CN112767540A - Automatic loading method for AGV - Google Patents
Automatic loading method for AGV Download PDFInfo
- Publication number
- CN112767540A CN112767540A CN202110055012.4A CN202110055012A CN112767540A CN 112767540 A CN112767540 A CN 112767540A CN 202110055012 A CN202110055012 A CN 202110055012A CN 112767540 A CN112767540 A CN 112767540A
- Authority
- CN
- China
- Prior art keywords
- goods
- agv
- carriage
- information
- stacking
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Pending
Links
- 238000011068 loading method Methods 0.000 title claims abstract description 44
- 238000001514 detection method Methods 0.000 claims abstract description 47
- 238000000034 method Methods 0.000 claims description 17
- 238000013135 deep learning Methods 0.000 claims description 15
- 238000012549 training Methods 0.000 claims description 11
- 238000009434 installation Methods 0.000 claims description 9
- 230000000007 visual effect Effects 0.000 claims description 9
- 238000004806 packaging method and process Methods 0.000 claims description 6
- 238000000513 principal component analysis Methods 0.000 claims description 6
- 230000032258 transport Effects 0.000 description 8
- 238000005516 engineering process Methods 0.000 description 5
- 238000004364 calculation method Methods 0.000 description 1
- 238000004891 communication Methods 0.000 description 1
- 230000003247 decreasing effect Effects 0.000 description 1
- 238000010586 diagram Methods 0.000 description 1
- 230000006870 function Effects 0.000 description 1
- 230000006872 improvement Effects 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000003287 optical effect Effects 0.000 description 1
- 238000012856 packing Methods 0.000 description 1
- 230000008569 process Effects 0.000 description 1
- 238000012545 processing Methods 0.000 description 1
Images
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T17/00—Three dimensional [3D] modelling, e.g. data description of 3D objects
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01B—MEASURING LENGTH, THICKNESS OR SIMILAR LINEAR DIMENSIONS; MEASURING ANGLES; MEASURING AREAS; MEASURING IRREGULARITIES OF SURFACES OR CONTOURS
- G01B11/00—Measuring arrangements characterised by the use of optical techniques
- G01B11/24—Measuring arrangements characterised by the use of optical techniques for measuring contours or curvatures
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01S—RADIO DIRECTION-FINDING; RADIO NAVIGATION; DETERMINING DISTANCE OR VELOCITY BY USE OF RADIO WAVES; LOCATING OR PRESENCE-DETECTING BY USE OF THE REFLECTION OR RERADIATION OF RADIO WAVES; ANALOGOUS ARRANGEMENTS USING OTHER WAVES
- G01S17/00—Systems using the reflection or reradiation of electromagnetic waves other than radio waves, e.g. lidar systems
- G01S17/88—Lidar systems specially adapted for specific applications
- G01S17/89—Lidar systems specially adapted for specific applications for mapping or imaging
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
- G06Q10/00—Administration; Management
- G06Q10/06—Resources, workflows, human or project management; Enterprise or organisation planning; Enterprise or organisation modelling
- G06Q10/063—Operations research, analysis or management
- G06Q10/0631—Resource planning, allocation, distributing or scheduling for enterprises or organisations
- G06Q10/06312—Adjustment or analysis of established resource schedule, e.g. resource or task levelling, or dynamic rescheduling
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
- G06Q10/00—Administration; Management
- G06Q10/08—Logistics, e.g. warehousing, loading or distribution; Inventory or stock management
- G06Q10/083—Shipping
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V20/00—Scenes; Scene-specific elements
- G06V20/10—Terrestrial scenes
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T2200/00—Indexing scheme for image data processing or generation, in general
- G06T2200/08—Indexing scheme for image data processing or generation, in general involving all processing steps from image acquisition to 3D model generation
Landscapes
- Engineering & Computer Science (AREA)
- Business, Economics & Management (AREA)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Human Resources & Organizations (AREA)
- Economics (AREA)
- Theoretical Computer Science (AREA)
- Strategic Management (AREA)
- Entrepreneurship & Innovation (AREA)
- Quality & Reliability (AREA)
- General Business, Economics & Management (AREA)
- Tourism & Hospitality (AREA)
- Development Economics (AREA)
- Marketing (AREA)
- Operations Research (AREA)
- Multimedia (AREA)
- Game Theory and Decision Science (AREA)
- Educational Administration (AREA)
- Computer Graphics (AREA)
- Geometry (AREA)
- Software Systems (AREA)
- Computer Networks & Wireless Communication (AREA)
- Electromagnetism (AREA)
- Radar, Positioning & Navigation (AREA)
- Remote Sensing (AREA)
- Traffic Control Systems (AREA)
Abstract
The invention discloses an automatic loading method of an AGV, which comprises the following steps: the AGVS scheduling system schedules the AGV provided with the scanning equipment to surround the truck carriage for a circle, issues an instruction to the three-dimensional scene reconstruction module, and the three-dimensional scene reconstruction module reconstructs a three-dimensional scene of the carriage; after the three-dimensional scene graph is collected, the three-dimensional scene reconstruction module transmits the point cloud data to the target object type detection module; the target object type detection module identifies the types of goods in the carriage, judges whether stacking is possible or not, calculates stacking information, and packages and sends the stacking information to the task management system WMS; and after receiving the stacking information, the task management system WMS allocates the cargo carrying or loading and unloading task to the stacking type AGV. Not only reduced the work load of manual handling, solved the goods moreover and stacked the problem that the kind is confused, can acquire simultaneously that the goods information carries out unified management, realized automatic transport moreover, improved handling efficiency.
