed, is the angular velocity, mF is the virtual rotation mass of the virtual structure, J F is the virtual moment of inertia of the virtual structure, is the virtual force applied to the virtual structure, is the virtual torque applied to the virtual structure. Then, according to the local virtual structure dynamics model, the local virtual structure state vector is constructed, where the expression of the local virtual structure state vector is: In the formula, is the state vector of the local virtual structure of the ith robot, r T Fi is the virtual structure position vector, v T Fi is the virtual structure velocity vector, θ Fi is the deflection angle of the virtual structure, ω Fiis the angular velocity of the virtual structure. Then, according to the local virtual structure state vector, the desired state of the virtual structure is constructed, where the expression of the desired state of the virtual structure is: Where, X d f is the expected state of the virtual structure, is the desired position vector, is the desired velocity vector, θ d F is the desired deflection angle, ω d F is the desired angular velocity. It can be understood that X d f It is a description of the ideal state of the virtual structure. In the subsequent motion control, each vehicle will continuously adjust its actual state. To get as close to X as possible d f , ensuring the overall coordination and stability of the formation. The virtual structure of the formation contains virtual individuals corresponding to each actual individual in the formation, and the state of the virtual individual is the expected state of the actual individual; combined with V2X communication technology, the vehicles in the convoy maintain the geometric relationship of the formation, provide continuous perception between the convoy and its surrounding environment, detect changes and states of the surrounding environment in real time, and provide environmental basis for subsequent operations such as obstacle avoidance.

[0127] In some embodiments, in step S102, when the target kinematics model is a single-vehicle kinematics model, constructing the target kinematics model according to the relative positions of the vehicle and the center of mass of the structure may include but is not limited to the following steps:

[0128] According to the relative position of the center of mass of the vehicle and the structure, the velocity equation of the single-drive-wheel vehicle is constructed;

[0129] According to the single-drive wheel bicycle motion speed equation, construct the left front wheel bicycle motion decomposition speed vector, the left rear wheel bicycle motion decomposition speed vector, the right front wheel bicycle motion decomposition speed vector, and the right rear wheel bicycle motion decomposition speed vector;

[0130] Perform filling transformation on the decomposed velocity vector of the left front wheel bicycle motion to obtain the left front wheel bicycle motion equation group;

[0131] Perform filling transformation on the decomposed velocity vector of the left rear wheel bicycle motion to obtain the left rear wheel bicycle motion equation group;

[0132] Perform filling transformation on the decomposed velocity vector of the right front wheel bicycle motion to obtain the right front wheel bicycle motion equation group;

[0133] Perform filling transformation on the decomposed velocity vector of the right rear wheel bicycle motion to obtain the right rear wheel bicycle motion equation group;

[0134] According to the left front wheel bicycle motion equation group, the left rear wheel bicycle motion equation group, the right front wheel bicycle motion equation group and the right rear wheel bicycle motion equation group, a central speed bicycle motion matrix is ​​constructed;

[0135] The coefficient matrix in the center speed bicycle motion matrix is ​​inverted to obtain the bicycle kinematic model.

[0136] In some embodiments, in the basic control and navigation of a bicycle, the bicycle kinematic model is used to calculate and control the motion path of the vehicle to ensure that the vehicle travels along the predetermined trajectory. When creating a coordinate system, the CENTER coordinate system can be established with the geometric center of the robot as the origin. The robot's forward direction is defined as the positive direction of the X-axis; the direction perpendicular to the left is defined as the positive direction of the Y-axis; and the Z-axis extends outward perpendicular to the plane to comply with the right-hand rule. Using the linear velocity and angular velocity in any direction [v c ω] T To describe the speed of CENTER. Where the linear speed v c It can be decomposed into two component velocities along the coordinate axis, expressed as [v cx v cy ω] T . Linear speed [v cx v cy ] T The direction along the positive direction of the coordinate axis is positive, and the reverse direction is negative; when the robot rotates counterclockwise, the angular velocity is positive, and vice versa. c ω] T , decompose it into each driving wheel, and use the planar rigid body motion law analysis method. The speed of a single driving wheel is synthesized by the center point speed and angular velocity. The single driving wheel bicycle motion speed equation can be constructed based on the relative position of the vehicle and the structure mass center. The expression of the single driving wheel bicycle motion speed equation is: i =v c +ω c × i (i=1,2,3,4), where v i represents the total speed of the ith driving wheel, v c represents the linear velocity of the center point, ω c represents the angular velocity of the center point, r i represents the vector from the center point to the contact point between the i-th driving wheel and the ground. It can be understood that ω c ×r represents the angular velocity ω of the robot cThe speed generated by the spin. The direction and magnitude of these velocity components follow the cross multiplication calculation rule, and the equations all use vector operations, which conform to the parallelogram law. Then, according to the single-drive wheel bicycle motion speed equation, the left front wheel bicycle motion decomposition speed vector, the left rear wheel bicycle motion decomposition speed vector, the right front wheel bicycle motion decomposition speed vector, and the right rear wheel bicycle motion decomposition speed vector are constructed. Among them, the driving wheel f l (left front wheel) as an example for analysis. l The total velocity along the coordinate system xf l Y decomposition, the expression of the velocity vector of the left front wheel bicycle motion decomposition is: In the formula, [v 1x v 1y ] T Indicates that v1 is in the coordinate system xf l The y component of velocity, [v cx v cy ] T Indicates v c The component speed, [r x r y ] T Indicates that r1 is in the coordinate system xf l y corresponds to the component under the coordinate axis, θ1 represents the component between v1 and the coordinate axis f l x angle, v1 represents the total speed of the first driving wheel (left front wheel). l (left rear wheel) for analysis l The velocity is along the coordinate system xr l Y decomposition, the expression of the velocity vector of the left rear wheel bicycle motion decomposition is: In the formula, [v 2x v 2y ] T Indicates v2 in coordinate system xr l The y component of velocity, [v cx v cy ] T Indicates v c The component speed, [r x r y ] T Represents r2 in the coordinate system xr l y corresponds to the component under the coordinate axis, θ2 represents the component between v2 and the coordinate axis xr l Angle, v2 represents the combined speed of the second driving wheel (left rear wheel). r (right rear wheel) for analysis, and point r r The total velocity along the coordinate system xr r Y decomposition, the expression of the decomposition velocity vector of the right rear wheel bicycle motion is: In the formula, [v3x v 3y ]T represents v3 in the coordinate system xr r The y component of velocity, [v cx v cy ] T Indicates v c The component speed, [r x r y ] T Represents r3 in the coordinate system xr r y corresponds to the component under the coordinate axis, θ3 represents the component between v3 and the coordinate axis xr r Angle, v3 represents the combined speed of the third driving wheel (right rear wheel). r (right front wheel) for analysis, point f r The total velocity along the coordinate system xf r Y decomposition, the expression of the decomposition velocity vector of the right front wheel bicycle motion is: In the formula, [v 4x v 4y ] T Indicates v4 in coordinate system xf r The y component of velocity, [v cx v cy ] T Indicates v c The component speed, [r x r y ] T Indicates that r4 is in the coordinate system xf r y corresponds to the component under the coordinate axis, θ4 represents the component between v4 and the coordinate axis xf r Angle, v4 represents the combined speed of the fourth driving wheel (right front wheel). Then the bicycle kinematic model can be integrated into the equation relationship between the speed of the robot center point and the speed of the four wheels, and the left front wheel bicycle motion decomposition speed vector is filled and transformed to obtain the left front wheel bicycle motion equation group, where the expression of the left front wheel bicycle motion equation group is: The left rear wheel bicycle motion decomposition velocity vector is filled and transformed to obtain the left rear wheel bicycle motion equation group, where the expression of the left rear wheel bicycle motion equation group is: The right rear wheel bicycle motion equation group is obtained by filling and transforming the decomposed velocity vector of the right rear wheel bicycle motion. The expression of the right rear wheel bicycle motion equation group is: The right front wheel bicycle motion equation group is obtained by filling and transforming the decomposed velocity vector of the right front wheel bicycle motion. The expression of the right front wheel bicycle motion equation group is: Then, according to the left front wheel bicycle motion equation group, the left rear wheel bicycle motion equation group, the right front wheel bicycle motion equation group and the right rear wheel bicycle motion equation group, the center speed bicycle motion matrix is ​​constructed, where the expression of the center speed bicycle motion matrix is: In the formula, It can be understood that in the center speed bicycle motion matrix, the left side is the center speed and the right side is the speed of the four wheels, each with its own coefficient matrix K and T, which can be filled in according to the matrix operation rules. Finally, the coefficient matrix in the center speed bicycle motion matrix is ​​inverted to obtain the bicycle kinematic model, that is, the forward motion model, where the inverse expression of the coefficient matrix is: In the formula, K + is the inverse of the coefficient matrix, r 2 Represents the square of the characteristic length of the vehicle chassis. After inverting the coefficient matrix, the expression of the bicycle kinematic model is obtained as follows: In the formula, Furthermore, the inverse kinematic model can be calculated based on the bicycle kinematic model (forward kinematic model), where the expression of the inverse kinematic model is:

