AGV robot scheduling method and system of intelligent factory and storage medium

Abstract

The invention relates to the technical field of AGV robot scheduling and control, and discloses an AGV robot scheduling method and system of an intelligent factory and a storage medium. The method comprises the steps of collecting bearing sensing data to identify a product height value according to a product carrying instruction and a transportation route, and if the value is higher than a reference value, generating an auxiliary scheduling instruction; next, nearby non-loaded AGVs are searched to serve as temporary robots, scheduling values are obtained in combination with task load and distance sorting of the temporary robots, and the optimal robot is selected to serve as a slave robot to be paired with the original AGV robot to jointly bear products; and finally, the master and slave robots respectively respond to the matched scheduling instructions. According to the method, through master-slave cooperation and accurate scheduling, the AGV transportation stability and efficiency are improved.

Description

Technical Field

[0001] This application relates to the technical field of AGV robot scheduling and control, and in particular to an AGV robot scheduling method, system and storage medium for a smart factory. Background Technology

[0002] AGV robots are core equipment in the material handling process of factory production, undertaking the automated transfer of raw materials, semi-finished products, and finished products. Through preset paths or autonomous navigation, they can move accurately and efficiently between various processes in the workshop, reducing human intervention, avoiding material transportation delays, ensuring the continuity and stability of the production process, significantly improving the factory's production efficiency and intelligence level, and serving as a key support for smart factories to achieve flexible production.

[0003] In existing packaging workshops, corrugated cardboard raw materials and finished corrugated boxes are relatively large, and manual transportation is limited by handling capacity and space, resulting in small quantities transported at a time. This leads to increased transportation frequency, consuming a large amount of manpower and easily causing low transportation efficiency due to human factors, making it difficult to meet the material turnover needs of large-scale production. To address this, the industry has widely adopted AGV robots to replace manual labor, enabling large-volume transportation of cardboard raw materials or finished cardboard boxes and improving transportation efficiency.

[0004] However, when using AGV robots to transport cardboard raw materials or finished cardboard boxes, these items need to be stacked to increase transport capacity, but excessive stacking can shift the center of gravity upwards. During AGV robot acceleration, deceleration, or turning, the items are prone to swaying due to inertia, and in severe cases, they may fall, causing not only material loss but also potentially affecting workshop equipment operation and personnel safety. Therefore, there is an urgent need for a scheduling and control method that can improve the stability of AGV robot transportation. Summary of the Invention

[0005] To improve the stability of AGV robots transporting packaged products, this application provides an AGV robot scheduling method, system, and storage medium for smart factories.

[0006] Firstly, this application provides a method for scheduling AGV robots in a smart factory, employing the following technical solution:

[0007] A method for scheduling AGV robots in a smart factory includes the following steps:

[0008] Based on product handling instructions and transportation routes, the AGV robot's load-bearing sensor data is collected, and the product height value is identified from the load-bearing sensor data; if the product height value is higher than the preset reference height value, an auxiliary scheduling instruction is generated.

[0009] In response to the auxiliary scheduling command, search for other AGV robots whose shortest distance to the transportation route is within a preset first distance range; read the working status of the other AGV robots, and mark the AGV robots whose working status is unloaded as temporary robots;

[0010] Obtain the task list of the temporary robot, sort the temporary robots according to the number of tasks in the task list to obtain a first scheduling value; calculate the temporary distance value between the temporary robot and the AGV robot carrying the product, sort the temporary robots according to the temporary distance value to obtain a second scheduling value; calculate the temporary scheduling value according to the first scheduling value and the second scheduling value, and sort the robots according to the temporary scheduling value to obtain the optimal robot;

[0011] The optimal robot is paired and bound to the AGV robot carrying the product. The optimal robot is controlled to move to the AGV robot carrying the product and jointly carry the product. The AGV robot carrying the product is the master robot and the optimal robot is the slave robot.

[0012] According to the controlled instructions of the master robot, the master scheduling instructions and slave scheduling instructions are matched from the preset scheduling instruction library. The master robot responds to the master scheduling instructions, and the slave robot responds to the slave scheduling instructions.

[0013] By adopting the above technical solution and using master-slave AGV pairing for transport, the risk of unstable center of gravity of stacked materials is effectively reduced. First, auxiliary scheduling is triggered based on the product height to accurately select suitable temporary robots; then, the optimal slave robot is determined through a dual sorting based on task volume and distance, ensuring collaborative efficiency. The master and slave robots move collaboratively according to the matched instructions, reducing the inertial effects of acceleration, deceleration, and turning, and preventing materials from shaking and falling. This improves transportation stability, adapts to the needs of large-scale production turnover, and reduces material loss and safety hazards.