Description
Technical Field
The invention belongs to the field of AGV automation, and relates to an AGV automatic loading method.
Background
Automatic loading transport is the last process that AGV is automatic, and when present AGV transports goods, need know the world coordinate system that the goods was located in advance or need the corresponding target point coordinate of goods transport, but because the freight train parking position is inaccurate, the goods species is unclear in the packing box, the position that the goods was laid in the carriage is unset etc. reason, so present pier container loading still adopts and sends the goods to the container by artifical traditional fork truck of driving in, but often this kind of transport mode can be because of artificial subjective factor, the goods of stacking is obscure easily, hardly with goods information entry system unified management, and adopt artifical fork truck transport hardly to realize 24 hours uninterrupted operation, the enterprise is difficult to realize full-automatic flow.
Disclosure of Invention
The invention aims to: the method for automatically loading the AGV based on the deep learning technology is provided.
The technical scheme of the invention is as follows: an automatic loading method of an AGV is applied to a system comprising a control computing device, the AGV provided with a scanning device, a vision device and a stacking type AGV, wherein an AGVS scheduling system, a three-dimensional scene reconstruction module, a target object type detection module and a task management system WMS are integrated in the control computing device, and the automatic loading method of the AGV comprises the following steps:
step 1, dispatching the AGV provided with the scanning equipment to surround a truck carriage for a circle through the AGVS dispatching system, issuing an instruction to the three-dimensional scene reconstruction module, and triggering the three-dimensional scene reconstruction module to reconstruct a three-dimensional scene of the carriage;
and 4, after the task management system WMS receives the stacking information, distributing the goods carrying or loading and unloading task to the stacking type AGV, and assigning the stacking type AGV to complete the corresponding goods carrying or loading and unloading task.
Through including control computing equipment, install scanning device's AGV, vision equipment, stacking type AGV's system models the carriage scene, combine target object type detection module to discern the goods kind in the carriage, and determine the stack information of goods, accomplish cargo handling or loading and unloading task by task management system WMS dispatch stacking type AGV, not only reduced the work load of artifical loading and unloading, and solved the problem that the goods stacked the kind and obscure, can acquire cargo information simultaneously and carry out unified management, and automatic handling has been realized, and the handling efficiency is improved.
The further technical scheme is as follows: the scanning device comprises at least a laser sensor or a vision sensor.
The further technical scheme is as follows: the laser sensor comprises a single-line laser radar, and the installation direction of the single-line laser radar is a vertical direction;
the step 1 comprises the following steps:
the AGVS scheduling system schedules the AGV provided with the scanning equipment to control the single-line laser radar to scan the vertical surface of the vehicle body space, and the AGV surrounds the truck carriage for one circle to complete the scanning of the whole environment of the carriage;
and converting the collected three-dimensional point cloud of the carriage into point cloud information of a world coordinate system by combining the real-time position and posture coordinate of the AGV and the installation position and posture of the single-line laser radar relative to the vehicle body, and reconstructing a three-dimensional scene of the carriage.
Through the vertical installation of the single line laser radar, the AGV drives the single line laser radar to surround the carriage for a circle, data of all environments of the carriage can be collected, and point cloud information of the carriage in a world coordinate system can be obtained by combining the position and posture coordinates of the AGV and the installation position and posture of the single line laser radar, so that the three-dimensional scene of the carriage can be reconstructed conveniently.