[0137] In some embodiments, in step S102, when the target kinematics model is a single-vehicle oblique displacement kinematics model, constructing the target kinematics model according to the relative position of the vehicle and the center of mass of the structure may include but is not limited to the following steps:

[0138] According to the relative position of the center of mass of the vehicle and the structure, the velocity equation of the oblique displacement of a vehicle with a single driving wheel is constructed;

[0139] According to the single-drive-wheel bicycle oblique displacement velocity equation, construct the left front wheel bicycle oblique displacement velocity vector, the left rear wheel bicycle oblique displacement velocity vector, the right front wheel bicycle oblique displacement velocity vector, and the right rear wheel bicycle oblique displacement velocity vector;

[0140] Perform filling transformation on the decomposed velocity vector of the left front wheel bicycle oblique displacement to obtain the left front wheel bicycle oblique displacement equation group;

[0141] Perform filling transformation on the decomposed velocity vector of the left rear wheel bicycle oblique displacement to obtain the left rear wheel bicycle oblique displacement equation group;

[0142] Perform filling transformation on the decomposed velocity vector of the right front wheel bicycle oblique displacement to obtain the right front wheel bicycle oblique displacement equation group;

[0143] Perform filling transformation on the decomposed velocity vector of the right rear wheel bicycle oblique displacement to obtain the right rear wheel bicycle oblique displacement equation group;

[0144] According to the left front wheel bicycle oblique shift equation group, the left rear wheel bicycle oblique shift equation group, the right front wheel bicycle oblique shift equation group and the right rear wheel bicycle oblique shift equation group, a central speed bicycle oblique shift matrix is ​​constructed;

[0145] The coefficient matrix in the center speed bicycle tilt matrix is ​​inverted to obtain the bicycle tilt kinematic model.

[0146] In some embodiments, in the bicycle tilt kinematic model, the velocity of CENTER is known to be v c , which can be decomposed into two component velocities along the coordinate axis, expressed as [v cx v cy ] T , and decompose it to each driving wheel, then the speed of a single driving wheel is synthesized by the speed of the point CENTER. We can first construct the single driving wheel single vehicle oblique speed equation based on the relative position of the vehicle and the structure mass center, where the single driving wheel single vehicle oblique speed equation is expressed as: i =v cx +v cy (i=1,2,3,4), where v i represents the total speed of the ith driving wheel, v cx represents the component velocity of the linear velocity of the center point in the x direction, v cy represents the component velocity of the linear velocity of the center point in the y direction. Then, according to the single driving wheel bicycle oblique displacement velocity equation, the left front wheel bicycle oblique displacement velocity vector, the left rear wheel bicycle oblique displacement velocity vector, the right front wheel bicycle oblique displacement velocity vector, and the right rear wheel bicycle oblique displacement velocity vector are constructed. Among them, the driving wheel f l (left front wheel) as an example for analysis. l The total velocity along the coordinate system xf l Y decomposition, the expression of the decomposition velocity vector of the left front wheel bicycle oblique displacement is: In the formula, [v 1x v 1y ] T Indicates that v1 is in the coordinate system xf l The y component of velocity, [v cx v cy ] T Indicates v c The velocity component of v1 and the coordinate axis f l x angle, v1 represents the total speed of the first driving wheel (left front wheel). l (left rear wheel) for analysis l The velocity is along the coordinate system xr l Y decomposition, the expression of the decomposition velocity vector of the left rear wheel bicycle oblique displacement is: In the formula, [v 2x v 2y ] T Indicates v2 in coordinate system xr l The y component of velocity, [v cx v cy ] T Indicates v cThe velocity component of v2 and the coordinate axis xr l Angle, v2 represents the combined speed of the second driving wheel (left rear wheel). r (right rear wheel) for analysis, and point r r The total velocity along the coordinate system xr r Y decomposition, the expression of the decomposition velocity vector of the right rear wheel bicycle oblique displacement is: In the formula, [v 3x v 3y ]T represents v3 in the coordinate system xr r The y component of velocity, [v cx v cy ]T represents v c The velocity component of v3 and the coordinate axis xr r Angle, v3 represents the combined speed of the third driving wheel (right rear wheel). r (right front wheel) for analysis, point f r The total velocity along the coordinate system xf r Y decomposition, the expression of the decomposition velocity vector of the right front wheel bicycle oblique displacement is: In the formula, [v 4x v 4y ]T represents v4 in the coordinate system xf r The y component of velocity, [v cx v cy ] T Indicates v c The velocity component of v4 and the coordinate axis xf r The angle, v4 represents the combined speed of the fourth driving wheel (right front wheel). Then the bicycle oblique transfer model can be integrated into the equation relationship between the speed of the robot center point and the speed of the four wheels. The left front wheel bicycle oblique transfer decomposition speed vector is filled and transformed to obtain the left front wheel bicycle oblique transfer equation group, where the expression of the left front wheel bicycle oblique transfer equation group is: The left rear wheel bicycle oblique displacement decomposition velocity vector is filled and transformed to obtain the left rear wheel bicycle oblique displacement equation group, where the expression of the left rear wheel bicycle oblique displacement equation group is:

