AGV multi-vehicle intelligent collaborative operation method, system and medium based on the Internet of Things

Abstract

The present invention discloses an AGV multi-vehicle intelligent collaborative operation method, system and medium based on the Internet of Things. A laser positioning device and a visual device are used to construct a 3D point cloud map and a basic visual map, and the laser and visual positioning information of the AVG forklift is collected through the Internet of Things. The navigation analysis platform controls the AVG forklift in real time, performs positioning analysis and calculates the deviation value according to the navigation mode (laser or vision). When the deviation exceeds the threshold, the system uses the information of the corresponding module for real-time correction. At the same time, the system records the forklift positioning deviation, marks the deviation-intensive area and adjusts the task plan to ensure accurate navigation and efficient warehousing operations. The present invention makes full use of different navigation modes. When performing subsequent positioning analysis, the positioning data is judged and analyzed for deviation based on navigation analysis of different modes and different map analysis, thereby improving positioning accuracy and navigation safety, and at the same time improving driving safety in complex dynamic environments.

Description

Technical Field

[0001] The present invention relates to the field of AGV navigation analysis, and more specifically, to an AGV multi-vehicle intelligent collaborative operation method, system and medium based on the Internet of Things. Background Art

[0002] With the rapid development of Internet of Things technology, AGV has been widely used in the fields of intelligent logistics, warehouse management, etc. However, the traditional AGV operation mode mostly relies on a single navigation technology, such as laser navigation or visual navigation, which has problems such as low positioning accuracy and poor environmental adaptability. Its positioning data source is relatively single, and it lacks the response to the deviation of real-time position, which greatly reduces the operation efficiency of AGV.

[0003] Therefore, there is an urgent need to develop an intelligent collaborative operation method for multiple AGVs based on the Internet of Things. Summary of the invention

[0004] The present invention overcomes the defects of the prior art and proposes an AGV multi-vehicle intelligent collaborative operation method, system and medium based on the Internet of Things.

[0005] The first aspect of the present invention provides an AGV multi-vehicle intelligent collaborative operation method based on the Internet of Things, comprising:

[0006] S1: Based on the preset laser positioning device, the laser signal is emitted and collected in the target storage area, and a 3D point cloud map is constructed based on the laser signal;

[0007] S2: Based on the preset visual device, multi-angle image data of the target storage area is collected and a basic visual map is constructed, and the 3D point cloud map is associated with the basic visual map by map path;

[0008] S3: Based on the laser module and the visual module, real-time positioning data is collected from multiple target AVG forklifts, and laser positioning information and visual positioning information are obtained. All positioning information is uploaded to the navigation analysis platform through the Internet of Things.

[0009] S4: through the preset task plan, multiple target AVG forklifts are navigated and controlled in real time, and the navigation mode is switched in real time in different navigation tasks, and the current navigation mode is judged. If the current navigation mode is laser module navigation, the laser positioning information collected in real time is imported into the point cloud map for positioning analysis, and the second positioning information in the corresponding basic visual map is obtained through map association, and the positioning deviation is calculated based on the laser positioning information and the second positioning information to obtain a first deviation value. The visual positioning information is collected in real time based on the visual module, and the positioning deviation is calculated based on the visual positioning information and the second positioning information to obtain a second deviation value. If the first and second deviation values ​​are both greater than the threshold, the laser positioning information is corrected in real time through the visual positioning information, and the real-time navigation of the target AVG forklift is performed through the corrected positioning;

[0010] S5: Determine the current navigation mode. If the current navigation mode is visual module navigation, import the real-time collected visual positioning information into the basic visual map for positioning analysis and positioning deviation value calculation in the same way as S4, and correct the visual positioning information in real time through the actual laser positioning information;

[0011] S6: Within a time period, the positioning deviations of multiple target AVG forklifts are recorded, all deviation points are counted in the point cloud map, and the distribution information of the deviation points is obtained. The dense areas of the deviation points are marked according to the distribution information of the deviation points and the preset task plan is adjusted.

[0012] In this solution, based on the preset laser positioning device, laser signals are emitted and collected in the target storage area, and a 3D point cloud map is constructed based on the laser signals, specifically:

[0013] Determine the target storage area and initialize the scanning parameters of the preset laser positioning device;

[0014] Scan the target storage area at multiple angles and collect spatial coordinate point data through a preset laser positioning device;

[0015] Constructing an original point cloud map through the spatial coordinate point data;

[0016] Identify and mark fixed equipment, paths, and buildings in the original point cloud map to form a complete 3D point cloud map.

[0017] In this solution, based on the preset visual device, multi-angle image data of the target storage area is collected and a basic visual map is constructed, and the 3D point cloud map is associated with the basic visual map by map path, specifically:

[0018] Through the preset visual device, multi-angle image data of the target storage area is collected to obtain multi-angle area image data;

[0019] Build a basic model based on the area and contour information of the target storage area;

[0020] Based on the multi-angle area image data, paths and objects are identified and marked, and based on the identification results, a basic model is filled to form a basic visual map.

[0021] In this solution, the 3D point cloud map and the basic visual map are associated with the map path, specifically:

[0022] Mark the buildings and fixed equipment in the point cloud map as key points, and obtain corresponding location information to form first key point information;

[0023] Based on the first key point information, marking and recording the position information of the corresponding key point from the basic visual map to form the second key point information;

[0024] Based on the 3D point cloud map and the basic visual map, the first key point information and the second key point information are mapped with respect to position information association, and key point association information is formed;

[0025] Extracting path information from the 3D point cloud map and the basic visual map to obtain first path information and second path information respectively;

[0026] Performing location information association mapping on the first path information and the second path information, and forming path point association information;

[0027] The key point association information and the path point association information are integrated to form map path association data.

[0028] In this solution, the S4 step is specifically:

[0029] Obtain multiple navigation tasks based on the preset task plan, select the navigation mode based on the distance, accuracy, and route running time of the navigation task, and switch the navigation mode in real time during the operation of the preset task plan;

[0030] In a preset judgment cycle, the current navigation mode is judged. If the current navigation mode is laser module navigation, the real-time collected laser positioning information and the task navigation route are obtained, and the laser positioning information is imported into the point cloud map for positioning analysis. The laser positioning information is positioned and associated through the path point association information in the map path association data, and the associated positioning in the basic visual map is obtained to form the second positioning information;

[0031] Get multiple key points based on map path association data;

[0032] Using the laser positioning information, respectively calculating the coordinate distances between the laser positioning position point and a plurality of key points in the point cloud map to form a first set of distance data;

[0033] Based on the second positioning information, respectively calculating the coordinate distances between the second positioning point and the plurality of key points in the basic visual map to form a second set of distance data;

[0034] Perform vector conversion on the first set of distance data and the second set of distance data to form two vectors, calculate the distance between the two vectors based on the standard Euclidean distance, and use the calculation result as the first deviation value;

[0035] Based on the visual module, the visual positioning information is collected in real time, and the coordinate deviation based on the basic visual map is calculated by using the visual positioning information and the second positioning information to obtain a second deviation value;

[0036] Determine whether the first deviation value and the second deviation value are both greater than a preset threshold value. If so, analyze the location point of the visual positioning corresponding to the point cloud map based on the visual positioning information and the map path association data to obtain the third positioning information;

[0037] The laser positioning information is corrected in real time by the third positioning information, and the target AVG forklift is navigated in real time after the positioning is corrected.