The further technical scheme is as follows: the vision equipment comprises at least one camera, the camera is fixedly installed outside the compartment, the horizontal visual field range and the vertical visual field range of the camera are determined according to the vehicle specification, and the scanning range of the camera completely covers the whole compartment;
the step 3 comprises the following steps:
training a training model of goods to be loaded and unloaded by using a yolov3 target detection deep learning method through the target object type detection module;
for the trained model, dividing the image scanned by the camera into S × S grids by the yolov3 target detection deep learning method, detecting cargos in each grid area, and determining the type of the cargos;
outputting the relative pixel difference between the central pixel point of each identified cargo and the focal point of the camera by using a bounding box;
calculating the actual relative coordinates of the goods relative to the central point of the camera by utilizing the length, width and height information of the recognized goods, and calculating the world coordinates of each goods in the carriage according to the coordinates of the camera under a world coordinate system;
according to the point cloud data collected by the three-dimensional scene reconstruction module, carrying out cargo position identification based on a point cloud segmentation-clustering-model growth model;
obtaining three main directions of point cloud by using a Principal Component Analysis (PCA) method, obtaining a centroid coordinate and a covariance of the goods, corresponding to the calculated world coordinate of the goods, and corresponding the type and the coordinate of the carried goods;
after the goods types in the carriage are identified, the information of the identified goods types is called from a goods type database, and whether stacking is available is judged; the goods category database is used for storing the volume of various goods and information on whether the goods can be stacked or not;
and calculating available free area in the carriage according to the cargo coordinates, and uniformly packaging and sending coordinate information, free area information and information whether stacking is possible to be performed or not to the task management system WMS.
Through the deep learning technology, the goods kind is discerned to the goods picture that the camera obtained, confirms the position and the stack information of goods to realize the automatic transport to the goods.
The further technical scheme is as follows: the training model for training the goods to be loaded and unloaded by the target object type detection module by using the yolov3 target detection deep learning method comprises the following steps:
and when the newly added goods are of the type, the newly added goods are trained again through the target object type detection module.
In order to ensure the accuracy of goods loading and unloading, newly added goods are trained in time, the accuracy of a deep learning technology in recognition is ensured, and no missing goods types exist.
The further technical scheme is as follows: the detecting the goods in each grid area comprises the following steps:
and determining the grids with the coordinates of the center position of the goods in the grid area as target grids, and detecting the corresponding goods through the target grids.
The identified image area is divided into a plurality of lattices, and the cargos are detected through the target lattices at the center of the cargos in the lattice area, so that the repeated detection of the cargos is avoided, the detection area is reduced, and the detection efficiency is improved.
The further technical scheme is as follows: and the central pixel point of the identified goods is the coordinate of the central position of the bounding box of the object obtained by target grid prediction.
The further technical scheme is as follows: the step 4 comprises the following steps:
after the task management system WMS receives information whether the goods can be stacked and the position and posture coordinates of the goods which can be stacked in the carriage and are sent by the target object type detection module, the information of the goods which need to be loaded and unloaded is sent to the AGVS scheduling system according to the position information of the corresponding carriage;
after the AGVS scheduling system receives the corresponding carriage position information sent by the task management system WMS, the AGVS scheduling system plans a corresponding route and sends the task to the idle stacking type AGV, and the idle stacking type AGV is allocated to complete the corresponding automatic loading and unloading task.
And the AGVS scheduling system allocates the idle stacking AGV to complete the loading and unloading task according to the acquired storage position information of the carriage, so that unified scheduling is realized, the manual workload is reduced, and the stacking accuracy is ensured.
The invention has the advantages that:
through including control computing equipment, install scanning device's AGV, vision equipment, stacking type AGV's system models the carriage scene, combine target object type detection module to discern the goods kind in the carriage, and determine the stack information of goods, accomplish cargo handling or loading and unloading task by task management system WMS dispatch stacking type AGV, not only reduced the work load of artifical loading and unloading, and solved the problem that the goods stacked the kind and obscure, can acquire cargo information simultaneously and carry out unified management, and automatic handling has been realized, and the handling efficiency is improved.
Drawings
The invention is further described with reference to the following figures and examples:
FIG. 1 is a block diagram of a system architecture for implementing an AGV automatic loading method according to the present disclosure;
FIG. 2 is a flow chart of an AGV automatic loading method provided by the present application;
FIG. 3 is a schematic illustration of a truck park as provided herein;
fig. 4 is a schematic view of a mounting position of a camera provided in the present application.
Detailed Description
Example (b): the application provides an AGV automatic loading method, which is applied to a system comprising a control computing device, an AGV provided with a scanning device, a visual device and a stacking type AGV shown in figure 1, wherein an AGVS scheduling system, a three-dimensional scene reconstruction module, a target object type detection module and a task management system WMS are integrated in the control computing device.
The control computing device is a device with data processing and signal transceiving functions and can be a computer.
Optionally, the scanning device, the AGV with the scanning device installed, the vision device, and the stack AGV may be in wired or wireless communication with the control computing device.
Optionally, the scanning device comprises at least a laser sensor or a vision sensor.
Optionally, the laser sensor includes a single line laser radar, and the installation direction of the single line laser radar is a vertical direction.
Optionally, the vision device includes at least one camera, the camera is fixedly installed outside the vehicle cabin, the horizontal visual field range and the vertical visual field range of the camera are determined according to the vehicle specification, and the scanning range of the camera completely covers the whole vehicle cabin. Alternatively, the camera is preferably an RGB camera.