[0147] The right rear wheel bicycle oblique displacement decomposition velocity vector is filled and transformed to obtain the right rear wheel bicycle oblique displacement equation group, where the expression of the right rear wheel bicycle oblique displacement equation group is:

[0148] The right front wheel bicycle oblique displacement decomposition velocity vector is filled and transformed to obtain the right front wheel bicycle oblique displacement equation group, where the expression of the right front wheel bicycle oblique displacement equation group is: Then, according to the left front wheel bicycle oblique shift equation group, the left rear wheel bicycle oblique shift equation group, the right front wheel bicycle oblique shift equation group and the right rear wheel bicycle oblique shift equation group, the center speed bicycle oblique shift matrix is ​​constructed, wherein the expression of the center speed bicycle oblique shift matrix is:

[0149] In the formula, It can be understood that in the center speed bicycle tilt matrix, the left side is the center speed and the right side is the speed of the four wheels, each with its own coefficient matrix K and T, which can be filled in according to the matrix operation rules. Finally, the coefficient matrix in the center speed bicycle tilt matrix is ​​inverted to obtain the bicycle tilt kinematic model, that is, the forward motion model, where the inverse expression of the coefficient matrix is: In the formula, K + is the inverse of the coefficient matrix. After inverting the coefficient matrix, the expression of the bicycle oblique shift model is obtained as follows: In the formula, Furthermore, the inverse kinematic model can be calculated based on the bicycle oblique displacement kinematic model (forward kinematic model), where the expression of the inverse kinematic model is:

[0150] In some embodiments, in step S102, when the target kinematics model is a multi-vehicle kinematics model, constructing the target kinematics model according to the relative positions of the vehicles and the center of mass of the structure may include but is not limited to the following steps:

[0151] Calculate the distance from the cargo to the cargo rotation point based on the expected linear velocity and the expected angular velocity;

[0152] Decompose the expected linear speed to obtain the decomposition equations of the cargo linear speed;

[0153] Construct the vehicle linear velocity decomposition equations according to the distance from the cargo to the cargo rotation point and the expected angular velocity;

[0154] The multi-vehicle kinematic model is constructed based on the distance from the cargo to the cargo rotation point, the relative position of the vehicle and the structure's center of mass, the cargo linear velocity decomposition equation group and the vehicle linear velocity decomposition equation group.

[0155] In some embodiments, a coordinate system may be established with the cargo center as the origin. By decomposing the speed of the cargo center, the component speeds of each vehicle in the x and y directions are calculated, thereby obtaining the combined speed of each vehicle. Figure 3 As shown in the figure, dark blue represents vehicles and light blue represents cargo. Taking vehicle No. 1 as an example, the distance r from the cargo to the cargo rotation point can be calculated using the expected linear velocity v and expected angular velocity ω of the cargo center point. The calculation formula for the distance from the cargo to the cargo rotation point is: In the formula, r is the distance from the cargo to the cargo rotation point, v is the expected linear velocity, and ω is the expected angular velocity. It can be understood that expectation refers to the expected motion state during the control process. Then the expected linear velocity is decomposed to obtain the cargo linear velocity decomposition equation group, where the expression of the cargo linear velocity decomposition equation group is: In the formula, (v x , v y ) is the decomposition of the linear velocity, and θ is the desired angle. Then, according to the distance from the cargo to the cargo rotation point and the desired angular velocity, the vehicle linear velocity decomposition equation group is constructed, where the expression of the vehicle linear velocity decomposition equation group is: In the formula, (v ix , v iy ) is the decomposition of the linear velocity of the i-th vehicle, ω is the desired angular velocity, r ix is the distance from the cargo to the cargo rotation point in the x direction, r iy is the component of the distance from the cargo to the cargo rotation point in the y direction. Finally, according to the distance from the cargo to the cargo rotation point, the relative position of the vehicle and the structure mass center, the cargo linear velocity decomposition equation group and the vehicle linear velocity decomposition equation group, a multi-vehicle kinematic model is constructed, where the expression of the multi-vehicle kinematic model is: Wherein, Steer_point represents the point around which the vehicle rotates, also called the center of rotation (or turning point), i.e., the center point of the vehicle when it is rotating, a represents the wheelbase of the vehicle, i.e., the distance between the front axle and the rear axle, b represents the track of the vehicle, i.e., the lateral distance between the left and right wheels, v1 is the updated linear velocity of the vehicle, and ω1 is the updated angular velocity of the vehicle.

[0156] In some embodiments, in step S103, the vehicle position and vehicle posture are calculated using the extended Kalman filter algorithm according to the vehicle sensor data, which may include but is not limited to the following steps:

[0157] The vehicle sensor data is used as the input of the extended Kalman filter algorithm to calculate the prior estimate of the state vector and the prior estimate covariance matrix;

[0158] Calculate the Kalman gain matrix based on the prior estimated covariance matrix;

[0159] Calculate the posterior estimate of the state variable according to the prior estimate and the Kalman gain matrix;

[0160] Calculate the posterior estimated covariance matrix according to the prior estimated covariance matrix and the Kalman gain matrix;

[0161] The posterior estimated value of the state variable and the posterior estimated covariance matrix are outputted to obtain the vehicle position and vehicle posture.