[0038] In this solution, the step S5 is specifically as follows:

[0039] In a preset judgment cycle, the current navigation mode is judged. If the current navigation mode is visual module navigation, the real-time collected visual positioning information and the task navigation route are obtained, and the visual positioning information is imported into the basic visual map for positioning analysis. The visual positioning information is positioned and associated through the path point association information in the map path association data, and the associated positioning in the point cloud map is obtained to form the second positioning information;

[0040] Get multiple key points based on map path association data;

[0041] By using the visual positioning information, respectively calculating the coordinate distances between the visual positioning position point and a plurality of key points in the basic visual map to form a first set of distance data;

[0042] Based on the second positioning information, respectively calculating the coordinate distances between the second positioning point and the plurality of key points in the point cloud map to form a second set of distance data;

[0043] Perform vector conversion on the first set of distance data and the second set of distance data to form two vectors, calculate the distance between the two vectors based on the standard Euclidean distance, and use the calculation result as the first deviation value;

[0044] Based on the real-time collection of laser positioning information by the laser module, coordinate deviation calculation based on the point cloud map is performed through the laser positioning information and the second positioning information to obtain a second deviation value;

[0045] Determine whether the first deviation value and the second deviation value are both greater than a preset threshold value. If so, analyze the location point of the laser positioning corresponding to the basic visual map based on the laser positioning information and the map path association data to obtain the third positioning information;

[0046] The visual positioning information is corrected in real time through the third positioning information, and the real-time navigation of the target AVG forklift is performed after the corrected positioning.

[0047] In this solution, multiple target AVG forklifts are positioned and deviated within a period of time, all deviation points are counted in the point cloud map, and the distribution information of the deviation points is obtained. The dense areas of the deviation points are marked and the preset task plan is adjusted according to the distribution information of the deviation points, specifically:

[0048] In a period of time, multiple target AVG forklifts are analyzed for real-time positioning deviation. If positioning correction occurs, the corresponding position point is recorded in the point cloud map and marked as a deviation point;

[0049] Generate deviation point distribution information based on all deviation points;

[0050] By using the deviation point distribution information, a deviation point dense area is screened out in the point cloud map, and based on the preset task plan, the navigation path tasks passing through the dense area are screened to obtain the navigation task plan to be adjusted;

[0051] Perform path complexity analysis on the navigation task plan to be adjusted, split and adjust the tasks, and generate an optimized task plan.

[0052] The second aspect of the present invention further provides an AGV multi-vehicle intelligent collaborative operation system based on the Internet of Things, the system comprising: a memory, a processor, the memory comprising an AGV multi-vehicle intelligent collaborative operation program based on the Internet of Things, and the AGV multi-vehicle intelligent collaborative operation program based on the Internet of Things is executed by the processor to implement the following steps:

[0053] S1: Based on the preset laser positioning device, the laser signal is emitted and collected in the target storage area, and a 3D point cloud map is constructed based on the laser signal;

[0054] S2: Based on the preset visual device, multi-angle image data of the target storage area is collected and a basic visual map is constructed, and the 3D point cloud map is associated with the basic visual map by map path;

[0055] S3: Based on the laser module and the visual module, real-time positioning data is collected from multiple target AVG forklifts, and laser positioning information and visual positioning information are obtained. All positioning information is uploaded to the navigation analysis platform through the Internet of Things.

[0056] S4: through the preset task plan, multiple target AVG forklifts are navigated and controlled in real time, and the navigation mode is switched in real time in different navigation tasks, and the current navigation mode is judged. If the current navigation mode is laser module navigation, the laser positioning information collected in real time is imported into the point cloud map for positioning analysis, and the second positioning information in the corresponding basic visual map is obtained through map association, and the positioning deviation is calculated based on the laser positioning information and the second positioning information to obtain a first deviation value. The visual positioning information is collected in real time based on the visual module, and the positioning deviation is calculated based on the visual positioning information and the second positioning information to obtain a second deviation value. If the first and second deviation values ​​are both greater than the threshold, the laser positioning information is corrected in real time through the visual positioning information, and the real-time navigation of the target AVG forklift is performed through the corrected positioning;

[0057] S5: Determine the current navigation mode. If the current navigation mode is visual module navigation, import the real-time collected visual positioning information into the basic visual map for positioning analysis and positioning deviation value calculation in the same way as S4, and correct the visual positioning information in real time through the actual laser positioning information;

[0058] S6: Within a time period, the positioning deviations of multiple target AVG forklifts are recorded, all deviation points are counted in the point cloud map, and the distribution information of the deviation points is obtained. The dense areas of the deviation points are marked according to the distribution information of the deviation points and the preset task plan is adjusted.

[0059] The third aspect of the present invention also provides a computer-readable storage medium, which includes an AGV multi-vehicle intelligent collaborative operation program based on the Internet of Things. When the AGV multi-vehicle intelligent collaborative operation program based on the Internet of Things is executed by a processor, the steps of the AGV multi-vehicle intelligent collaborative operation method based on the Internet of Things as described in any one of the above items are implemented.

[0060] The present invention discloses an AGV multi-vehicle intelligent collaborative operation method, system and medium based on the Internet of Things. A laser positioning device and a visual device are used to construct a 3D point cloud map and a basic visual map, and the laser and visual positioning information of the AVG forklift is collected through the Internet of Things. The navigation analysis platform controls the AVG forklift in real time, performs positioning analysis and calculates the deviation value according to the navigation mode (laser or vision). When the deviation exceeds the threshold, the system uses the information of the corresponding module for real-time correction. At the same time, the system records the forklift positioning deviation, marks the deviation-intensive area and adjusts the task plan to ensure accurate navigation and efficient warehousing operations. The present invention makes full use of different navigation modes. When performing subsequent positioning analysis, the positioning data is judged and analyzed for deviation based on navigation analysis of different modes and different map analysis, thereby improving positioning accuracy and navigation safety, and at the same time improving driving safety in complex dynamic environments. BRIEF DESCRIPTION OF THE DRAWINGS

[0061] Figure 1 A flow chart of an AGV multi-vehicle intelligent collaborative operation method based on the Internet of Things of the present invention is shown;

[0062] Figure 2 A block diagram of an AGV multi-vehicle intelligent collaborative operation system based on the Internet of Things of the present invention is shown. DETAILED DESCRIPTION

[0063] In order to more clearly understand the above-mentioned purpose, features and advantages of the present invention, the present invention is further described in detail below in conjunction with the accompanying drawings and specific embodiments. It should be noted that the embodiments of the present application and the features in the embodiments can be combined with each other without conflict.

[0064] In the following description, many specific details are set forth to facilitate a full understanding of the present invention. However, the present invention may also be implemented in other ways different from those described herein. Therefore, the protection scope of the present invention is not limited to the specific embodiments disclosed below.

[0065] Figure 1 The flowchart of the present invention shows a method for intelligent collaborative operation of multiple AGVs based on the Internet of Things.

[0066] like Figure 1 As shown, the first aspect of the present invention provides an AGV multi-vehicle intelligent collaborative operation method based on the Internet of Things, comprising:

[0067] S1: Based on the preset laser positioning device, the laser signal is emitted and collected in the target storage area, and a 3D point cloud map is constructed based on the laser signal;

[0068] S2: Based on the preset visual device, multi-angle image data of the target storage area is collected and a basic visual map is constructed, and the 3D point cloud map is associated with the basic visual map by map path;

[0069] S3: Based on the laser module and the visual module, real-time positioning data is collected from multiple target AVG forklifts, and laser positioning information and visual positioning information are obtained. All positioning information is uploaded to the navigation analysis platform through the Internet of Things.