The dispatching system dispatches the AGV provided with the single-line laser head to surround the truck carriage for a circle, and issues the instruction to the three-dimensional scene reconstruction module, and the three-dimensional scene reconstruction module reconstructs a carriage three-dimensional scene; after the three-dimensional scene graph is collected, the AGVS scheduling system sends a collection stopping instruction to the three-dimensional scene reconstruction module, then the three-dimensional scene reconstruction module transmits the point cloud data cluster to the target object type detection module, the target object type detection module identifies the types of goods in the carriage, judges whether the goods can be stacked or not, calculates the coordinates of the goods and the available free area in the carriage, and uniformly packages and sends the information to the task management system WMS; after receiving the carriage goods information, the task management system WMS allocates goods carrying or loading and unloading tasks to the stacking type AGV, and assigns the tasks to complete corresponding tasks.
As shown in FIG. 2, the AGV automatic loading method includes the following steps.
Step 1, dispatching the AGV provided with the scanning equipment to surround the truck carriage for a circle through an AGVS dispatching system, issuing an instruction to a three-dimensional scene reconstruction module, and triggering the three-dimensional scene reconstruction module to reconstruct a three-dimensional scene of the carriage.
For an AGV equipped with a single line lidar, step 1 includes the following steps.
The method comprises the steps that firstly, an AGVS dispatching system is used for dispatching, an AGV provided with scanning equipment controls a single-line laser radar to scan a vertical surface of a vehicle body space, and the AGV surrounds a truck carriage for a circle to complete scanning of all environments of the carriage.
And converting the collected three-dimensional point cloud of the carriage into point cloud information of a world coordinate system by combining the real-time position coordinates of the AGV and the installation position of the single-line laser radar relative to the vehicle body, and reconstructing a three-dimensional scene of the carriage.
And 2, after the three-dimensional scene graph is acquired, issuing a collection stopping instruction to the three-dimensional scene reconstruction module through the AGVS scheduling system, and transmitting the point cloud data to the target object type detection module through the three-dimensional scene reconstruction module.
And 3, identifying the types of goods in the carriage through the target object type detection module, judging whether stacking is possible or not, calculating stacking information, and packaging and sending the stacking information to the task management system WMS through the target object type detection module.
The stacking information at least comprises the calculated coordinates of the goods, the available free area in the carriage and information whether the goods can be stacked.
In the first step, a training model of goods to be loaded and unloaded is trained by a target object type detection module by using a yolov3 target detection deep learning method.
The method comprises the steps of collecting sample information of goods in advance for marking, comparing the identified goods with the samples collected in advance when the goods are identified by the yolov3 target detection deep learning method, and matching when the similarity reaches the preset similarity.
In practical application, when a newly added cargo type exists, the newly added cargo is trained again through the target object type detection module.
And secondly, dividing the image scanned by the camera into S x S grids by a yolov3 target detection deep learning method for the trained model, detecting cargos in each grid area, and determining the type of the cargos.
Wherein, detect goods in each check district, include: and determining the grids of the coordinate with the goods center position in the grid area as target grids, and detecting the corresponding goods through the target grids.
And thirdly, outputting the relative pixel difference between the central pixel point of each identified cargo and the focal point of the camera by using a bounding box.
The central pixel point of the identified goods is the coordinate of the central position of the bounding box of the object obtained by target grid prediction.
And fourthly, calculating the actual relative coordinates of the cargos relative to the central point of the camera by utilizing the length, width and height information of the recognized cargos, and calculating the world coordinates of each cargo in the carriage according to the coordinates of the camera in a world coordinate system.
Because the scanned picture of the camera is identified to be two-dimensional information, the precision of the calculated world coordinate and the actual coordinate is poor, and the specific coordinate of the object to be transported in the vehicle is accurate, the cargo position coordinate calculation method is added in the method, and the steps from the fifth step to the sixth step are specifically realized.
And fifthly, identifying the cargo position based on a point cloud segmentation-clustering-model growth model according to the point cloud data collected by the three-dimensional scene reconstruction module.
And sixthly, obtaining three main directions of the point cloud by using a Principal Component Analysis (PCA) method, obtaining the centroid coordinates and the covariance of the goods, corresponding to the calculated world coordinates of the goods, and corresponding the types and the coordinates of the carried goods.
And seventhly, after the goods types in the carriage are identified, the information of the identified goods types is called from the goods type database, and whether stacking is available or not is judged.
The cargo type in the carriage can be identified through the trained model.
The goods category database is used for storing the volume of various goods and information whether the goods can be stacked or not.
And eighthly, calculating available free area in the carriage according to the goods coordinates, and uniformly packaging and sending the coordinate information, the free area information and the information whether stacking is possible or not to the task management system WMS.