[0162] In some embodiments, sensor fusion positioning can use an extended Kalman filter algorithm to fuse multiple vehicle sensor data, where the vehicle sensor data can include wheel odometer data, IMU inertial measurement unit data, GPS global positioning system data, and visual odometer data. The vehicle sensor data can be used as the input of the extended Kalman filter algorithm to calculate the prior estimate of the state vector and the prior estimate covariance matrix, where the calculation formula of the prior estimate of the state vector is: In the formula, is the prior estimate of the state vector, is the posterior estimate of the state vector at time t, u t+1 is the control matrix. The calculation formula of the prior estimated covariance matrix is: In the formula, is the a priori estimated covariance matrix, is the Jacobian matrix of the kinematic model, p t is the posterior estimated covariance matrix at time t, R t+1 is the dynamic noise covariance matrix. Then, the Kalman gain matrix is ​​calculated based on the prior estimated covariance matrix, where the calculation formula of the Kalman gain matrix is: In the formula, K t+1 is the Kalman gain matrix, is the Jacobian matrix of the observation model, Q t+1 is the observation noise covariance matrix. Then, according to the prior estimate and the Kalman gain matrix, the posterior estimate of the state variable is calculated, where the calculation formula of the posterior estimate of the state variable is: In the formula, is the posterior estimate of the state variable, z t+1 is the observation value at the t+1th moment, h(*) is the observation function, which describes the relationship between the state and the observation. Then, according to the prior estimation covariance matrix and the Kalman gain matrix, the posterior estimation covariance matrix is ​​calculated, where the calculation formula of the posterior estimation covariance matrix is: In the formula, p t+1 is the posterior estimated covariance matrix, and I is the unit matrix with the same dimension as the covariance matrix. Finally, the posterior estimated value of the state variable and the posterior estimated covariance matrix are outputted to obtain the vehicle position and vehicle posture. It can be understood that in the extended Kalman filter algorithm, the Kalman extended filter motion equation is: and In the formula, x t+1 represents the state variable at time t+1, x t represents the state variable at time t, u t+1 is the control matrix, f(x t ,u t+1 ) represents the state variable x at time t+1t+1 and the state variable x at time t t The functional relationship between t+1 Represents dynamic noise, assumed to be Gaussian white noise with a mean of 0 and a covariance of R t+1 , Z t+1 represents the observed value at time t+1, h(x t+1 ) represents the observation function, describing the relationship between the state and the observation, v t+1 is the observation noise, assuming it is a Gaussian white noise with a mean of 0 and a covariance of Q t+1 , R t+1 is the dynamic noise covariance matrix, Q t+1 is the observation noise covariance matrix. Then the state variables are expanded using the Taylor expansion formula, and the expansion formula is: In the formula, Expressed as the posterior estimate of the state vector at time t, is the Jacobian matrix of the kinematic model, whose value is the function exist u t+1 Conditions about x t The first-order derivative of . Among them, the calculation formula of the Jacobian matrix of the kinematic model is: Similarly, the observation equation can be expanded using the Taylor expansion formula as follows: In the formula, represents the prior estimate of the state vector at time t+1, is the Jacobian matrix of the observation model, whose value is the function exist Conditions about x t+1 The first-order derivative of . Among them, the calculation formula of the Jacobian matrix of the observation model is: It can be understood that the above expansion formula plays a key role in the prediction and update stages of the extended Kalman filter algorithm, ensuring the accuracy and reliability of multi-sensor fusion positioning, and can be widely used in robot positioning, navigation, multi-sensor fusion and other fields.

[0163] In some embodiments, in step S104, constructing map data according to radar data may include but is not limited to the following steps:

[0164] According to the local synchronous positioning map, the global synchronous positioning map and the gravity direction, the radar data is projected into a two-dimensional plane to obtain a sub-map;

[0165] According to the sub-image, rotation angle and observation value, calculate the position and pose of the radar frame in the sub-image coordinate system;

[0166] Calculate the hit ratio based on the poses of the sub-image and radar frame in the sub-image coordinate system;

[0167] According to the hit ratio, construct the hit probability update equation and the miss probability update equation;

[0168] According to the nonlinear least squares equation, the hit probability update equation and the miss probability update equation, the matching probability between the scan point and the sub-graph is calculated;

[0169] According to the matching probability between the scan points and the sub-map, the scan data is matched with the sub-map to obtain the map data.

[0170] In some embodiments, the radar data can be projected into a two-dimensional plane based on local synchronous positioning and mapping, global synchronous positioning and mapping, and gravity direction, using the gravity direction given by the IMU inertial measurement unit to obtain a sub-image. The scanning area of ​​each radar frame is called a sub-image, and nonlinear optimization scan matching can be used to align the scanning points and the sub-image. Then set the origin of the sub-image's coordinate system to 0∈R 2 , each radar scanning point is L, and the pose of the current radar frame in the sub-image coordinate system is T ξ , transfer the scanning point to the sub-image coordinate system, and calculate the position and pose of the radar frame in the sub-image coordinate system according to the sub-image, rotation angle and observation value. The calculation formula of the position and pose of the radar frame in the sub-image coordinate system is: Where, T ξp is the position of the radar frame in the sub-image coordinate system, ξ θ Represents the rotation angle, indicating the rotation of the current radar frame relative to the sub-image coordinate system; ξ x Represents the translation (X direction), which means the X-direction translation of the current radar frame relative to the sub-image coordinate system; ξ y represents the translation (Y direction), indicating the Y direction translation of the current radar frame relative to the sub-image coordinate system; Z represents the observation value, indicating the position of the radar scanning point in the sub-image coordinate system, and L is the radar scanning point. It can be understood that the sub-image is actually a probability grid map M: rZ×rZ→[p min , p max ], r is the grid map resolution, where p min Represents the lower limit of the free area; p max Represents the upper limit of the obstacle. For unobserved grids, you can customize a p miss or hit Represent the miss probability and hit probability respectively. Then, the hit ratio is calculated according to the position of the sub-image and radar frame in the sub-image coordinate system. The calculation formula of the hit ratio is: In the formula, odds(p) is the hit ratio, and p is the number of hits. In order to build and maintain an accurate probability map, reflect the obstacles and idle areas in the environment, and improve the map accuracy and stability, it is necessary to process the hit probability and the miss probability. The hit probability update equation and the miss probability update equation can be constructed based on the hit ratio. The expression of the hit probability update equation is: newi (x) = clamp[odds -1 (odds(M oldi (x))×odds(p hit ))|, where M newi (x) is the updated hit probability, M oldi (x) is the hit probability before updating, p hit is the hit probability of the unobserved grid. The expression of the miss probability update equation is: newj (x) = clamp[odds -1 (odds(M oldj (x))×odds(p miss ))], where M newj (x) is the updated miss probability, M oldj (x) is the miss probability before updating, p miss is the miss probability of the unobserved grid. The purpose of Scan Matching is to find the optimal pose so that the points in the Scan can be matched to the Submap with the maximum probability, which is a nonlinear least squares problem. The matching probability between the scan point and the submap can be calculated based on the nonlinear least squares equation, the hit probability update equation and the miss probability update equation. The expression of the matching probability between the scan point and the submap is: In the formula, h k Represents the scan point, indicating the kth point in the current radar scan data, M smooth represents the smoothed probability grid map, represents the sub-graph after smoothing, K is the total number of scan points, T ξ is the position and posture of the radar frame in the sub-image coordinate system. Finally, according to the matching probability between the scanning point and the sub-image, the scanning data is matched with the sub-image, and the matching is performed through SLAM mapping technology. The scanning to sub-image matching is used as a constraint to optimize the vehicle posture so that the scanning data is best matched with the sub-image to obtain map data to ensure accurate positioning and map updates, and improve navigation and mapping accuracy.