[0014] Optionally, the step of the master robot responding to the master scheduling command and the slave robot responding to the slave scheduling command further includes the following sub-steps:

[0015] Based on the main scheduling command, the main robot calculates motion data, which includes motion direction, motion speed, and motion acceleration.

[0016] The direction coefficient is matched from a preset direction symbol library according to the direction of motion;

[0017] The speed coefficient is calculated based on the movement speed and the preset reference speed;

[0018] The acceleration coefficient is calculated based on the motion acceleration and the preset reference acceleration;

[0019] The height adjustment coefficient is calculated based on the direction coefficient, the velocity coefficient, and the acceleration coefficient.

[0020] The height of the master robot can be adjusted inversely based on the height adjustment coefficient, or the height of the slave robot can be adjusted in a positive correlation based on the height adjustment coefficient, or the height of the master robot can be adjusted inversely based on the height adjustment coefficient while the height of the slave robot is adjusted in a positive correlation.

[0021] By adopting the above technical solution, stability is further improved through dynamic adjustment of the height of the master and slave robots. Based on the master robot's motion data, a height adjustment coefficient is calculated by combining direction, speed, and acceleration coefficients. The height of the master robot is adjusted in reverse, the height of the slave robot is adjusted in forward, or both are adjusted in tandem to adapt to the motion state in real time. This can dynamically balance the center of gravity of stacked materials, reduce inertial swaying during acceleration, deceleration, and turning, further reduce the risk of falling, and ensure a more stable and reliable transportation process.

[0022] Optionally, the step of matching the master scheduling instruction and the slave scheduling instruction from a preset scheduling instruction library based on the master robot's controlled instructions further includes the following sub-steps:

[0023] The master robot sends the controlled instruction to the background database and queries the background database for the master scheduling instruction and the slave scheduling instruction corresponding to the controlled instruction; the master robot receives the master scheduling instruction and the slave scheduling instruction sent by the background database; the master robot executes the master scheduling instruction and sends the slave scheduling instruction to the slave robot; the slave robot receives and executes the slave scheduling instruction.

[0024] Alternatively, the master robot sends the controlled instruction and the communication address of the slave robot to the backend. The backend queries the backend database for the master scheduling instruction and the slave scheduling instruction corresponding to the controlled instruction. The backend returns the master scheduling instruction to the master robot and sends the slave scheduling instruction to the slave robot. The master robot receives and executes the master scheduling instruction, and the slave robot receives and executes the slave scheduling instruction.

[0025] By adopting the above technical solution, the reliability of collaboration is improved by clarifying the master-slave robot command interaction logic. Both command transmission methods rely on the backend database to accurately match commands, ensuring synchronized actions between the master and slave robots. The former uses the master robot as an intermediary for command transmission, while the latter distributes commands directly from the backend, both ensuring efficient command delivery. This avoids command delays or mismatches, ensuring precise collaborative actions between the master and slave robots, and further improving transportation stability and response efficiency.

[0026] Optionally, the step of collecting the load-bearing sensor data of the AGV robot and identifying the product height value from the load-bearing sensor data further includes the following sub-steps:

[0027] The carrying sensing data is obtained from the image sensor at a preset position on the AGV robot, and the image data of the product to be carried is obtained based on the image sensor.

[0028] The image recognition algorithm identifies the pixel coordinates of the product in the image data.

[0029] Obtain the real-time position and shooting angle of the image sensor;

[0030] The product height is calculated based on the real-time location, shooting angle, and pixel coordinates.

[0031] or,

[0032] The load-bearing sensing data is from the distance sensor on the top or side of the AGV robot. The distance sensor emits continuous detection signals as it rotates from the bottom to the top of the stacked load-bearing products.

[0033] The feedback signal of the continuous detection signal is analyzed to obtain the initial feedback signal and the abrupt feedback signal;

[0034] The product height value is calculated based on the time and phase of the mutation feedback signal, as well as the time and phase of the initial feedback signal.

[0035] By adopting the above technical solutions, the product height value can be accurately obtained by using an image sensor combined with an image recognition algorithm or by using a distance sensor to dynamically scan and analyze feedback signals.

