Background
Currently, an Automatic Guided Vehicle (AGV) has become an important device in intelligent manufacturing, advanced logistics and intelligent factories, and has the purposes of facilitating product transportation, improving production efficiency, reducing production cost and the like. The so-called automatic guidance means that the vehicle travels along a preset path, and currently, the common guidance methods include magnetic navigation, laser navigation, two-dimensional code navigation, visual navigation, and the like. Magnetic navigation is to lay magnetic strips on a preset running path, the existing use environment needs to be changed, and the engineering quantity is not small; laser navigation is susceptible to light changes, is not easy to identify transparent objects, and is relatively expensive; the two-dimension code navigation needs to paste a plurality of two-dimension codes on a preset driving path, so that the appearance of an application environment is influenced, and the two-dimension codes are easily damaged or covered; the visual navigation has lower requirements on hardware and higher requirements on an algorithm level, but the image information acquired based on the visual navigation system has the potential of acquiring or mining richer information, so the AGV navigation control algorithm based on the visual navigation increasingly becomes a research hotspot in academic and engineering circles.
At present, visual navigation is mainly divided into visual navigation based on vehicle-mounted visual equipment and visual navigation based on a wireless network and the visual equipment. The visual navigation based on the vehicle-mounted visual equipment needs complex deep neural network learning and training, and only local environment information and position information can be mastered; the visual navigation based on the wireless network and the visual equipment does not need to transform the field environment, can master the environment information and the position information which are nearly global, is convenient for the global optimization of system design, has low requirements on the sensor and low cost, and is an AGV navigation mode with good application prospect.
Disclosure of Invention
In view of the above, there is a need for an automatic navigation cloud server for controlling the travel of a transport vehicle.
An automatic navigation cloud server is applied to a scene and comprises a transport vehicle, a destination, a camera device and a cloud server, wherein the cloud server plans a reference path for the transport vehicle to travel to the destination, the camera device shoots an image of the transport vehicle, the cloud server analyzes current pose information of the transport vehicle according to the image and gives ideal pose information of a forward reference point of the transport vehicle according to the reference path, and calculates a travel control instruction of the transport vehicle according to the current pose information of the transport vehicle and the ideal pose information of the forward reference point, and the cloud server transmits the travel control instruction to the transport vehicle so as to control the transport vehicle to travel to the destination according to the travel control instruction.
An automatic navigation control method is applied to a scene, and comprises the following steps:
planning a reference path of the transport vehicle to the destination;
shooting images of the transport vehicle in the scene;
analyzing the current pose information of the transport vehicle according to the image and giving ideal pose information of a forward reference point of the transport vehicle according to the reference path;
calculating a running control instruction of the transport vehicle according to the current pose information of the transport vehicle and the ideal pose information of the forward reference point;
and controlling the transport vehicle to run to the destination according to the running control instruction.
The automatic navigation cloud server calculates the running control instruction, so that the transport vehicle can quickly return to the reference path when running deviation occurs, efficient deviation correction is realized, and the running accuracy and efficiency of the transport vehicle are greatly improved.
Detailed Description
Referring to fig. 1, an automatic navigation cloud server 100 according to a preferred embodiment of the present invention includes a transportation vehicle 10, a destination 20, a camera 30, and a cloud server 40. The transporter 10, the camera device 30 and the cloud server 40 are in communication connection through a wireless network so as to transmit data and commands to each other. The automatic navigation cloud server 100 is arranged in a scene, such as a storage bin.
Referring to fig. 2, a scene 500 is provided with the destination 20 and the transportation vehicle 10, and a camera device 30 is further disposed for capturing images in the scene 500, where the images include the transportation vehicle 10. The cloud server 40 extracts the current pose information of the transporter 10 through a visual positioning algorithm. The cloud server 40 also plans a reference path for the transportation vehicle 10 to travel to the destination 20, and calculates pose information of a forward reference point. In the present embodiment, the reference path is a linear path composed of a plurality of points in order, and the forward reference point is a reference point at which the carriage 10 travels forward in the reference path. The cloud server 40 calculates a control instruction for the traveling of the transporter 10 according to the current pose information of the transporter 10 and the pose information of the forward reference point. The transporter 10 travels toward the destination 20 under the guidance of the cloud server 40.
The transport Vehicle 10 is an Automated Guided Vehicle (AGV) and is controlled by the cloud server 40 to move within the scene. In this embodiment, the transportation vehicle 10 is controlled by the cloud server 40 to drive toward the destination 20 in the scene. The transporter 10 includes a first wireless communication module 12 for establishing a wireless communication connection with the cloud server 40 for transmitting data and commands.