[0171] In some embodiments, in step S105, path planning is performed according to a preset grid map to obtain a global path and a local path, which may include but is not limited to the following steps:

[0172] According to the preset grid map and the delta algorithm, determine whether there are obstacles on the straight path from the parent node to the neighbor node;

[0173] If there are no obstacles on the straight path from the parent node to the neighbor node, the grid nodes are smoothly connected to obtain the global path;

[0174] According to the preset grid map, the adjusted pure pursuit algorithm is used to combine the linear speed adjusted by distance and curvature with the speed-based look-ahead, and combined with the time-scaled collision detector to obtain the local path.

[0175] In some embodiments, the path planning can be performed using the Theta* algorithm according to the preset grid map, and the path planning is performed on the drawn grid map, and the path from the parent node of the current node to its adjacent node is set. It is determined whether there is an obstacle on the straight path from the parent node to the neighbor node. If there is no obstacle on the straight path from the parent node to the neighbor node, it does not need to be added as a node in the path. It can be directly connected from the parent node to the neighbor node, and the grid nodes are smoothly connected. The smooth connection of the path is allowed to move diagonally with the underlying grid indirectly or non-indirectly to obtain a global path. The obtained global path will be sent to the motion module as a series of path points, and the car will be driven to start moving along the path. Then, according to the preset grid map, the Regulated Pure Pursuit algorithm is used to combine the linear speed adjusted by distance and curvature with the speed-based look-ahead to achieve smooth driving in narrow spaces. And by combining the linear speed adjusted by distance and curvature with the time-scaled collision detector, the regulated Pure Pursuit algorithm can handle high starting deviations on the path, and avoid collisions in narrow spaces without overshooting, and obtain a local path.

[0176] In some embodiments, in step S106, the driving parameters are calculated by the target selector according to the vehicle position, vehicle posture, map data, global path and local path, which may include but is not limited to the following steps:

[0177] When the target selector is a behavior selector, environmental obstacles are identified based on the lidar perception information and map data to obtain the target obstacle position, which includes the static obstacle position and the dynamic obstacle position;

[0178] Calculate the relative position of the transport vehicle based on the vehicle position and vehicle posture;

[0179] Calculate the location of cargo based on the relative positions of multiple transport vehicles;

[0180] According to the target obstacle position and cargo position, calculate the distance between cargo and obstacle and the position relationship between cargo and obstacle;

[0181] When the target selector is a speed selector, the frame line equations are calculated based on the cargo length and cargo width;

[0182] Calculate the distance from the obstacle point to the cargo based on the frame line equations and the coordinates of the obstacle point;

[0183] According to the global path and the local path, determine whether the distance from the obstacle point to the cargo is less than the distance of the termination operation range;

[0184] If the distance from the obstacle point to the cargo is less than the range distance for terminating the operation, the operation is terminated; otherwise, it is determined whether the distance from the obstacle point to the cargo is less than the range distance for deceleration;

[0185] If the distance from the obstacle to the cargo is less than the required deceleration range, the driving speed is calculated based on the deceleration ratio;

[0186] When the target selector is a posture corrector, the vehicle linear velocity and the vehicle angular velocity are calculated according to the cargo posture information and the vehicle posture information;

[0187] Calculate the vehicle position coordinates and the rotation center point offset according to the vehicle linear velocity and vehicle angular velocity;

[0188] According to the oblique linear velocity and the oblique angular velocity, the oblique moving distance, the oblique composite velocity and the oblique steering angle are calculated.

[0189] In some embodiments, different target selectors are used to calculate different driving parameters. When the target selector is a behavior selector, the behavior selector is used to adaptively adjust speed changes, posture corrections, drive selections, etc. The data required includes the posture information of the cargo, the position and posture information of each robot, and the laser radar's recognition information of obstacles. First, based on the laser radar perception information and map data, the object information perceived by the laser radar can be compared with the map data to instantly identify and classify obstacles in the environment, thereby obtaining the target obstacle position, wherein the target obstacle position can include a static obstacle position and a dynamic obstacle position. Then, based on the position information fused using the extended Kalman filter, including the vehicle position and the vehicle posture, the relative position of the transport vehicle is calculated, and then the cargo position is calculated based on the relative positions of multiple transport vehicles. And combining these two parts of information, the distance between the cargo and the obstacle and the position relationship between the cargo and the obstacle are calculated based on the target obstacle position and the cargo position.

[0190] In some embodiments, before speed selection, a range structure based on a cargo model can be adopted to distinguish what operations should be taken when a dynamic obstacle is at a certain position. The range structure can be defined according to the characteristics of the cargo and transportation requirements. When the target selector is a speed selector, the frame equation set can be calculated based on the cargo length and cargo width. The expression of the frame equation set is as follows: In the formula, l represents the length of the cargo, w represents the width of the cargo; θ represents the angle between the diagonal of the cargo and the horizontal line; L1 represents the range distance to terminate the running activity; L2 represents the range distance for deceleration operation; (x0, y0) represents the coordinates of the dynamic obstacle; L represents the distance between the dynamic obstacle and the cargo; represents the deceleration ratio; vi represents the speed after deceleration; x represents the abscissa, and y represents the ordinate. Among them, the frame equation set contains the formulas of three frames, namely the range contour of the cargo, the frame to terminate the running activity, and the range for deceleration operation. Assume that there is a dynamic obstacle point with coordinates (x0, y0) for the cargo at this time. The distance of this point, that is, the distance from the obstacle point to the cargo, can be deduced according to the frame equation set and the obstacle point coordinates. The calculation formula for the distance from the obstacle point to the cargo is: In the formula, L is the distance from the obstacle point to the cargo. Next, in the speed decision stage, in the speed judge, according to the global path and the local path, it can be judged whether the distance from the obstacle point to the cargo is less than the range distance to terminate the running activity. If the distance from the obstacle point to the cargo is less than the range distance to terminate the running activity, that is, L < L1, a stop instruction is issued to terminate the running activity; if the distance from the obstacle point to the cargo is greater than the range distance to terminate the running activity, that is, L > L1, the cargo does not need to stop and enters the speed selector link to decide whether to perform an operation to reduce the expected speed or keep moving forward at the expected speed. It can be judged whether the distance from the obstacle point to the cargo is less than the deceleration range distance. If the distance from the obstacle point to the cargo is less than the deceleration range distance, that is, L < L2, the speed needs to be reduced. The driving speed can be calculated according to the deceleration ratio The calculation formula for the driving speed is as follows: In the formula, v i is the speed after deceleration, is the deceleration ratio. If the distance from the obstacle point to the cargo is greater than the deceleration range distance, that is, L > L2, the expected speed can be maintained to move forward. It can be understood that when the dynamic obstacle is within the frame to terminate the running activity and the range for deceleration operation, an operation to reduce the expected speed needs to be performed; if the dynamic object is outside the range for deceleration operation, the expected speed can be maintained to move forward.