It should be noted that the target object type detection module needs to provide as many cargo samples as possible for the deep learning training model, the more training samples are, the higher the accuracy of identifying the cargo is, the more accurate the size of the cargo encircled by the bounding box is, and when there is a new cargo type, the target object type detection module needs to be informed in advance to retrain the new cargo.
And 4, after the task management system WMS receives the stacking information, distributing the goods carrying or loading and unloading task to the stacking type AGV, and assigning the stacking type AGV to complete the corresponding goods carrying or loading and unloading task.
Step 4 comprises the following steps:
firstly, after receiving information whether goods can be stacked, coordinates of the goods and position and posture coordinates of the stackable goods in a carriage sent by a target object type detection module, a task management system WMS sends information of the goods to be loaded and unloaded to an AGVS scheduling system according to the position information of the corresponding carriage;
and step two, after the AGVS scheduling system receives the corresponding carriage position information sent by the task management system WMS, the AGVS scheduling system plans a corresponding route and sends the task to the idle stacking type AGV, and the idle stacking type AGV is allocated to complete a corresponding automatic loading and unloading task.
Illustratively, with reference to fig. 3, in practical applications, a large truck with a container is parked in a frame area, a parking positioning block is arranged at the rear wheel position of the truck parking space, the truck needs to be backed into a specified position along an aligned central line as much as possible, the error is controlled to be 2cm, the error is reflected between the rear wheel and the block, the left-right deviation during parking needs to be controlled to be 0.5m, after the truck is parked, the compartment wings at two sides of the container are opened, and the truck is accurately loaded and unloaded by manually operating the RFID notification management system. And then the AGVS scheduling system starts to schedule the AGV provided with the scanning equipment to acquire the carriage information and perform subsequent modeling.
Referring to fig. 4 in combination, cameras are respectively placed on two sides of a truck parking space, a shooting range of each camera covers a complete carriage, for example, 4 cameras are shown in fig. 4, two cameras are respectively placed on two sides of each parking space to acquire pictures, and in practical application, the number of the cameras can be increased or decreased according to the specification of a truck, and the layout of the cameras is adjusted.
To sum up, the AGV automatic loading method that this application provided, through including control computing equipment, install scanning device's AGV, visual equipment, stack type AGV's system models the carriage scene, combine target object type detection module to discern the goods kind in the carriage, and determine the stack information of goods, accomplish cargo handling or the task of loading and unloading by task management system WMS dispatch stack type AGV, not only reduced the work load of manual loading and unloading, and solved the problem that the goods stacked kind obscures, can acquire cargo information simultaneously and carry out unified management, and realized automatic transport, transport efficiency has been improved.
In addition, the single-line laser radar is vertically installed, the single-line laser radar is driven by the AGV to surround the carriage for one circle, data of all environments of the carriage can be collected, and point cloud information of the carriage in a world coordinate system can be obtained by combining the position and posture coordinates of the AGV and the installation position and posture of the single-line laser radar, so that the three-dimensional scene of the carriage can be reconstructed conveniently.
In addition, through the deep learning technology, the goods type is identified from the goods picture obtained by the camera, and the position and stacking information of the goods are determined, so that the automatic carrying of the goods is realized.
In addition, in order to ensure the accuracy of goods loading and unloading, newly added goods are trained in time, the accuracy of the deep learning technology in recognition is ensured, and no missing goods types exist.
In addition, the identified image area is divided into a plurality of grids, and the goods are detected through the target grids at the center of the goods in the grid area, so that repeated detection of the goods is avoided, the detection area is reduced, and the detection efficiency is improved.
In addition, the AGVS scheduling system allocates the idle stacking type AGV to complete loading and unloading tasks according to the acquired storage position information of the carriage, so that uniform scheduling is realized, the manual workload is reduced, and the stacking accuracy is ensured.
The terms "first" and "second" are used for descriptive purposes only and are not to be construed as indicating or implying relative importance or implying a number of the indicated technical features. Thus, a defined feature of "first", "second", may explicitly or implicitly include one or more of that feature. In the description of the present application, "a plurality" means two or more unless otherwise specified.
The above-mentioned serial numbers of the embodiments of the present application are merely for description and do not represent the merits of the embodiments.
It will be understood by those skilled in the art that all or part of the steps for implementing the above embodiments may be implemented by hardware, or may be implemented by a program instructing relevant hardware, where the program may be stored in a computer-readable storage medium, and the above-mentioned storage medium may be a read-only memory, a magnetic disk, an optical disk, or the like.
The above description is only exemplary of the present application and should not be taken as limiting the present application, as any modification, equivalent replacement, or improvement made within the spirit and principle of the present application should be included in the protection scope of the present application.