[0070] S4: through the preset task plan, multiple target AVG forklifts are navigated and controlled in real time, and the navigation mode is switched in real time in different navigation tasks, and the current navigation mode is judged. If the current navigation mode is laser module navigation, the laser positioning information collected in real time is imported into the point cloud map for positioning analysis, and the second positioning information in the corresponding basic visual map is obtained through map association, and the positioning deviation is calculated based on the laser positioning information and the second positioning information to obtain a first deviation value. The visual positioning information is collected in real time based on the visual module, and the positioning deviation is calculated based on the visual positioning information and the second positioning information to obtain a second deviation value. If the first and second deviation values ​​are both greater than the threshold, the laser positioning information is corrected in real time through the visual positioning information, and the real-time navigation of the target AVG forklift is performed through the corrected positioning;

[0071] S5: Determine the current navigation mode. If the current navigation mode is visual module navigation, import the real-time collected visual positioning information into the basic visual map for positioning analysis and positioning deviation value calculation in the same way as S4, and correct the visual positioning information in real time through the actual laser positioning information;

[0072] S6: Within a time period, the positioning deviations of multiple target AVG forklifts are recorded, all deviation points are counted in the point cloud map, and the distribution information of the deviation points is obtained. The dense areas of the deviation points are marked according to the distribution information of the deviation points and the preset task plan is adjusted.

[0073] It should be noted that the basic visual map is a three-dimensional visualization map model. The navigation analysis platform is a data platform built based on the current cargo transportation scene, which is used to store the corresponding AGV forklift positioning data, control information and cargo transportation data, which can be used for subsequent research on forklift operation data.

[0074] According to an embodiment of the present invention, based on the preset laser positioning device, laser signals are emitted and collected for the target storage area, and a 3D point cloud map is constructed based on the laser signals, specifically:

[0075] Determine the target storage area and initialize the scanning parameters of the preset laser positioning device;

[0076] Scan the target storage area at multiple angles and collect spatial coordinate point data through a preset laser positioning device;

[0077] Constructing an original point cloud map through the spatial coordinate point data;

[0078] Identify and mark fixed equipment, paths, and buildings in the original point cloud map to form a complete 3D point cloud map.

[0079] It should be noted that the target storage area is an area set by the user, generally the entire area of ​​a storage warehouse. The fixed equipment refers to items that are relatively fixedly placed in the target storage area, such as shelves, display equipment, lighting equipment, etc. The forklift can be navigated and operated in laser detection mode through the 3D point cloud map. The preset laser positioning device includes scanning equipment, laser radar, laser module in the forklift and other equipment.

[0080] According to an embodiment of the present invention, based on the preset visual device, multi-angle image data of the target storage area is collected and a basic visual map is constructed, and the 3D point cloud map is associated with the basic visual map by map path, specifically:

[0081] Through the preset visual device, multi-angle image data of the target storage area is collected to obtain multi-angle area image data;

[0082] Build a basic model based on the area and contour information of the target storage area;

[0083] Based on the multi-angle area image data, paths and objects are identified and marked, and based on the identification results, a basic model is filled to form a basic visual map.

[0084] It should be noted that the preset visual device includes a plurality of high-definition camera devices arranged in the target storage area, which are used to collect multi-angle images in the area and build a map model. The basic visual map is a map model built based on image data.

[0085] It is worth mentioning that different map models have corresponding advantages and disadvantages in different forklift task environments. The present invention makes full use of different navigation modes. In the subsequent positioning analysis, the positioning data is judged and analyzed for deviation based on navigation analysis of different modes and different map analysis, thereby improving positioning accuracy and navigation safety, and at the same time improving driving safety in complex dynamic environments.

[0086] According to an embodiment of the present invention, the 3D point cloud map is associated with the basic visual map by performing a map path, specifically:

[0087] Mark the buildings and fixed equipment in the point cloud map as key points, and obtain corresponding location information to form first key point information;

[0088] Based on the first key point information, marking and recording the position information of the corresponding key point from the basic visual map to form the second key point information;

[0089] Based on the 3D point cloud map and the basic visual map, the first key point information and the second key point information are mapped with respect to position information association, and key point association information is formed;

[0090] Extracting path information from the 3D point cloud map and the basic visual map to obtain first path information and second path information respectively;

[0091] Performing location information association mapping on the first path information and the second path information, and forming path point association information;

[0092] The key point association information and the path point association information are integrated to form map path association data.

[0093] It should be noted that the building refers to building obstacles, wall columns, etc. existing in the target storage area. The key point information includes key points and spatial position information. The position information is recorded as the position information in the basic visual map. In the first key point information and the second key point information, the first and second key points are points at the same position in the corresponding target storage area, and the information of the two exists in different map models. Among them, the extraction of path information includes the number of paths, path type, and path position information.

[0094] The key point association information is used to associate multiple key points set in the area in the two maps. After the association, the key point positions of the two maps can be found through the association information. For example, if a key point K1 is known in the point cloud map, the corresponding key point K2 in the basic visual map can be obtained through the association information. K1 and K2 are the same position point in the target storage area. Subsequently, the key points are used to analyze the degree of deviation of the real-time position.

[0095] In the path point association information, the present invention analyzes the path point information of both maps and associates the same path through the actual area location, so that the corresponding location point information in the other map can be quickly found based on a certain point in one of the maps.

[0096] It is worth mentioning that in this embodiment, buildings, fixed equipment, etc. are used as key points of the map, which can subsequently perform deviation analysis based on the key points based on a certain positioning. The key points are used as fixed points of a map to perform corresponding deviation calculations, which can subsequently improve the accuracy of positioning deviation analysis.

[0097] According to an embodiment of the present invention, the step S4 is specifically as follows:

[0098] Obtain multiple navigation tasks based on the preset task plan, select the navigation mode based on the distance, accuracy, and route running time of the navigation task, and switch the navigation mode in real time during the operation of the preset task plan;

[0099] In a preset judgment cycle, the current navigation mode is judged. If the current navigation mode is laser module navigation, the real-time collected laser positioning information and the task navigation route are obtained, and the laser positioning information is imported into the point cloud map for positioning analysis. The laser positioning information is positioned and associated through the path point association information in the map path association data, and the associated positioning in the basic visual map is obtained to form the second positioning information;

[0100] Get multiple key points based on map path association data;

[0101] Using the laser positioning information, respectively calculating the coordinate distances between the laser positioning position point and a plurality of key points in the point cloud map to form a first set of distance data;

[0102] Based on the second positioning information, respectively calculating the coordinate distances between the second positioning point and the plurality of key points in the basic visual map to form a second set of distance data;

[0103] Perform vector conversion on the first set of distance data and the second set of distance data to form two vectors, calculate the distance between the two vectors based on the standard Euclidean distance, and use the calculation result as the first deviation value;

[0104] Based on the visual module, the visual positioning information is collected in real time, and the coordinate deviation based on the basic visual map is calculated by using the visual positioning information and the second positioning information to obtain a second deviation value;

[0105] Determine whether the first deviation value and the second deviation value are both greater than a preset threshold value. If so, analyze the location point of the visual positioning corresponding to the point cloud map based on the visual positioning information and the map path association data to obtain the third positioning information;

[0106] The laser positioning information is corrected in real time by the third positioning information, and the target AVG forklift is navigated in real time after the positioning is corrected.