[0191] In some embodiments, during the posture correction stage, the posture information of the cargo is obtained by attaching sensors to the cargo. At the same time, the posture data of each vehicle is obtained from the vehicle's position information. By comparing the postures of the cargo and the vehicle, it can be determined whether a posture correction operation is required for a single vehicle or multiple vehicles. If the posture of the vehicle matches the posture of the cargo, no additional adjustment is required; if the comparison result between the posture of a vehicle and the cargo exceeds the allowable range of posture deviation, that is, the posture difference is large, then the vehicle needs to perform a separate posture correction step. Then adjust the posture of the vehicle during the movement to keep it consistent with the posture of the cargo. Figure 4 As shown, by comparing and judging the postures of the cargo and the vehicle, the vehicle that needs to be corrected can be determined to ensure the consistency of the posture between the vehicle and the cargo. When the target selector is a posture corrector, the vehicle linear velocity and vehicle angular velocity can be calculated based on the cargo posture information and the vehicle posture information, where it is assumed that the linear velocity is v and the angular velocity is ω in the original speed instruction. This can be obtained by the equation group To calculate, the equations for the vehicle linear velocity and vehicle angular velocity are expressed as: In the formula, v i is the corrected vehicle linear speed, ω i is the corrected vehicle angular velocity. For a vehicle, according to the vehicle linear velocity and vehicle angular velocity, the position of the fix_point sent to the vehicle in the vehicle coordinate system, the vehicle position coordinates can be expressed as: For the rotation center point steer_point, the fix-point can calculate the rotation center point offset in the rotation center point coordinate system according to the vehicle linear velocity and vehicle angular velocity. The expression of the rotation center point offset is: In the mode selection stage, the mode selector can select two different movement modes when turning for the movement mode of the four-wheel four-turn vehicle, namely the oblique movement mode and the steering mode. In the oblique movement mode, the vehicle will turn all wheels in one direction and travel at the same rotation speed, so that the vehicle and the cargo are in an oblique movement state as a whole; in the steering mode, the vehicle will adjust the steering angle of each wheel so that all wheels are radially toward the rotation point, and then determine the rotation speed of each wheel according to the relative position of the rotation point and the wheel, so that the vehicle and the cargo are in a rotational movement state as a whole. In the mode selection stage, according to the specific turning requirements and environmental conditions, the appropriate movement mode can be selected according to the design and control strategy of the vehicle. The steering mode is suitable for situations that require fast turning and high maneuverability, while the oblique movement mode is suitable for scenarios that require precise control. Assume that in the original speed command, the linear velocity is v and the angular velocity is ω. According to the oblique linear velocity and the oblique angular velocity, the oblique movement distance is calculated, where the calculation formula for the oblique movement distance is: In the formula, θ represents the rotation angle of the robot in the time interval Δt; l represents the movement distance of the robot in the tilt mode. The speed in the xAy direction can be deduced, and the tilt composite speed can be calculated according to the tilt linear speed and the tilt angular speed. The calculation formula of the tilt composite speed is:

[0192] In the formula, v i v xi and v yi The composite speed, v xi represents the velocity component of the robot in the x direction; v yi represents the velocity component of the robot in the y direction. Therefore, the oblique steering angle of each wheel can be calculated according to the oblique linear velocity and oblique angular velocity, where the calculation formula of the oblique steering angle is: Where θ is the tilt steering angle.

[0193] In some embodiments, in step S107, the coordinated transportation can be performed according to the desired state of the virtual structure, the target kinematic model and the driving parameters. To approach the expected state X of the virtual structure as much as possible d f. Ensure the overall coordination and stability of the formation. The virtual structure of the formation contains virtual individuals corresponding to each actual individual in the formation. The state of the virtual individual is the expected state of the actual individual. Combined with V2X communication technology, the vehicles in the fleet maintain the geometric relationship of the formation, provide continuous perception between the fleet and its surrounding environment, detect changes and states of the surrounding environment in real time, and provide environmental basis for subsequent operations such as obstacle avoidance. At the same time, the target kinematic model is used to coordinate single-vehicle movement, single-vehicle oblique movement or multiple movements, and the driving parameters are continuously adjusted in real time during transportation to keep the goods stable and the vehicles in a coordinated state, so as to achieve multi-vehicle coordinated transportation.

[0194] In some embodiments, the overall process of multi-vehicle coordinated transportation is as follows: Figure 5 As shown, vehicles can be formed into a platoon first, then a motion model is established, and sensor fusion positioning is performed. SLAM real-time positioning and map building technology is used to build a map, and path planning is performed. Finally, driving parameters are adjusted through behavior selection to achieve multi-vehicle collaborative transportation.

[0195] The beneficial effects of implementing the embodiments of the present invention include: the embodiments of the present invention first calculate the relative position of the vehicle and the center of mass of the structure and the desired state of the virtual structure according to the three-dimensional rectangular coordinate system, and construct a target kinematic model according to the relative position of the vehicle and the center of mass of the structure, and then calculate the vehicle position and vehicle posture according to the vehicle sensor data using the extended Kalman filter algorithm, and construct map data according to the radar data, and then perform path planning according to the preset grid map to obtain the global path and the local path, and calculate the driving parameters through the target selector according to the vehicle position, vehicle posture, map data, the global path and the local path, and finally perform collaborative transportation according to the desired state of the virtual structure, the target kinematic model and the driving parameters, so that collaborative transportation can be performed by controlling the vehicle parameters and the cargo posture, thereby improving the collaborative accuracy, driving stability and transportation efficiency.