Claims (8)
1. The automatic AGV loading method is characterized by being applied to a system comprising a control computing device, an AGV with a scanning device, a visual device and a stacking type AGV, wherein an AGVS scheduling system, a three-dimensional scene reconstruction module, a target object type detection module and a task management system WMS are integrated in the control computing device, and the automatic AGV loading method comprises the following steps:
step 1, dispatching the AGV provided with the scanning equipment to surround a truck carriage for a circle through the AGVS dispatching system, issuing an instruction to the three-dimensional scene reconstruction module, and triggering the three-dimensional scene reconstruction module to reconstruct a three-dimensional scene of the carriage;
step 2, after the three-dimensional scene graph is acquired, issuing a collection stopping instruction to the three-dimensional scene reconstruction module through the AGVS scheduling system, and transmitting point cloud data to the target object type detection module through the three-dimensional scene reconstruction module;
step 3, identifying the cargo type in the carriage through the target object type detection module, judging whether stacking is possible, calculating stacking information, wherein the stacking information at least comprises calculated cargo coordinates, available free area in the carriage and information whether the cargo is stackable, and packaging and sending the stacking information to the task management system WMS through the target object type detection module;
and 4, after the task management system WMS receives the stacking information, distributing the goods carrying or loading and unloading task to the stacking type AGV, and assigning the stacking type AGV to complete the corresponding goods carrying or loading and unloading task.
2. The AGV auto-loading method according to claim 1, wherein said scanning device includes at least a laser sensor or a vision sensor.
3. The AGV automatic loading method according to claim 2, wherein the laser sensor includes a single line laser radar, and an installation direction of the single line laser radar is a vertical direction;
the step 1 comprises the following steps:
the AGVS scheduling system schedules the AGV provided with the scanning equipment to control the single-line laser radar to scan the vertical surface of the vehicle body space, and the AGV surrounds the truck carriage for one circle to complete the scanning of the whole environment of the carriage;
and converting the collected three-dimensional point cloud of the carriage into point cloud information of a world coordinate system by combining the real-time position and posture coordinate of the AGV and the installation position and posture of the single-line laser radar relative to the vehicle body, and reconstructing a three-dimensional scene of the carriage.
4. The AGV automatic loading method according to claim 3, wherein the vision device includes at least one camera fixedly installed outside the car, a horizontal visual field range and a vertical visual field range of the camera are determined according to the specifications of the vehicle, and a scanning range of the camera completely covers the whole car;
the step 3 comprises the following steps:
training a training model of goods to be loaded and unloaded by using a yolov3 target detection deep learning method through the target object type detection module;
for the trained model, dividing the image scanned by the camera into S × S grids by the yolov3 target detection deep learning method, detecting cargos in each grid area, and determining the type of the cargos;
outputting the relative pixel difference between the central pixel point of each identified cargo and the focal point of the camera by using a bounding box;
calculating the actual relative coordinates of the goods relative to the central point of the camera by utilizing the length, width and height information of the recognized goods, and calculating the world coordinates of each goods in the carriage according to the coordinates of the camera under a world coordinate system;
according to the point cloud data collected by the three-dimensional scene reconstruction module, carrying out cargo position identification based on a point cloud segmentation-clustering-model growth model;
obtaining three main directions of point cloud by using a Principal Component Analysis (PCA) method, obtaining a centroid coordinate and a covariance of the goods, corresponding to the calculated world coordinate of the goods, and corresponding the type and the coordinate of the carried goods;
after the goods types in the carriage are identified, the information of the identified goods types is called from a goods type database, and whether stacking is available is judged; the goods category database is used for storing the volume of various goods and information on whether the goods can be stacked or not;
and calculating available free area in the carriage according to the cargo coordinates, and uniformly packaging and sending coordinate information, free area information and information whether stacking is possible to be performed or not to the task management system WMS.
5. The AGV automatic loading method according to claim 4, wherein said training model for training the cargo to be loaded and unloaded by said target object type detection module using yolov3 target detection deep learning method comprises:
and when the newly added goods are of the type, the newly added goods are trained again through the target object type detection module.
6. The AGV automatic loading method of claim 4, wherein said detecting the load in each grid section comprises:
and determining the grids with the coordinates of the center position of the goods in the grid area as target grids, and detecting the corresponding goods through the target grids.
7. The AGV automatic loading method according to claim 6, wherein the center pixel point of the recognized load is a coordinate indicating a center position of a bounding box of the object predicted by the target grid.