[0107] It should be noted that visual detection and laser detection have their own advantages and disadvantages. By analyzing the task volume, navigation distance, and environment of the corresponding preset task plan, the corresponding optimal mode (visual or laser mode) is selected for different tasks. In addition, based on the complexity of the environment, different AGV navigation models can also be selected based on the predetermined configuration mode, such as selecting the visual or laser mode based on the navigation accuracy requirement or the navigation distance requirement. At the same time, the mode can also be configured based on user settings. The preset task plan includes navigation tasks and AGV control tasks. The preset judgment cycle is a time interval set by the user, which is used to periodically judge the positioning accuracy and positioning correction. The size of the cycle is determined by the running time of the entire navigation task and the navigation task volume. The larger the navigation task volume, the more complicated the operation of the AGV forklift is generally. At this time, the judgment cycle can be shortened and the judgment frequency can be increased. The first set of distance data and the second set of distance data are respectively converted into two vectors. The number of dimensions of the vector is based on the number of values ​​of a set of distance data (equal to the number of key points), and the corresponding value of each dimension is the specific value of the corresponding distance data. In the two sets of distance data, the number of values ​​is equal. In the process of calculating the coordinate distance between the positioning point and multiple key points, if the distance is calculated using the laser position point on the point cloud map, the first key point (information) is used for calculation. If the distance is calculated using the position point (second positioning information) of the basic visual map, the second key point (information) is used for calculation. The first key point and the second key point exist in the two maps respectively, and the spatial positions of the two in the actual target storage area are consistent. The task navigation route is obtained based on the preset task plan.

[0108] It is worth mentioning that in a complex dynamic storage environment, there are often situations such as inaccurate positioning through a single navigation mode. For example, under a complex path task, the deviation between the path and the positioning will accumulate over time. For another example, after the navigation mode is set, the ambient light, road obstacles, and road surface flatness change dynamically. At this time, positioning deviation will occur in the set navigation mode. Therefore, the present invention analyzes the current navigation mode in real time, based on the corresponding position points, through map association, simulates and derives the positioning situation in another navigation mode, performs deviation analysis on the positioning of the two, and introduces the positioning information of another navigation mode under actual conditions, and performs secondary deviation analysis on the positioning information in another navigation mode and the positioning in the current navigation mode. If both deviations exceed expectations, it can be determined that there is a positioning deviation at present. This method can be effectively applied to the forklift operation process with complex dynamic environments and complex path tasks. Through deviation analysis, the positioning information in another navigation mode is corrected and positioned in real time to achieve high efficiency, sustainability and stability of forklift operation.

[0109] In addition, in traditional technologies, there is a lack of corresponding supplementary positioning analysis methods for situations where real-time positioning is inaccurate or there is a deviation in positioning, which may lead to certain defects in the real-time analysis and positioning process, affecting the overall AGV forklift navigation task planning and operation.

[0110] According to an embodiment of the present invention, the step S5 is specifically as follows:

[0111] In a preset judgment cycle, the current navigation mode is judged. If the current navigation mode is visual module navigation, the real-time collected visual positioning information and the task navigation route are obtained, and the visual positioning information is imported into the basic visual map for positioning analysis. The visual positioning information is positioned and associated through the path point association information in the map path association data, and the associated positioning in the point cloud map is obtained to form the second positioning information;

[0112] Get multiple key points based on map path association data;

[0113] By using the visual positioning information, respectively calculating the coordinate distances between the visual positioning position point and a plurality of key points in the basic visual map to form a first set of distance data;

[0114] Based on the second positioning information, respectively calculating the coordinate distances between the second positioning point and the plurality of key points in the point cloud map to form a second set of distance data;

[0115] Perform vector conversion on the first set of distance data and the second set of distance data to form two vectors, calculate the distance between the two vectors based on the standard Euclidean distance, and use the calculation result as the first deviation value;

[0116] Based on the real-time collection of laser positioning information by the laser module, coordinate deviation calculation based on the point cloud map is performed through the laser positioning information and the second positioning information to obtain a second deviation value;

[0117] Determine whether the first deviation value and the second deviation value are both greater than a preset threshold value. If so, analyze the location point of the laser positioning corresponding to the basic visual map based on the laser positioning information and the map path association data to obtain the third positioning information;

[0118] The visual positioning information is corrected in real time through the third positioning information, and the real-time navigation of the target AVG forklift is performed after the corrected positioning.

[0119] It should be noted that there are two cases for the second positioning information. The first case is that the current navigation mode in S4 is laser module navigation, and the positioning information is obtained by importing the laser positioning information into the point cloud map for correlation analysis. The second case is that the current navigation mode in S5 is visual module navigation, and the positioning information is obtained by importing the visual positioning information into the basic visual map for correlation analysis. Similarly, there are two cases for the third positioning information. In the invention, the methods of S4 and S5 are the same, based on the positioning correction of one navigation mode to another navigation mode.

[0120] According to an embodiment of the present invention, within a time period, multiple target AVG forklifts are positioned and deviated, all deviation points are counted in the point cloud map, and deviation point distribution information is obtained. The dense area of ​​the deviation points is marked and the preset task plan is adjusted according to the deviation point distribution information, specifically:

[0121] In a period of time, multiple target AVG forklifts are analyzed for real-time positioning deviation. If positioning correction occurs, the corresponding position point is recorded in the point cloud map and marked as a deviation point;

[0122] Generate deviation point distribution information based on all deviation points;

[0123] By using the deviation point distribution information, a deviation point dense area is screened out in the point cloud map, and based on the preset task plan, the navigation path tasks passing through the dense area are screened to obtain the navigation task plan to be adjusted;

[0124] Perform path complexity analysis on the navigation task plan to be adjusted, split and adjust the tasks, and generate an optimized task plan.

[0125] It should be noted that the time period is a preset time period. In the case of screening out the dense area of ​​deviation points in the point cloud map, the number of deviation points in each small area can be counted in the point cloud map, and small areas with a larger number can be screened out as dense areas. In the case of recording the corresponding position points in the point cloud map and marking them as deviation points, based on the embodiment of the present invention, distribution statistics can also be performed based on the basic visual map.

[0126] It is worth mentioning that in the process of navigation analysis, in some complex navigation task paths, due to the complexity of the tasks, multiple target AVG forklifts may easily have high-frequency positioning deviation problems in a certain area or a certain path when navigating. Therefore, the present invention counts the deviation points of the target storage area based on the map model within a time period, screens out dense areas with more deviation points in the statistical process, screens out and marks the corresponding navigation tasks in the dense areas, performs complexity analysis and task splitting on the corresponding navigation tasks, thereby simplifying the navigation tasks accordingly or adjusting the path planning, effectively reducing route deviations.

[0127] The navigation task plan to be adjusted is analyzed for path complexity and tasks are split and adjusted to generate an optimized task plan. Specifically, the path location, path length, and regional environmental conditions are analyzed comprehensively. In the subsequent task splitting and corresponding task adjustments are performed on the complex navigation task. For example, a long forklift navigation loading and unloading task is split into multiple tasks, or a complex task is split into multiple subtasks and assigned to multiple AVG forklifts for collaborative operation. The optimized task plan records the collaborative operation plan after optimization and adjustment. Through the present invention, the efficiency reduction and safety problems caused by positioning deviation are effectively reduced.

[0128] Figure 2 A block diagram of an AGV multi-vehicle intelligent collaborative operation system based on the Internet of Things of the present invention is shown.