[0036] Optionally, the step of calculating the temporary scheduling value based on the first scheduling value and the second scheduling value further includes the following steps:

[0037] The sequence number of the temporary robot in the first sorting is positively correlated with the number of tasks in the task list. The more tasks there are, the larger the sequence number in the first sorting, and the fewer tasks there are, the smaller the sequence number in the first sorting.

[0038] The sequence number of the temporary robot in the second sorting is positively correlated with the temporary distance value. The larger the temporary distance value, the larger the sequence number in the second sorting; the smaller the temporary distance value, the smaller the sequence number in the second sorting.

[0039] The first scheduling value and the second scheduling value are used to calculate the temporary scheduling value using a weighted average algorithm; the optimal robot is the temporary robot with the smallest temporary scheduling value.

[0040] By adopting the above technical solution, and through dual-dimensional sorting and weighted calculation, the scheduling priority of temporary robots is scientifically quantified. Robots with fewer tasks and shorter distances receive lower scheduling values ​​and participate in collaboration first, ensuring the efficient completion of the current transportation task while reducing interference with other tasks. Dynamically balancing resource utilization and response efficiency improves overall scheduling flexibility and system stability, achieving optimized configuration for multi-robot collaboration.

[0041] Optionally, the step of calculating the temporary scheduling value using a weighted average algorithm based on the first scheduling value and the second scheduling value further includes the following steps:

[0042] The weight of the first scheduling value is adjusted inversely based on the average number of tasks in the task lists of all the temporary robots; the smaller the average number, the greater the weight of the first scheduling value; the greater the average number, the smaller the weight of the first scheduling value.

[0043] The weight of the second scheduling value is adjusted according to the positive correlation between the dispersion value of the temporary distance values ​​of all the temporary robots and the dispersion value; the smaller the dispersion value, the smaller the weight of the second scheduling value; the larger the dispersion value, the larger the weight of the second scheduling value.

[0044] By adopting the above technical solution, intelligent weighting is achieved through dynamic adjustment of weights: when the workload of temporary robots is generally small, the workload weight is increased to prioritize the release of idle resources; when the distance distribution is relatively scattered, the distance weight is increased to filter neighboring robots. Resource utilization and response speed are dynamically balanced according to the real-time scheduling environment, avoiding decision-making biases caused by a single indicator.

[0045] Optionally, the step of controlling the optimal robot to move to the AGV robot carrying the product and jointly carrying the product includes the following sub-steps:

[0046] The lifting and load-bearing mechanism bears the load as the robot moves to the first position next to the center of the pallet carrying the product.

[0047] The main robot moves from the center of the pallet carrying the product to a second position next to the center of the pallet. The first position and the second position are located on both sides of the center of the pallet and are centrally symmetrical. The first position and the second position are located on the transport route.

[0048] By adopting the above technical solution, the master and slave robots share the load symmetrically on both sides of the pallet's center, ensuring that the center of gravity of the stacked products falls within the symmetrical center area supported by both robots, effectively preventing center of gravity shift. Simultaneously, both robots are located on the transport route, ensuring balanced force during coordinated movement, further reducing swaying during acceleration, deceleration, and turning, lowering the risk of falling, and improving transport stability.

[0049] Optionally, the step of controlling the optimal robot to move to the AGV robot carrying the product and jointly carrying the product includes the following sub-steps:

[0050] Obtain the dimensions of the master robot and the slave robot;

[0051] The aspect ratio of the main robot is calculated based on its dimensions, and the aspect ratio of the slave robot is calculated based on its dimensions.

[0052] The average aspect ratio is calculated based on the main aspect ratio and the secondary aspect ratio;

[0053] If the average aspect ratio is greater than the preset aspect ratio reference value, the main robot and the slave robot will adjust their length direction to be perpendicular to the forward direction of their current position on the transportation route, and move along the forward direction.

[0054] Calculate the average ratio of the aspect ratio to the preset aspect ratio reference value, and adjust the spacing between the first position and the second position according to the positive correlation of the average ratio.

[0055] By adopting the above technical solution, the collaborative posture and spacing of the robots are dynamically adjusted based on their aspect ratio: when the aspect ratio is too large, the robot's length direction is perpendicular to the forward direction to reduce the risk of lateral swaying; at the same time, the spacing is increased according to the ratio to enhance support stability. This optimizes the spatial layout of the master and slave robots and improves the center of gravity control capability.

[0056] Secondly, this application provides an AGV robot scheduling system for a smart factory, which adopts the following technical solution:

[0057] An AGV robot scheduling system for a smart factory includes a processor, wherein the processor performs the steps of the AGV robot scheduling method for a smart factory as described in any of the preceding claims.