8. An AGV automatic loading method according to any one of claims 4 to 7, wherein said step 4 includes:
after the task management system WMS receives information whether the goods can be stacked and the position and posture coordinates of the goods which can be stacked in the carriage and are sent by the target object type detection module, the information of the goods which need to be loaded and unloaded is sent to the AGVS scheduling system according to the position information of the corresponding carriage;
after the AGVS scheduling system receives the corresponding carriage position information sent by the task management system WMS, the AGVS scheduling system plans a corresponding route and sends the task to the idle stacking type AGV, and the idle stacking type AGV is allocated to complete the corresponding automatic loading and unloading task.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN202110055012.4A CN112767540A (en) | 2021-01-15 | 2021-01-15 | Automatic loading method for AGV |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN202110055012.4A CN112767540A (en) | 2021-01-15 | 2021-01-15 | Automatic loading method for AGV |
Publications (1)
Publication Number | Publication Date |
---|---|
CN112767540A true CN112767540A (en) | 2021-05-07 |
Family
ID=75701721
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN202110055012.4A Pending CN112767540A (en) | 2021-01-15 | 2021-01-15 | Automatic loading method for AGV |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN112767540A (en) |
Cited By (13)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN113192058A (en) * | 2021-05-21 | 2021-07-30 | 中国矿业大学(北京) | Intelligent brick pile loading system based on computer vision and loading method thereof |
CN113959337A (en) * | 2021-10-26 | 2022-01-21 | 中煤科工智能储装技术有限公司 | Method for calculating real-time accumulation amount of bulk cargo during loading based on single-line laser radar |
CN114455511A (en) * | 2021-12-27 | 2022-05-10 | 劢微机器人科技(深圳)有限公司 | Forklift loading method and equipment and computer readable storage medium |
CN114648233A (en) * | 2022-03-29 | 2022-06-21 | 苏州艾吉威机器人有限公司 | Dynamic station cargo carrying method and system |
CN114663001A (en) * | 2022-05-25 | 2022-06-24 | 广州市广播电视台 | Prompt message generation method applied to relay vehicle and electronic equipment |
CN115170648A (en) * | 2022-06-29 | 2022-10-11 | 广西柳工机械股份有限公司 | Carriage pose determining method and device |
CN115892823A (en) * | 2022-12-01 | 2023-04-04 | 国网智能科技股份有限公司 | Material storage and matching inspection integrated system and method |
CN115924798A (en) * | 2022-12-20 | 2023-04-07 | 广东嘉腾机器人自动化有限公司 | Forklift AGV goods taking method and system based on visual recognition |
CN115984195A (en) * | 2022-12-16 | 2023-04-18 | 秦皇岛燕大滨沅科技发展有限公司 | Three-dimensional point cloud-based carriage contour detection method and system |
JP7288568B1 (en) * | 2023-03-13 | 2023-06-08 | X-Prospect株式会社 | Automatic measurement system |
CN116597681A (en) * | 2023-04-06 | 2023-08-15 | 无锡汉尚建筑装饰工程有限公司 | Intelligent building operation management system based on GIS |
CN116757331A (en) * | 2023-08-11 | 2023-09-15 | 山东捷瑞数字科技股份有限公司 | Method, device, equipment and medium for generating stacking scheme based on industrial vision |
CN114265375B (en) * | 2021-11-25 | 2024-06-04 | 云南昆船智能装备有限公司 | System and method for storing and taking goods of flat truck by AGV |
Citations (4)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN107601062A (en) * | 2017-09-25 | 2018-01-19 | 中铁第四勘察设计院集团有限公司 | A kind of high ferro physical-distribution intelligent handling system and method based on AGV dollies |
CN108238460A (en) * | 2018-01-09 | 2018-07-03 | 富通集团(嘉善)通信技术有限公司 | Optical cable automatic loading system and method |
CN111470327A (en) * | 2020-05-25 | 2020-07-31 | 常州墨狄机器人科技有限公司 | Visual stacking equipment and visual information processing method thereof |
CN111652936A (en) * | 2020-08-06 | 2020-09-11 | 天津迦自机器人科技有限公司 | Three-dimensional sensing and stacking planning method and system for open container loading |
-
2021
- 2021-01-15 CN CN202110055012.4A patent/CN112767540A/en active Pending
Patent Citations (4)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN107601062A (en) * | 2017-09-25 | 2018-01-19 | 中铁第四勘察设计院集团有限公司 | A kind of high ferro physical-distribution intelligent handling system and method based on AGV dollies |
CN108238460A (en) * | 2018-01-09 | 2018-07-03 | 富通集团(嘉善)通信技术有限公司 | Optical cable automatic loading system and method |
CN111470327A (en) * | 2020-05-25 | 2020-07-31 | 常州墨狄机器人科技有限公司 | Visual stacking equipment and visual information processing method thereof |
CN111652936A (en) * | 2020-08-06 | 2020-09-11 | 天津迦自机器人科技有限公司 | Three-dimensional sensing and stacking planning method and system for open container loading |