[0129] The second aspect of the present invention further provides an AGV multi-vehicle intelligent collaborative operation system 2 based on the Internet of Things, the system comprising: a memory 21, a processor 22, the memory comprising an AGV multi-vehicle intelligent collaborative operation program based on the Internet of Things, and the AGV multi-vehicle intelligent collaborative operation program based on the Internet of Things is executed by the processor to implement the following steps:

[0130] S1: Based on the preset laser positioning device, the laser signal is emitted and collected in the target storage area, and a 3D point cloud map is constructed based on the laser signal;

[0131] S2: Based on the preset visual device, multi-angle image data of the target storage area is collected and a basic visual map is constructed, and the 3D point cloud map is associated with the basic visual map by map path;

[0132] S3: Based on the laser module and the visual module, real-time positioning data is collected from multiple target AVG forklifts, and laser positioning information and visual positioning information are obtained. All positioning information is uploaded to the navigation analysis platform through the Internet of Things.

[0133] S4: through the preset task plan, multiple target AVG forklifts are navigated and controlled in real time, and the navigation mode is switched in real time in different navigation tasks, and the current navigation mode is judged. If the current navigation mode is laser module navigation, the laser positioning information collected in real time is imported into the point cloud map for positioning analysis, and the second positioning information in the corresponding basic visual map is obtained through map association, and the positioning deviation is calculated based on the laser positioning information and the second positioning information to obtain a first deviation value. The visual positioning information is collected in real time based on the visual module, and the positioning deviation is calculated based on the visual positioning information and the second positioning information to obtain a second deviation value. If the first and second deviation values ​​are both greater than the threshold, the laser positioning information is corrected in real time through the visual positioning information, and the real-time navigation of the target AVG forklift is performed through the corrected positioning;

[0134] S5: Determine the current navigation mode. If the current navigation mode is visual module navigation, import the real-time collected visual positioning information into the basic visual map for positioning analysis and positioning deviation value calculation in the same way as S4, and correct the visual positioning information in real time through the actual laser positioning information;

[0135] S6: Within a time period, the positioning deviations of multiple target AVG forklifts are recorded, all deviation points are counted in the point cloud map, and the distribution information of the deviation points is obtained. The dense areas of the deviation points are marked according to the distribution information of the deviation points and the preset task plan is adjusted.

[0136] It should be noted that the basic visual map is a three-dimensional visualization map model. The navigation analysis platform is a data platform built based on the current cargo transportation scene, which is used to store the corresponding AGV forklift positioning data, control information and cargo transportation data, which can be used for subsequent research on forklift operation data.

[0137] According to an embodiment of the present invention, based on the preset laser positioning device, laser signals are emitted and collected for the target storage area, and a 3D point cloud map is constructed based on the laser signals, specifically:

[0138] Determine the target storage area and initialize the scanning parameters of the preset laser positioning device;

[0139] Scan the target storage area at multiple angles and collect spatial coordinate point data through a preset laser positioning device;

[0140] Constructing an original point cloud map through the spatial coordinate point data;

[0141] Identify and mark fixed equipment, paths, and buildings in the original point cloud map to form a complete 3D point cloud map.

[0142] It should be noted that the target storage area is an area set by the user, generally the entire area of ​​a storage warehouse. The fixed equipment refers to items that are relatively fixedly placed in the target storage area, such as shelves, display equipment, lighting equipment, etc. The forklift can be navigated and operated in laser detection mode through the 3D point cloud map. The preset laser positioning device includes scanning equipment, laser radar, laser module in the forklift and other equipment.

[0143] According to an embodiment of the present invention, based on the preset visual device, multi-angle image data of the target storage area is collected and a basic visual map is constructed, and the 3D point cloud map is associated with the basic visual map by map path, specifically:

[0144] Through the preset visual device, multi-angle image data of the target storage area is collected to obtain multi-angle area image data;

[0145] Build a basic model based on the area and contour information of the target storage area;

[0146] Based on the multi-angle area image data, paths and objects are identified and marked, and based on the identification results, a basic model is filled to form a basic visual map.

[0147] It should be noted that the preset visual device includes a plurality of high-definition camera devices arranged in the target storage area, which are used to collect multi-angle images in the area and build a map model. The basic visual map is a map model built based on image data.

[0148] It is worth mentioning that different map models have corresponding advantages and disadvantages in different forklift task environments. The present invention makes full use of different navigation modes. In the subsequent positioning analysis, the positioning data is judged and analyzed for deviation based on navigation analysis of different modes and different map analysis, thereby improving positioning accuracy and navigation safety, and at the same time improving driving safety in complex dynamic environments.

[0149] According to an embodiment of the present invention, the 3D point cloud map is associated with the basic visual map by performing a map path, specifically:

[0150] Mark the buildings and fixed equipment in the point cloud map as key points, and obtain corresponding location information to form first key point information;

[0151] Based on the first key point information, marking and recording the position information of the corresponding key point from the basic visual map to form the second key point information;

[0152] Based on the 3D point cloud map and the basic visual map, the first key point information and the second key point information are mapped with respect to position information association, and key point association information is formed;

[0153] Extracting path information from the 3D point cloud map and the basic visual map to obtain first path information and second path information respectively;

[0154] Performing location information association mapping on the first path information and the second path information, and forming path point association information;

[0155] The key point association information and the path point association information are integrated to form map path association data.

[0156] It should be noted that the building refers to building obstacles, wall columns, etc. existing in the target storage area. The key point information includes key points and spatial position information. The position information is recorded as the position information in the basic visual map. In the first key point information and the second key point information, the first and second key points are points at the same position in the corresponding target storage area, and the information of the two exists in different map models. Among them, the extraction of path information includes the number of paths, path type, and path position information.

[0157] The key point association information is used to associate multiple key points set in the area in the two maps. After the association, the key point positions of the two maps can be found through the association information. For example, if a key point K1 is known in the point cloud map, the corresponding key point K2 in the basic visual map can be obtained through the association information. K1 and K2 are the same position point in the target storage area. Subsequently, the key points are used to analyze the degree of deviation of the real-time position.

[0158] In the path point association information, the present invention analyzes the path point information of both maps and associates the same path through the actual area location, so that the corresponding location point information in the other map can be quickly found based on a certain point in one of the maps.

[0159] It is worth mentioning that in this embodiment, buildings, fixed equipment, etc. are used as key points of the map, which can subsequently perform deviation analysis based on the key points based on a certain positioning. The key points are used as fixed points of a map to perform corresponding deviation calculations, which can subsequently improve the accuracy of positioning deviation analysis.

[0160] According to an embodiment of the present invention, the step S4 is specifically as follows:

[0161] Obtain multiple navigation tasks based on the preset task plan, select the navigation mode based on the distance, accuracy, and route running time of the navigation task, and switch the navigation mode in real time during the operation of the preset task plan;

[0162] In a preset judgment cycle, the current navigation mode is judged. If the current navigation mode is laser module navigation, the real-time collected laser positioning information and the task navigation route are obtained, and the laser positioning information is imported into the point cloud map for positioning analysis. The laser positioning information is positioned and associated through the path point association information in the map path association data, and the associated positioning in the basic visual map is obtained to form the second positioning information;

[0163] Get multiple key points based on map path association data;

[0164] Using the laser positioning information, respectively calculating the coordinate distances between the laser positioning position point and a plurality of key points in the point cloud map to form a first set of distance data;

[0165] Based on the second positioning information, respectively calculating the coordinate distances between the second positioning point and the plurality of key points in the basic visual map to form a second set of distance data;

[0166] Perform vector conversion on the first set of distance data and the second set of distance data to form two vectors, calculate the distance between the two vectors based on the standard Euclidean distance, and use the calculation result as the first deviation value;

[0167] Based on the visual module, the visual positioning information is collected in real time, and the coordinate deviation based on the basic visual map is calculated by using the visual positioning information and the second positioning information to obtain a second deviation value;

[0168] Determine whether the first deviation value and the second deviation value are both greater than a preset threshold value. If so, analyze the location point of the visual positioning corresponding to the point cloud map based on the visual positioning information and the map path association data to obtain the third positioning information;

[0169] The laser positioning information is corrected in real time by the third positioning information, and the target AVG forklift is navigated in real time after the positioning is corrected.