[0058] Thirdly, this application provides a storage medium, which adopts the following technical solution:

[0059] A storage medium storing a program that, when executed by a processor, implements the steps of the AGV robot scheduling method for a smart factory as described above.

[0060] In summary, this application includes at least one of the following beneficial technical effects:

[0061] By using mechanisms such as symmetrical load-bearing by master and slave robots and dynamic height adjustment, the center of gravity of stacked materials is balanced, reducing the risk of swaying and falling during acceleration, deceleration and turning, thereby reducing material loss and safety hazards.

[0062] Precise scheduling is triggered based on product height value. The optimal slave robot is selected by weighting the task volume and distance. The weights are adaptively adjusted to balance the current task and the overall resources, thereby improving the speed of collaborative response.

[0063] It is designed to transport large-volume materials such as corrugated cardboard, and uses sensing technology to accurately identify height, posture and spacing for dynamic adjustment. It adapts to different stacking states and workshop environments, supporting the needs of large-scale production turnover. Attached Figure Description

[0064] Figure 1 This is a flowchart illustrating the steps of an AGV robot scheduling method in a smart factory.

[0065] Figure 2 This is a flowchart illustrating the steps of the master robot responding to the master scheduling command and the slave robot responding to the slave scheduling command.

[0066] Figure 3 This diagram illustrates the steps of controlling the optimal robot to move to the AGV robot carrying the product and jointly carrying the product. Detailed Implementation

[0067] The embodiments of this application are described in detail below, and examples of the embodiments are shown in the accompanying drawings.

[0068] In the description of this specification, the references to "certain embodiments," "one embodiment," "some embodiments," "illustrative embodiment," "example," "specific example," or "some examples" refer to specific features, structures, materials, or characteristics described in connection with the described embodiment or example, which are included in at least one embodiment or example of this application. In this specification, the illustrative expressions of the above terms do not necessarily refer to the same embodiment or example. Furthermore, the specific features, structures, materials, or characteristics described may be combined in any suitable manner in one or more embodiments or examples.

[0069] This application discloses an AGV robot scheduling method for a smart factory, aiming to solve the problem of swaying and falling when existing AGV robots transport stacked materials (such as corrugated cardboard and corrugated boxes) due to the upward shift of the center of gravity, thereby improving transportation stability and production efficiency. (Refer to...) Figure 1 Specifically, it includes the following steps:

[0070] When the AGV robot receives a product handling instruction, it plans an initial transport path based on a pre-set transport route. Simultaneously, sensors deployed on the AGV robot, such as image sensors and distance sensors, collect data on the product to be transported on the platform. This data is then analyzed to identify the actual stacking height of the product, i.e., the product height. The identified product height is compared with a pre-set reference height: if the product height is lower than or equal to the reference height, the current stacking is stable, and the AGV robot transports the product independently along the conventional path; if the product height is higher than the reference height, a risk of instability is identified, and an auxiliary scheduling instruction is automatically generated, triggering a collaborative transport mechanism. The pre-set reference height is determined based on the product material and stacking stability test results, such as the critical safe stacking height for corrugated cardboard.

[0071] In response to the aforementioned auxiliary scheduling instructions, initiate the search and selection process for nearby AGV robots:

[0072] Based on the current transportation route, a preset first distance range is defined, such as within 5 meters on both sides of the transportation route. This range is dynamically adjusted according to the workshop layout to search for other AGV robots within this range.

[0073] Read the real-time working status of the searched AGV robots, including load status and task execution progress, filter out AGV robots with a working status of "unloaded" and no urgent tasks, and mark them as temporary robots, i.e. potential collaborators.

[0074] To ensure efficient collaboration and rational use of resources, temporary robots are prioritized, specifically including:

[0075] First Dimension Sorting - Task Quantity Dimension: Obtain the task list of each temporary robot and count the number of tasks to be executed; sort the temporary robots according to the number of tasks. The fewer the number of tasks, the higher the ranking. Convert the sorting result into the first scheduling value. The smaller the value, the higher the priority.

[0076] Second-dimensional sorting - distance dimension: The positioning system calculates the straight-line distance between each temporary robot and the AGV robot currently carrying the product, i.e., the temporary distance value; a second sorting is performed based on the temporary distance value, the closer the distance, the higher the ranking, which is converted into a second scheduling value, the smaller the value, the higher the priority;

[0077] Comprehensive ranking: The first and second scheduling values ​​are calculated using a weighted average algorithm to obtain the temporary scheduling value. The weights can be dynamically adjusted according to the real-time load of the workshop. The temporary robots are sorted in ascending order of their temporary scheduling values, and the temporary robot at the top of the ranking is selected as the optimal robot.