The destination 20 is fixedly arranged at a position within the scene. In this embodiment, the destination 20 may be a rack for carrying cargo, and the transporter 10 is driven toward the destination 20 to perform loading and unloading interactions.
The camera device 30 is fixedly disposed on a ceiling in the scene, and is configured to capture images in the scene, including capturing the movement of the transporter 10 in the scene. The image information captured by the imaging device 30 is transmitted to the cloud server 40.
The cloud server 40 is configured to collect image information of the camera device 30, analyze the movement of the transporter 10, and control the movement of the transporter 10. The cloud server 40 analyzes the position and posture of the transporter 10 and its surrounding environmental conditions by analyzing the movement of the transporter 10. It is understood that the cloud server 40 further includes a plurality of 4G/5G base stations disposed within or near the scene for communicating instructions and information to the transporter 10.
The cloud server 40 includes a path planning module 42, a navigation module 44, a transporter control module 46, an analysis module 48, and a second wireless communication module 49. The second wireless communication module 49 establishes a wireless communication connection with the first wireless communication module 12 of the transporter 10 to transmit data and commands. The path planning module 42 is configured to plan the reference path traveled by the vehicle 10 toward the destination 20. The navigation module 44 is configured to provide navigation information to the transporter 10 according to the reference path. The transporter control module 46 is configured to control the transporter 10 to travel to the destination 20 according to the navigation information. The analysis module 48 receives the image information transmitted by the camera device 30 and analyzes the current pose information of the transporter 10. Specifically, the analysis module 48 extracts characteristic information of the transporter 10 from the image information. In this embodiment, the characteristic information of the transporter 10 is that the analysis module 48 extracts the current pose information (x, y, θ) of the transporter 10 based on a visual positioning algorithm, where the parameters x and y are the position information of the transporter 10 in the scene global coordinate system, and θ is the heading angle information of the transporter 10.
After the path planning module 42 formulates the reference path, the analysis module 48 analyzes the current position information (x, y, θ) of the transporter 10 and the ideal pose information (x) of the selected forward reference pointr,yr,θr) Wherein the parameter xrAnd yrPosition information of the forward reference point in the scene global coordinate system, θrAnd the ideal course angle information of the forward reference point. Based on the current pose information (x, y, theta) of the transporter 10 and the ideal pose information (x) of the forward reference pointr,yr,θr) The analysis module 48 calculates three deviation amounts including: an angular deviation amount Δ θ, a distance deviation amount Δ d, and another angular deviation amount Δ φ. The angular deviation Δ θ is defined as an angle between the heading angle of the transporter 10 and the ideal heading angle of the forward reference point, i.e., Δ θ - θrRepresenting the need for delta theta angle adjustment of the vehicle 10. The distance deviation Δ d is defined as the minimum distance from the current position of the vehicle 10 to the reference path, i.e., Δ d min { norm [ (X, y) - (X) }i,Yi)],i∈[1,N]Norm represents the Euclidean distance, N represents the number of points quantized on the reference path, and the smaller N is, the larger the distance between two points is, and the lower the resolution is; said (X)i,Yi) Is the position parameter of the ith point constituting the reference path. The further angular deviation Δ φ is defined as a heading angle of the transporter 10 and the destination 20 and the transportThe angle between the line connecting the current positions of the truck 10, i.e., Δ φ - θ. And calculating a running control command (v, omega) of the transport vehicle 10 based on a PID (proportion, integral and differential) algorithm according to the three deviation values, wherein the parameter v is the linear speed of the transport vehicle 10, and the parameter omega is the angular speed of the transport vehicle 10. Wherein the computational mathematical model is represented as
ω=PID(Δθ,Δd,Δφ)=PID1(Δθ)+PID2(Δd)+PID3(Δφ)
Wherein the content of the first and second substances,
wherein, Ki,P,i∈[1,3]Is a proportional element gain coefficient, Ki,I,i∈[1,3]Is an integral unit gain coefficient, Ki,D,i∈[1,3]Is the differential cell gain factor. The analysis module 48 transmits the travel control command (v, ω) to the transporting vehicle 10 through the second wireless communication module 49 to control the traveling of the transporting vehicle 10.
The automatic navigation cloud server 100 calculates the driving control command (v, ω) so that the transport vehicle 10 can quickly return to the reference path during driving deviation, thereby realizing efficient deviation correction and greatly improving the driving accuracy and efficiency of the transport vehicle 10.
In view of the above, although the preferred embodiments of the present invention have been disclosed for illustrative purposes, the present invention is not limited to the above-described embodiments, and those skilled in the relevant art can make various modifications and applications without departing from the scope of the basic technical idea of the present invention.