[0170] It should be noted that visual detection and laser detection have their own advantages and disadvantages. By analyzing the task volume, navigation distance, and environment of the corresponding preset task plan, the corresponding optimal mode (visual or laser mode) is selected for different tasks. In addition, based on the complexity of the environment, different AGV navigation models can also be selected based on the predetermined configuration mode, such as selecting the visual or laser mode based on the navigation accuracy requirement or the navigation distance requirement. At the same time, the mode can also be configured based on user settings. The preset task plan includes navigation tasks and AGV control tasks. The preset judgment cycle is a time interval set by the user, which is used to periodically judge the positioning accuracy and positioning correction. The size of the cycle is determined by the running time of the entire navigation task and the navigation task volume. The larger the navigation task volume, the more complicated the operation of the AGV forklift is generally. At this time, the judgment cycle can be shortened and the judgment frequency can be increased. The first set of distance data and the second set of distance data are respectively converted into two vectors. The number of dimensions of the vector is based on the number of values ​​of a set of distance data (equal to the number of key points), and the corresponding value of each dimension is the specific value of the corresponding distance data. In the two sets of distance data, the number of values ​​is equal. In the process of calculating the coordinate distance between the positioning point and multiple key points, if the distance is calculated using the laser position point on the point cloud map, the first key point (information) is used for calculation. If the distance is calculated using the position point (second positioning information) of the basic visual map, the second key point (information) is used for calculation. The first key point and the second key point exist in the two maps respectively, and the spatial positions of the two in the actual target storage area are consistent. The task navigation route is obtained based on the preset task plan.

[0171] It is worth mentioning that in a complex dynamic storage environment, there are often situations such as inaccurate positioning through a single navigation mode. For example, under a complex path task, the deviation between the path and the positioning will accumulate over time. For another example, after the navigation mode is set, the ambient light, road obstacles, and road surface flatness change dynamically. At this time, positioning deviation will occur in the set navigation mode. Therefore, the present invention analyzes the current navigation mode in real time, based on the corresponding position points, through map association, simulates and derives the positioning situation in another navigation mode, performs deviation analysis on the positioning of the two, and introduces the positioning information of another navigation mode under actual conditions, and performs secondary deviation analysis on the positioning information in another navigation mode and the positioning in the current navigation mode. If both deviations exceed expectations, it can be determined that there is a positioning deviation at present. This method can be effectively applied to the forklift operation process with complex dynamic environments and complex path tasks. Through deviation analysis, the positioning information in another navigation mode is corrected and positioned in real time to achieve high efficiency, sustainability and stability of forklift operation.

[0172] In addition, in traditional technologies, there is a lack of corresponding supplementary positioning analysis methods for situations where real-time positioning is inaccurate or there is a deviation in positioning, which may lead to certain defects in the real-time analysis and positioning process, affecting the overall AGV forklift navigation task planning and operation.

[0173] According to an embodiment of the present invention, the step S5 is specifically as follows:

[0174] In a preset judgment cycle, the current navigation mode is judged. If the current navigation mode is visual module navigation, the real-time collected visual positioning information and the task navigation route are obtained, and the visual positioning information is imported into the basic visual map for positioning analysis. The visual positioning information is positioned and associated through the path point association information in the map path association data, and the associated positioning in the point cloud map is obtained to form the second positioning information;

[0175] Get multiple key points based on map path association data;

[0176] By using the visual positioning information, respectively calculating the coordinate distances between the visual positioning position point and a plurality of key points in the basic visual map to form a first set of distance data;

[0177] Based on the second positioning information, respectively calculating the coordinate distances between the second positioning point and the plurality of key points in the point cloud map to form a second set of distance data;

[0178] Perform vector conversion on the first set of distance data and the second set of distance data to form two vectors, calculate the distance between the two vectors based on the standard Euclidean distance, and use the calculation result as the first deviation value;

[0179] Based on the real-time collection of laser positioning information by the laser module, coordinate deviation calculation based on the point cloud map is performed through the laser positioning information and the second positioning information to obtain a second deviation value;

[0180] Determine whether the first deviation value and the second deviation value are both greater than a preset threshold value. If so, analyze the location point of the laser positioning corresponding to the basic visual map based on the laser positioning information and the map path association data to obtain the third positioning information;

[0181] The visual positioning information is corrected in real time through the third positioning information, and the real-time navigation of the target AVG forklift is performed after the corrected positioning.

[0182] It should be noted that there are two cases for the second positioning information. The first case is that the current navigation mode in S4 is laser module navigation, and the positioning information is obtained by importing the laser positioning information into the point cloud map for correlation analysis. The second case is that the current navigation mode in S5 is visual module navigation, and the positioning information is obtained by importing the visual positioning information into the basic visual map for correlation analysis. Similarly, there are two cases for the third positioning information. In the invention, the methods of S4 and S5 are the same, based on the positioning correction of one navigation mode to another navigation mode.

[0183] According to an embodiment of the present invention, within a time period, multiple target AVG forklifts are positioned and deviated, all deviation points are counted in the point cloud map, and deviation point distribution information is obtained. The dense area of ​​the deviation points is marked and the preset task plan is adjusted according to the deviation point distribution information, specifically:

[0184] In a period of time, multiple target AVG forklifts are analyzed for real-time positioning deviation. If positioning correction occurs, the corresponding position point is recorded in the point cloud map and marked as a deviation point;

[0185] Generate deviation point distribution information based on all deviation points;

[0186] By using the deviation point distribution information, a deviation point dense area is screened out in the point cloud map, and based on the preset task plan, the navigation path tasks passing through the dense area are screened to obtain the navigation task plan to be adjusted;

[0187] Perform path complexity analysis on the navigation task plan to be adjusted, split and adjust the tasks, and generate an optimized task plan.

[0188] It should be noted that the time period is a preset time period. In the case of screening out the dense area of ​​deviation points in the point cloud map, the number of deviation points in each small area can be counted in the point cloud map, and small areas with a larger number can be screened out as dense areas. In the case of recording the corresponding position points in the point cloud map and marking them as deviation points, based on the embodiment of the present invention, distribution statistics can also be performed based on the basic visual map.

[0189] It is worth mentioning that in the process of navigation analysis, in some complex navigation task paths, due to the complexity of the tasks, multiple target AVG forklifts may easily have high-frequency positioning deviation problems in a certain area or a certain path when navigating. Therefore, the present invention counts the deviation points of the target storage area based on the map model within a time period, screens out dense areas with more deviation points in the statistical process, screens out and marks the corresponding navigation tasks in the dense areas, performs complexity analysis and task splitting on the corresponding navigation tasks, thereby simplifying the navigation tasks accordingly or adjusting the path planning, effectively reducing route deviations.