[0078] The optimal robot is paired with the AGV robot currently carrying the product and a binding relationship is established. The AGV robot carrying the product is the master robot, and the optimal robot is the slave robot.

[0079] The slave robot moves along the shortest path to the location of the master robot, and the two work together to switch loads. The slave robot moves to the preset load-bearing position on the product tray and lifts the load-bearing mechanism, sharing the weight of the product with the master robot. The master robot then adjusts its position to balance the center of gravity, ensuring that the force is evenly distributed between the two. At this point, the product is supported by the master and slave robots working together to form a stable dual-machine support structure.

[0080] After the master robot generates controlled instructions based on transportation requirements, it matches the corresponding master scheduling instructions and slave scheduling instructions from a pre-set scheduling instruction library that stores cooperative motion parameters for different scenarios:

[0081] The main scheduling command contains the motion parameters of the main robot, such as driving speed and turning angle;

[0082] The slave robot keeps synchronized with the master robot to ensure that the slave robot's movement trajectory and acceleration / deceleration rhythm are completely matched with the master robot. For example, when the master robot decelerates, the slave robot decelerates synchronously in the same proportion.

[0083] The master robot and slave robot respond to corresponding commands to achieve coordinated movement: during acceleration, deceleration, turning and other actions, the two maintain consistent motion to avoid swaying of the product being carried due to relative displacement.

[0084] Using the above methods, AGV robots can dynamically trigger collaborative transportation based on the stacking height of the products they carry. Through scientific screening and collaborative control, the risk of instability of the center of gravity of highly stacked materials can be effectively reduced, material falling and loss during transportation can be reduced, and the turnover efficiency of workshop materials can be improved, meeting the large-scale and highly stable production needs of smart factories.

[0085] Reference Figure 2 To further improve transportation stability during the process of the master robot responding to the master scheduling command and the slave robot responding to the slave scheduling command, this embodiment introduces a dynamic height adjustment mechanism, which specifically includes the following sub-steps:

[0086] After receiving the master scheduling command, the master robot parses the motion parameters in the command and calculates the motion data at the current moment, specifically including:

[0087] Direction of movement: The current direction of the robot is determined by the navigation system and expressed in angle values, such as 0° representing due east and 90° representing due north;

[0088] Movement speed: Instantaneous travel speed calculated based on the speed parameters in the command and the actual load conditions, in m / s;

[0089] Motion acceleration: The acceleration value calculated based on the rate of change of velocity, in m / s². 2 It is used to characterize how fast a change in velocity occurs.

[0090] The system queries the preset direction symbol library for the direction coefficient corresponding to the current direction of motion. The direction symbol library is a pre-established mapping table that maps different ranges of motion directions to corresponding coefficient values, as detailed below:

[0091] Straight-line driving scenario: When the direction of movement is 0° (due east), 90° (due north), 180° (due west), or 270° (due south), it is mapped to a direction coefficient +1.0, indicating that no additional center of gravity adjustment is required when driving in a straight line;

[0092] Left turn scenario: When the direction of movement is in the range of 45° to 135°, it is mapped to a direction coefficient of -0.5, which means that when turning left, the center of gravity needs to be adjusted to the right.

[0093] Right turn scenario: When the direction of movement is in the range of 225° to 315°, it is mapped to a direction coefficient of +0.5, which means that when turning right, the center of gravity needs to be adjusted to the left.

[0094] Special angle scenarios: For other angle values, the corresponding direction coefficients are calculated using a linear interpolation algorithm to ensure the continuity of coefficient changes.

[0095] The current velocity v is compared with the preset reference velocity v0, and the velocity coefficient kv is obtained by division: kv = v / v0;

[0096] When the speed of movement equals the reference speed, the speed coefficient is 1.0, indicating that no adjustment is needed;

[0097] When the speed of movement is greater than the reference speed, the speed coefficient is greater than 1.0, indicating that the effort to adjust the center of gravity needs to be increased;

[0098] When the speed of movement is less than the reference speed, the speed coefficient is less than 1.0, which means that the adjustment of the center of gravity can be appropriately reduced.

[0099] The current acceleration a is compared with the preset reference acceleration a0, and the acceleration coefficient ka is obtained by division: ka = a / a0;

[0100] When the acceleration is positive, it is in an acceleration state. The acceleration coefficient is positive, indicating that the center of gravity needs to be adjusted forward.