Cited By (15)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN113192058A (en) * | 2021-05-21 | 2021-07-30 | 中国矿业大学(北京) | Intelligent brick pile loading system based on computer vision and loading method thereof |
CN113959337B (en) * | 2021-10-26 | 2024-03-15 | 中煤科工智能储装技术有限公司 | Bulk cargo loading real-time accumulation amount calculating method based on single-line laser radar |
CN113959337A (en) * | 2021-10-26 | 2022-01-21 | 中煤科工智能储装技术有限公司 | Method for calculating real-time accumulation amount of bulk cargo during loading based on single-line laser radar |
CN114265375B (en) * | 2021-11-25 | 2024-06-04 | 云南昆船智能装备有限公司 | System and method for storing and taking goods of flat truck by AGV |
CN114455511A (en) * | 2021-12-27 | 2022-05-10 | 劢微机器人科技(深圳)有限公司 | Forklift loading method and equipment and computer readable storage medium |
CN114648233A (en) * | 2022-03-29 | 2022-06-21 | 苏州艾吉威机器人有限公司 | Dynamic station cargo carrying method and system |
CN114663001A (en) * | 2022-05-25 | 2022-06-24 | 广州市广播电视台 | Prompt message generation method applied to relay vehicle and electronic equipment |
CN115170648A (en) * | 2022-06-29 | 2022-10-11 | 广西柳工机械股份有限公司 | Carriage pose determining method and device |
CN115892823A (en) * | 2022-12-01 | 2023-04-04 | 国网智能科技股份有限公司 | Material storage and matching inspection integrated system and method |
CN115984195B (en) * | 2022-12-16 | 2023-09-29 | 秦皇岛燕大滨沅科技发展有限公司 | Carriage contour detection method and system based on three-dimensional point cloud |
CN115984195A (en) * | 2022-12-16 | 2023-04-18 | 秦皇岛燕大滨沅科技发展有限公司 | Three-dimensional point cloud-based carriage contour detection method and system |
CN115924798A (en) * | 2022-12-20 | 2023-04-07 | 广东嘉腾机器人自动化有限公司 | Forklift AGV goods taking method and system based on visual recognition |
JP7288568B1 (en) * | 2023-03-13 | 2023-06-08 | X-Prospect株式会社 | Automatic measurement system |
CN116597681A (en) * | 2023-04-06 | 2023-08-15 | 无锡汉尚建筑装饰工程有限公司 | Intelligent building operation management system based on GIS |
CN116757331A (en) * | 2023-08-11 | 2023-09-15 | 山东捷瑞数字科技股份有限公司 | Method, device, equipment and medium for generating stacking scheme based on industrial vision |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN112767540A (en) | Automatic loading method for AGV | |
CN111461107B (en) | Material handling method, apparatus and system for identifying a region of interest | |
CN101685001B (en) | System for measuring three-dimensional profile | |
KR100996978B1 (en) | Method and apparatus for quay container crane-based automated optical container code recognition with positional identification | |
US20200231386A1 (en) | Systems and methods for automated guided vehicle control | |
CN109678055A (en) | Mobile hanging device control system and method, server, mobile hanging device | |
EP3203264A1 (en) | Method of imaging an object for tracking and documentation in transportation and storage | |
CN109969989B (en) | Driving strategy determination method, intelligent forklift and storage medium | |
CN113759906B (en) | Vehicle alignment method and device, computer equipment and storage medium | |
CN114030805B (en) | Warehouse system, shuttle vehicle for warehouse system and navigation method of shuttle vehicle | |
US20190077600A1 (en) | Systems and Methods for Yard Management at Distribution Centers | |
CN114648233A (en) | Dynamic station cargo carrying method and system | |
CN114455511A (en) | Forklift loading method and equipment and computer readable storage medium | |
CN113283845A (en) | Intelligent storage yard management system and method | |
CN113052531A (en) | Unmanned logistics vehicle distribution method and system based on motion analysis | |
CN116757350A (en) | Unmanned forklift cluster scheduling processing system | |
CN113841101A (en) | Method for creating an environment map for use in autonomous navigation of a mobile robot | |
CN116958255A (en) | Intelligent loading system based on three-dimensional point cloud data processing technology and using method | |
KR102529256B1 (en) | Method And Apparatus for Managing Cargo Unloading Plan by Using Cloud | |
CN117011362A (en) | Method for calculating cargo volume and method for dynamically calculating volume rate | |
CN115909204A (en) | Method and device for determining vehicle motion strategy based on container spreader detection | |
US11960299B2 (en) | System for autonomous and semi-autonomous material handling in an outdoor yard | |
CN115334130A (en) | Intelligent network operation service platform and method | |
CN114758163A (en) | Forklift movement control method and device, electronic equipment and storage medium | |
CN113947360A (en) | Scrap steel tracing method and system for scrap steel storage yard |
Legal Events
Date | Code | Title | Description |
---|---|---|---|
PB01 | Publication | ||
PB01 | Publication | ||
SE01 | Entry into force of request for substantive examination | ||
SE01 | Entry into force of request for substantive examination |