[0196] like Figure 6As shown, an embodiment of the present invention also provides a collaborative transport device, including a connecting plate 201, an air spring 202, a connecting shaft 203, a bearing seat 204, a laser radar 205, an inertial sensor 206, an angular velocity sensor 207, a computer host 208, a display 209, a voltage stabilizer 210, and a router 211. Among them, the connecting plate 201 is arranged at the top of the device, attached to the pallet or the payload, and the goods can be carried by the connecting plate. The air spring 202 is connected to the connecting plate, and the air spring can be used as an airbag, the lower end of which is fixed on the rotating shaft of the rotating joint, and the upper end is provided with a valve core and a connecting plate, which can be inflated and deflated by the valve core. At the same time, the central axis of the airbag is aligned with the rotating axis of the rotating joint, and is hinged to the bottom plate through a bearing, and the bottom plate is fixed to the transport device. The connecting shaft 203 is connected to the air spring, and the air spring and the connecting plate can be supported on the bearing seat through the connecting shaft. The bearing seat 204 is connected to the connecting shaft, and the bearing seat can be set on the upper layer of the device. The bearing seat, the connecting plate, the air spring, the connecting shaft and the angular velocity sensor are combined into a rigid-flexible coupling connector to buffer the instantaneous pressure between the cargo and the device. The laser radar 205 is used to monitor the first driving state, which may include position and speed. The laser radar and the aluminum plate can be fixed together by screws, and the opening of the power data line of the laser radar is reserved on the aluminum plate, wherein the aluminum plate is fixed on the aluminum frame according to the European standard, and the aluminum frame has a coupling at the bottom, which is responsible for connecting and fixing the angle sensor. Among them, the aluminum plate is located on the upper layer of the aluminum frame, and the aluminum frame is divided into three layers, including the upper layer, the middle layer and the lower layer. The inertial sensor 206 is used to monitor the second driving state, which may include acceleration, tilt, impact, vibration, rotation and multi-degree-of-freedom motion. The inertial sensor can be fixed at the center of the trolley by screws or pasting, and the inertial sensor can be set on the lower layer of the aluminum frame. The angular velocity sensor 207 is connected to the bearing seat, and the angular velocity sensor can be used to monitor the third driving state, which may include posture and position. Among them, the angle sensor can be set in the rotating joint, and the real-time direction sampling can be performed through the angular velocity sensor for feedback control. The computer host 208 is connected to the acrylic plate, and the multi-vehicle cooperative transportation method of this embodiment can be executed by the computer host according to the first driving state, the second driving state and the third driving state to control the operation of each component. Among them, the acrylic plate can be set in the middle layer of the aluminum frame. The display 209 is connected to the acrylic plate, and the driving parameters and related driving states can be displayed through the display. The voltage stabilizer 210 is used to output a stable voltage. The voltage output by the chassis can be converted into a voltage adapted to each component by the voltage stabilizer to provide a stable voltage for each component. The voltage stabilizer can be set in the lower layer of the aluminum frame. The router 211 is connected to the acrylic plate, and the communication connection between the laser radar, inertial sensor, angular velocity sensor, display, voltage stabilizer and the computer host can be provided through the router.Furthermore, laser radar, inertial sensor, computer host, display, voltage stabilizer, and router can be combined into a real vehicle transport platform. By combining the real vehicle transport platform and the rigid-flexible coupling connector into an unmanned transport platform, multi-vehicle collaborative transportation can be carried out, wherein the frame structure of the unmanned transport platform is as follows. Figure 7 shown.

[0197] The beneficial effects of implementing the embodiments of the present invention include: the embodiments of the present invention provide a collaborative transport device including a connecting plate, an air spring, a connecting shaft, a bearing seat, a laser radar, an inertial sensor, an angular velocity sensor, a computer host, a display, a voltage regulator, and a router. Among them, the connecting plate is arranged at the top, the air spring is connected to the connecting plate, the connecting shaft is connected to the air spring, the bearing seat is connected to the connecting shaft, the aluminum plate is arranged on the upper layer of the aluminum frame, the bearing seat and the laser radar are both arranged on the aluminum frame, the inertial sensor and the voltage regulator are both arranged on the lower layer of the aluminum frame, the angular velocity sensor is connected to the bearing seat, the computer host, the display, and the router are all arranged on the acrylic plate, the acrylic plate is arranged on the middle layer of the aluminum frame, and the computer host is used to execute the multi-vehicle collaborative transport method of this embodiment, so that collaborative transport can be carried out by controlling vehicle parameters and cargo posture, thereby improving collaborative accuracy, driving stability and transportation efficiency. At the same time, the embodiments of the present invention solve the problems of huge labor costs, low economic benefits, and poor adaptability to work scenarios in the transportation of large-scale goods, and improve the efficiency, accuracy and safety of large-scale transportation fleet formation.

[0198] like Figure 8 As shown, the embodiment of the present invention further provides a multi-vehicle cooperative transport system, including a cargo to be transported 301 and at least two cooperative transport devices 302. When the number of cooperative transport devices is 2, the cooperative transport devices can be set at both ends of the cargo to facilitate turning and moving forward. The relative position relationship between the cargo and the cooperative transport device is shown in the top view as follows: Fig. 9 When the number of cooperative transport devices is 4, the cooperative transport devices can be set at the four corners of the cargo, so that the cargo is in the center of the multi-vehicle cooperative transport system to keep the cargo stable.

[0199] The beneficial effects of implementing the embodiments of the present invention include: the embodiments of the present invention provide a multi-vehicle cooperative transport system including goods to be transported and at least two cooperative transport devices, and the goods to be transported can be cooperatively transported by multiple cooperative transport devices, thereby improving the cooperative accuracy, driving stability and transportation efficiency.

[0200] The preferred embodiments of the present application are described above with reference to the accompanying drawings, but the scope of the rights of the present application is not limited thereto. Any modification, equivalent substitution and improvement made by a person skilled in the art without departing from the scope and essence of the present application should be within the scope of the rights of the present application.

Claims

1. A multi-vehicle coordinated transportation method, characterized in that: The following steps are involved: According to the three-dimensional rectangular coordinate system, the relative position of the center of mass of the vehicle and the transport structure and the desired state of the virtual structure are calculated; According to the relative position of the vehicle and the center of mass of the transport structure, a target kinematic model is constructed, wherein the target kinematic model includes a single-vehicle kinematic model, a single-vehicle oblique shift kinematic model and a multi-vehicle kinematic model; According to the vehicle sensor data, the extended Kalman filter algorithm is used to calculate the vehicle position and vehicle posture; Construct map data based on radar data; According to the preset grid map, path planning is performed to obtain the global path and local path; Calculating driving parameters through a target selector according to the vehicle position, the vehicle posture, the map data, the global path and the local path, the target selector includes a behavior selector, a speed selector, and a posture corrector, and the driving parameters include a distance between the cargo and the obstacle, a positional relationship between the cargo and the obstacle, a driving speed, a vehicle position coordinate, a rotation center point offset, a slant moving distance, a slant synthetic speed and a slant steering angle; performing collaborative transportation according to the desired state of the virtual structure, the target kinematic model and the driving parameters; The step of calculating the relative position of the vehicle and the center of mass of the transport structure and the desired state of the virtual structure according to the three-dimensional rectangular coordinate system includes: Constructing a group of centroid position equations according to the three-dimensional rectangular coordinate system; Connect the center of each sub-support area according to the support mode and the center of mass position equation group to obtain a stable support area; According to the stable support area, a set of equations of three side lengths of the supporting triangle is calculated; Calculate the perimeter of the supporting triangle according to the three-side length equation group; Calculate the area of ​​the supporting triangle according to the perimeter; Calculating the support angle based on the area and the centroid height; Calculating the relative position of the vehicle and the mass center of the transport structure based on the plurality of support angles; According to the three-dimensional rectangular coordinate system, a local virtual structure dynamic model is constructed by using a virtual structure method; Constructing a local virtual structure state vector according to the local virtual structure dynamics model; The virtual structure desired state is constructed according to the local virtual structure state vector.