[0190] The navigation task plan to be adjusted is analyzed for path complexity and tasks are split and adjusted to generate an optimized task plan. Specifically, the path location, path length, and regional environmental conditions are analyzed comprehensively. In the subsequent task splitting and corresponding task adjustments are performed on the complex navigation task. For example, a long forklift navigation loading and unloading task is split into multiple tasks, or a complex task is split into multiple subtasks and assigned to multiple AVG forklifts for collaborative operation. The optimized task plan records the collaborative operation plan after optimization and adjustment. Through the present invention, the efficiency reduction and safety problems caused by positioning deviation are effectively reduced.

[0191] The third aspect of the present invention also provides a computer-readable storage medium, which includes an AGV multi-vehicle intelligent collaborative operation program based on the Internet of Things. When the AGV multi-vehicle intelligent collaborative operation program based on the Internet of Things is executed by a processor, the steps of the AGV multi-vehicle intelligent collaborative operation method based on the Internet of Things as described in any one of the above items are implemented.

[0192] The present invention discloses an AGV multi-vehicle intelligent collaborative operation method, system and medium based on the Internet of Things. A laser positioning device and a visual device are used to construct a 3D point cloud map and a basic visual map, and the laser and visual positioning information of the AVG forklift is collected through the Internet of Things. The navigation analysis platform controls the AVG forklift in real time, performs positioning analysis and calculates the deviation value according to the navigation mode (laser or vision). When the deviation exceeds the threshold, the system uses the information of the corresponding module for real-time correction. At the same time, the system records the forklift positioning deviation, marks the deviation-intensive area and adjusts the task plan to ensure accurate navigation and efficient warehousing operations. The present invention makes full use of different navigation modes. When performing subsequent positioning analysis, the positioning data is judged and analyzed for deviation based on navigation analysis of different modes and different map analysis, thereby improving positioning accuracy and navigation safety, and at the same time improving driving safety in complex dynamic environments.

[0193] In the several embodiments provided in the present application, it should be understood that the disclosed devices and methods can be implemented in other ways. The device embodiments described above are only schematic. For example, the division of the units is only a logical function division. There may be other division methods in actual implementation, such as: multiple units or components can be combined, or can be integrated into another system, or some features can be ignored or not executed. In addition, the coupling, direct coupling, or communication connection between the components shown or discussed can be through some interfaces, and the indirect coupling or communication connection of the devices or units can be electrical, mechanical or other forms.

[0194] The units described above as separate components may or may not be physically separated, and the components displayed as units may or may not be physical units; they may be located in one place or distributed on multiple network units; some or all of the units may be selected according to actual needs to achieve the purpose of the present embodiment.

[0195] In addition, all functional units in the embodiments of the present invention may be integrated into one processing unit, or each unit may be separately used as a unit, or two or more units may be integrated into one unit; the above-mentioned integrated units may be implemented in the form of hardware or in the form of hardware plus software functional units.

[0196] A person of ordinary skill in the art can understand that: all or part of the steps of implementing the above method embodiment can be completed by hardware related to program instructions, and the aforementioned program can be stored in a computer-readable storage medium. When the program is executed, it executes the steps of the above method embodiment; and the aforementioned storage medium includes: a mobile storage device, a read-only memory (ROM), a random access memory (RAM), a magnetic disk or an optical disk, and other media that can store program codes.

[0197] Alternatively, if the above-mentioned integrated unit of the present invention is implemented in the form of a software function module and sold or used as an independent product, it can also be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the embodiment of the present invention can be essentially or partly reflected in the form of a software product that contributes to the prior art. The computer software product is stored in a storage medium and includes several instructions for a computer device (which can be a personal computer, server, or network device, etc.) to execute all or part of the methods described in each embodiment of the present invention. The aforementioned storage medium includes: various media that can store program codes, such as mobile storage devices, ROM, RAM, magnetic disks or optical disks.

[0198] The above is only a specific embodiment of the present invention, but the protection scope of the present invention is not limited thereto. Any person skilled in the art can easily think of changes or substitutions within the technical scope disclosed by the present invention, which should be included in the protection scope of the present invention. Therefore, the protection scope of the present invention should be based on the protection scope of the claims.

Claims

1. An AGV multi-vehicle intelligent collaborative operation method based on the Internet of Things, characterized in that: include: S1: Based on the preset laser positioning device, the laser signal is emitted and collected in the target storage area, and a 3D point cloud map is constructed based on the laser signal; S2: Based on the preset visual device, multi-angle image data of the target storage area is collected and a basic visual map is constructed, and the 3D point cloud map is associated with the basic visual map by map path; S3: Based on the laser module and the visual module, real-time positioning data is collected from multiple target AVG forklifts, and laser positioning information and visual positioning information are obtained. All positioning information is uploaded to the navigation analysis platform through the Internet of Things. S4: through the preset task plan, multiple target AVG forklifts are navigated and controlled in real time, and the navigation mode is switched in real time in different navigation tasks, and the current navigation mode is judged. If the current navigation mode is laser module navigation, the laser positioning information collected in real time is imported into the point cloud map for positioning analysis, and the second positioning information in the corresponding basic visual map is obtained through map association, and the positioning deviation is calculated based on the laser positioning information and the second positioning information to obtain a first deviation value. The visual positioning information is collected in real time based on the visual module, and the positioning deviation is calculated based on the visual positioning information and the second positioning information to obtain a second deviation value. If the first and second deviation values ​​are both greater than the threshold, the laser positioning information is corrected in real time through the visual positioning information, and the real-time navigation of the target AVG forklift is performed through the corrected positioning; S5: Determine the current navigation mode. If the current navigation mode is visual module navigation, import the real-time collected visual positioning information into the basic visual map for positioning analysis and positioning deviation value calculation in the same way as S4, and correct the visual positioning information in real time through the actual laser positioning information; S6: Within a time period, the positioning deviations of multiple target AVG forklifts are recorded, all deviation points are counted in the point cloud map, and the distribution information of the deviation points is obtained. The dense areas of the deviation points are marked according to the distribution information of the deviation points and the preset task plan is adjusted; Among them, based on the preset visual device, multi-angle image data of the target storage area is collected and a basic visual map is constructed, and the 3D point cloud map is associated with the basic visual map by map path, specifically: Through the preset visual device, multi-angle image data of the target storage area is collected to obtain multi-angle area image data; Build a basic model based on the area and contour information of the target storage area; Based on the multi-angle regional image data, paths and objects are identified and marked, and based on the identification results, a basic model is filled to form a basic visual map; Mark the buildings and fixed equipment in the point cloud map as key points, and obtain corresponding location information to form first key point information; Based on the first key point information, marking and recording the position information of the corresponding key point from the basic visual map to form the second key point information; Based on the 3D point cloud map and the basic visual map, the first key point information and the second key point information are mapped with respect to position information association, and key point association information is formed; Extracting path information from the 3D point cloud map and the basic visual map to obtain first path information and second path information respectively; Performing location information association mapping on the first path information and the second path information, and forming path point association information; The key point association information and the path point association information are integrated to form map path association data.

2. The method for intelligent collaborative operation of multiple AGVs based on the Internet of Things according to claim 1 is characterized in that: Based on the preset laser positioning device, the laser signal is emitted and collected in the target storage area, and a 3D point cloud map is constructed based on the laser signal, specifically: Determine the target storage area and initialize the scanning parameters of the preset laser positioning device; Scan the target storage area at multiple angles and collect spatial coordinate point data through a preset laser positioning device; Constructing an original point cloud map through the spatial coordinate point data; Identify and mark fixed equipment, paths, and buildings in the original point cloud map to form a complete 3D point cloud map.