2. The method according to claim 1, characterized in that When the target kinematics model is a single-vehicle kinematics model, constructing the target kinematics model according to the relative positions of the vehicle and the center of mass of the transport structure includes: Constructing a velocity equation for a vehicle with a single drive wheel based on the relative positions of the vehicle and the center of mass of the transport structure; According to the single-drive-wheel bicycle motion speed equation, construct the left front wheel bicycle motion decomposition speed vector, the left rear wheel bicycle motion decomposition speed vector, the right front wheel bicycle motion decomposition speed vector, and the right rear wheel bicycle motion decomposition speed vector; Performing a filling transformation on the left front wheel bicycle motion decomposition velocity vector to obtain a left front wheel bicycle motion equation group; Performing a filling transformation on the left rear wheel bicycle motion decomposition velocity vector to obtain a left rear wheel bicycle motion equation group; Performing a filling transformation on the right front wheel bicycle motion decomposition velocity vector to obtain a right front wheel bicycle motion equation group; Performing a filling transformation on the right rear wheel bicycle motion decomposition velocity vector to obtain a right rear wheel bicycle motion equation group; Constructing a center speed bicycle motion matrix according to the left front wheel bicycle motion equation group, the left rear wheel bicycle motion equation group, the right front wheel bicycle motion equation group and the right rear wheel bicycle motion equation group; The coefficient matrix in the center speed bicycle motion matrix is ​​inverted to obtain the bicycle kinematic model.

3. The method according to claim 1, characterized in that When the target kinematics model is a multi-vehicle kinematics model, constructing the target kinematics model according to the relative positions of the vehicles and the center of mass of the transport structure includes: Calculate the distance from the cargo to the cargo rotation point based on the expected linear velocity and the expected angular velocity; Decomposing the expected linear speed to obtain a group of decomposition equations of the cargo linear speed; Constructing a vehicle linear velocity decomposition equation group according to the distance from the cargo to the cargo rotation point and the expected angular velocity; The multi-vehicle kinematic model is constructed according to the distance from the cargo to the cargo rotation point, the relative position of the vehicle and the center of mass of the transport structure, the cargo linear velocity decomposition equation group and the vehicle linear velocity decomposition equation group.

4. The method according to claim 1, characterized in that: The method of calculating the vehicle position and vehicle posture using an extended Kalman filter algorithm based on the vehicle sensor data includes: Using the vehicle sensor data as input to the extended Kalman filter algorithm, calculating a priori estimates of a state vector and a priori estimated covariance matrix; Calculating a Kalman gain matrix based on the a priori estimated covariance matrix; Calculating a posteriori estimates of state variables according to the prior estimates and the Kalman gain matrix; Calculating a posterior estimation covariance matrix according to the prior estimation covariance matrix and the Kalman gain matrix; The state variable posterior estimation value and the posterior estimation covariance matrix are output processed to obtain the vehicle position and the vehicle posture.

5. The method according to claim 1, characterized in that The step of constructing map data according to radar data includes: According to the local synchronous positioning map, the global synchronous positioning map and the gravity direction, the radar data is projected into a two-dimensional plane to obtain a sub-map; Calculate the position and posture of the radar frame in the sub-image coordinate system according to the sub-image, the rotation angle and the observation value; Calculating a hit ratio according to the positions and poses of the sub-image and the radar frame in the sub-image coordinate system; According to the hit ratio, construct a hit probability update equation and a miss probability update equation; Calculating the matching probability between the scan point and the sub-graph according to the nonlinear least squares equation, the hit probability update equation and the miss probability update equation; According to the matching probability between the scan points and the sub-map, the scan data is matched with the sub-map to obtain the map data.

6. The method according to claim 1, characterized in that The path planning is performed according to the preset grid map to obtain the global path and the local path, including: According to the preset grid map and the delta algorithm, determine whether there is an obstacle on the straight path from the parent node to the neighbor node; If there are no obstacles on the straight path from the parent node to the neighbor node, the grid nodes are smoothly connected to obtain the global path; According to the preset grid map, the local path is obtained by combining the linear speed adjusted by distance and curvature with the speed-based look-ahead using an adjusted pure pursuit algorithm and combining it with a time-scaled collision detector.

7. The method according to claim 1, characterized in that The calculating of the driving parameters by a target selector according to the vehicle position, the vehicle posture, the map data, the global path and the local path comprises: When the target selector is a behavior selector, environmental obstacles are identified according to the laser radar perception information and the map data to obtain the target obstacle position, where the target obstacle position includes a static obstacle position and a dynamic obstacle position; Calculating the relative position of the transport vehicle according to the vehicle position and the vehicle posture; Calculating the location of the cargo based on the relative positions of the plurality of transport vehicles; Calculating the distance between the cargo and the obstacle and the positional relationship between the cargo and the obstacle according to the target obstacle position and the cargo position; When the target selector is a speed selector, the frame line equations are calculated based on the cargo length and cargo width; Calculate the distance from the obstacle point to the cargo based on the frame line equation group and the coordinates of the obstacle point; According to the global path and the local path, determining whether the distance from the obstacle point to the cargo is less than the distance of the terminating operation range; If the distance from the obstacle point to the cargo is less than the distance within the range of termination of operation, the operation is terminated; otherwise, it is determined whether the distance from the obstacle point to the cargo is less than the distance within the range of required deceleration; If the distance from the obstacle to the cargo is less than the range of distance that requires deceleration, the driving speed is calculated according to the deceleration ratio; When the target selector is a posture corrector, the vehicle linear velocity and the vehicle angular velocity are calculated according to the cargo posture information and the vehicle posture information; Calculating the vehicle position coordinates and the rotation center point offset according to the vehicle linear velocity and the vehicle angular velocity; The tilt movement distance, the tilt synthesis speed and the tilt steering angle are calculated based on the tilt linear speed and the tilt angular speed.

8. A coordinated transport device, characterized in that: include: A connecting plate, the connecting plate is used to carry goods; an air spring connected to the connecting plate; A connecting shaft connected to the air spring; A bearing seat connected to the connecting shaft and arranged on an upper layer of the device; A laser radar, wherein the laser radar is used to monitor a first driving state; an inertial sensor, wherein the inertial sensor is used to monitor a second driving state; an angular velocity sensor, the angular velocity sensor being connected to the bearing seat and being used for monitoring a third driving state; A computer host, wherein the computer host is used to execute the multi-vehicle cooperative transportation method according to any one of claims 1 to 7 according to the first driving state, the second driving state and the third driving state; A display, the display being used to display driving parameters; A voltage stabilizer, the voltage stabilizer is used to output a stable voltage; A router is used to provide communication connection between the laser radar, the inertial sensor, the angular velocity sensor, the display, the voltage stabilizer and the computer host.

9. A multi-vehicle cooperative transportation system, characterized in that: include: Goods to be transported and at least two coordinated transport devices as claimed in claim 8.

Images