3. The method for intelligent collaborative operation of multiple AGVs based on the Internet of Things according to claim 1 is characterized in that: The S4 step is specifically as follows: Obtain multiple navigation tasks based on the preset task plan, select the navigation mode by analyzing the distance, accuracy, and route running time of the navigation task, and switch the navigation mode in real time during the operation of the preset task plan; In a preset judgment cycle, the current navigation mode is judged. If the current navigation mode is laser module navigation, the real-time collected laser positioning information and the task navigation route are obtained, and the laser positioning information is imported into the point cloud map for positioning analysis. The laser positioning information is positioned and associated through the path point association information in the map path association data, and the associated positioning in the basic visual map is obtained to form the second positioning information; Get multiple key points based on map path association data; Using the laser positioning information, respectively calculating the coordinate distances between the laser positioning position point and a plurality of key points in the point cloud map to form a first set of distance data; Based on the second positioning information, respectively calculating the coordinate distances between the second positioning point and the plurality of key points in the basic visual map to form a second set of distance data; Perform vector conversion on the first set of distance data and the second set of distance data to form two vectors, calculate the distance between the two vectors based on the standard Euclidean distance, and use the calculation result as the first deviation value; Based on the visual module, the visual positioning information is collected in real time, and the coordinate deviation based on the basic visual map is calculated by using the visual positioning information and the second positioning information to obtain a second deviation value; Determine whether the first deviation value and the second deviation value are both greater than a preset threshold value. If so, analyze the location point of the visual positioning corresponding to the point cloud map based on the visual positioning information and the map path association data to obtain the third positioning information; The laser positioning information is corrected in real time by the third positioning information, and the target AVG forklift is navigated in real time after the positioning is corrected.

4. The method for intelligent collaborative operation of multiple AGVs based on the Internet of Things according to claim 3 is characterized in that: The step S5 is specifically as follows: In a preset judgment cycle, the current navigation mode is judged. If the current navigation mode is visual module navigation, the real-time collected visual positioning information and the task navigation route are obtained, and the visual positioning information is imported into the basic visual map for positioning analysis. The visual positioning information is positioned and associated through the path point association information in the map path association data, and the associated positioning in the point cloud map is obtained to form the second positioning information; Get multiple key points based on map path association data; By using the visual positioning information, respectively calculating the coordinate distances between the visual positioning position point and a plurality of key points in the basic visual map to form a first set of distance data; Based on the second positioning information, respectively calculating the coordinate distances between the second positioning point and the plurality of key points in the point cloud map to form a second set of distance data; Perform vector conversion on the first set of distance data and the second set of distance data to form two vectors, calculate the distance between the two vectors based on the standard Euclidean distance, and use the calculation result as the first deviation value; Based on the real-time collection of laser positioning information by the laser module, coordinate deviation calculation based on the point cloud map is performed through the laser positioning information and the second positioning information to obtain a second deviation value; Determine whether the first deviation value and the second deviation value are both greater than a preset threshold value. If so, analyze the location point of the laser positioning corresponding to the basic visual map based on the laser positioning information and the map path association data to obtain the third positioning information; The visual positioning information is corrected in real time through the third positioning information, and the real-time navigation of the target AVG forklift is performed after the positioning is corrected.

5. The method for intelligent collaborative operation of multiple AGVs based on the Internet of Things according to claim 4 is characterized in that: In a time period, multiple target AVG forklifts are positioned and deviated, all deviation points are counted in the point cloud map, and the deviation point distribution information is obtained. The dense area of ​​the deviation points is marked and the preset task plan is adjusted according to the deviation point distribution information, specifically: In a period of time, multiple target AVG forklifts are analyzed for real-time positioning deviation. If positioning correction occurs, the corresponding position point is recorded in the point cloud map and marked as a deviation point; Generate deviation point distribution information based on all deviation points; By using the deviation point distribution information, a deviation point dense area is screened out in the point cloud map, and based on the preset task plan, the navigation path tasks passing through the dense area are screened to obtain the navigation task plan to be adjusted; Perform path complexity analysis on the navigation task plan to be adjusted, split and adjust the tasks, and generate an optimized task plan.

6. An AGV multi-vehicle intelligent collaborative operation system based on the Internet of Things, characterized in that: The system includes: a memory and a processor, wherein the memory includes an AGV multi-vehicle intelligent collaborative operation program based on the Internet of Things, and when the AGV multi-vehicle intelligent collaborative operation program based on the Internet of Things is executed by the processor, the following steps are implemented: S1: Based on the preset laser positioning device, the laser signal is emitted and collected in the target storage area, and a 3D point cloud map is constructed based on the laser signal; S2: Based on the preset visual device, multi-angle image data of the target storage area is collected and a basic visual map is constructed, and the 3D point cloud map is associated with the basic visual map by map path; S3: Based on the laser module and the visual module, real-time positioning data is collected from multiple target AVG forklifts, and laser positioning information and visual positioning information are obtained. All positioning information is uploaded to the navigation analysis platform through the Internet of Things. S4: through the preset task plan, multiple target AVG forklifts are navigated and controlled in real time, and the navigation mode is switched in real time in different navigation tasks, and the current navigation mode is judged. If the current navigation mode is laser module navigation, the laser positioning information collected in real time is imported into the point cloud map for positioning analysis, and the second positioning information in the corresponding basic visual map is obtained through map association, and the positioning deviation is calculated based on the laser positioning information and the second positioning information to obtain a first deviation value. The visual positioning information is collected in real time based on the visual module, and the positioning deviation is calculated based on the visual positioning information and the second positioning information to obtain a second deviation value. If the first and second deviation values ​​are both greater than the threshold, the laser positioning information is corrected in real time through the visual positioning information, and the real-time navigation of the target AVG forklift is performed through the corrected positioning; S5: Determine the current navigation mode. If the current navigation mode is visual module navigation, import the real-time collected visual positioning information into the basic visual map for positioning analysis and positioning deviation value calculation in the same way as S4, and correct the visual positioning information in real time through the actual laser positioning information; S6: Within a time period, the positioning deviations of multiple target AVG forklifts are recorded, all deviation points are counted in the point cloud map, and the distribution information of the deviation points is obtained. The dense areas of the deviation points are marked according to the distribution information of the deviation points and the preset task plan is adjusted; Among them, based on the preset visual device, multi-angle image data of the target storage area is collected and a basic visual map is constructed, and the 3D point cloud map is associated with the basic visual map by map path, specifically: Through the preset visual device, multi-angle image data of the target storage area is collected to obtain multi-angle area image data; Build a basic model based on the area and contour information of the target storage area; Based on the multi-angle regional image data, paths and objects are identified and marked, and based on the identification results, a basic model is filled to form a basic visual map; Mark the buildings and fixed equipment in the point cloud map as key points, and obtain corresponding location information to form first key point information; Based on the first key point information, marking and recording the position information of the corresponding key point from the basic visual map to form the second key point information; Based on the 3D point cloud map and the basic visual map, the first key point information and the second key point information are mapped with respect to position information association, and key point association information is formed; Extracting path information from the 3D point cloud map and the basic visual map to obtain first path information and second path information respectively; Performing location information association mapping on the first path information and the second path information, and forming path point association information; The key point association information and the path point association information are integrated to form map path association data.

7. A computer-readable storage medium, characterized in that: The computer-readable storage medium includes an AGV multi-vehicle intelligent collaborative operation program based on the Internet of Things. When the AGV multi-vehicle intelligent collaborative operation program based on the Internet of Things is executed by a processor, the steps of the AGV multi-vehicle intelligent collaborative operation method based on the Internet of Things as described in any one of claims 1 to 5 are implemented.

Images