[0101] When the acceleration is negative, it is in a deceleration state. A negative acceleration coefficient indicates that the center of gravity needs to be adjusted backward.

[0102] The larger the absolute value of the acceleration, the larger the absolute value of the coefficient, indicating that a stronger force is needed to adjust the center of gravity.

[0103] Multiply the direction coefficient kd, velocity coefficient kv, and acceleration coefficient ka to obtain the final height adjustment coefficient k: k = kd × kv × ka;

[0104] The height adjustment coefficient comprehensively reflects the influence of motion direction, speed and acceleration on the stability of the center of gravity. Its value can be positive or negative, and the larger the absolute value, the greater the adjustment required.

[0105] Based on the calculated height adjustment coefficient, the heights of the master robot and slave robot are dynamically adjusted. The adjustment methods are divided into three modes:

[0106] Independent adjustment mode for main robot: When the height adjustment coefficient is negative, the height of the main robot is reduced by using an inverse correlation adjustment method, such as reducing it by 5-10cm, while keeping the height of the main robot unchanged, so that the overall center of gravity is lowered and shifted to the side of the main robot.

[0107] Independent adjustment mode from the robot: When the height adjustment coefficient is positive, the height of the slave robot is increased by positive correlation adjustment, such as by 5-10cm, while keeping the height of the main robot unchanged, so that the overall center of gravity moves up and shifts to the side of the slave robot.

[0108] Cooperative adjustment mode: When the absolute value of the height adjustment coefficient is large, the height of the master robot and the slave robot are adjusted at the same time. That is, the height of the master robot is lowered while the height of the slave robot is raised, forming a cooperative adjustment effect of one lowering and one raising, which further enhances the center of gravity adjustment capability.

[0109] During height adjustment, the system ensures adjustment accuracy through closed-loop control: real-time monitoring of height sensor data, and when the actual adjustment reaches 95% of the target value, a fine-tuning mechanism is activated, ultimately controlling the adjustment error within ±0.5cm.

[0110] Through the aforementioned dynamic height adjustment mechanism, the AGV robot can automatically adjust its center of gravity position according to its real-time motion status during transportation: when turning, it shifts its center of gravity to the inside of the turn to reduce the risk of tipping over due to centrifugal force; when accelerating, it adjusts its center of gravity forward to prevent materials from tilting backward; and when decelerating, it adjusts its center of gravity backward to prevent materials from tilting forward.

[0111] In the step of matching master scheduling instructions and slave scheduling instructions from a preset scheduling instruction library according to the master robot's controlled instructions, this embodiment provides two instruction interaction methods to ensure the accuracy and reliability of the collaborative actions of the master and slave robots.

[0112] Method 1: Main robot relay mode:

[0113] In this mode, the master robot acts as a command relay hub, responsible for interacting with the backend database and distributing commands to the slave robots. The specific steps are as follows:

[0114] After receiving the initial controlled command, such as "transport to warehouse A," the main robot transmits the command to the backend database via the workshop's wireless communication network, such as Wi-Fi 6 or 5G. The command content includes key parameters such as task type, target location, and priority.

[0115] The backend database performs a matching query in a pre-defined scheduling instruction library based on the received controlled instructions. The scheduling instruction library uses a key-value pair structure for storage, with each controlled instruction corresponding to a set of predefined master and slave scheduling instructions, for example:

[0116] Main dispatch instructions include speed curves (e.g., "accelerate uniformly to 1.5m / s"), steering strategies (e.g., "right turn radius 2m"), and start / stop control parameters;

[0117] From the scheduling instructions: follow-up strategies synchronized with the master scheduling instructions (such as "maintain a 1m gap"), load distribution ratios (such as "bear 40% of the weight").

[0118] The backend database packages the matched master and slave scheduling commands into a single package and sends it to the master robot. After parsing the data packet, the master robot forwards the slave scheduling command to the slave robot via a dedicated communication channel, such as Bluetooth Mesh.

[0119] The master robot and slave robot execute their respective scheduling commands and ensure synchronized movements through a real-time status feedback mechanism, such as reporting their position and posture every 100ms. If a command execution deviation occurs, such as a position error exceeding 5cm, the system triggers a compensation algorithm for dynamic adjustment.

[0120] Method 2: Direct distribution mode from the backend:

[0121] In this mode, the backend database acts as the command center, directly distributing commands to the master and slave robots. The specific steps are as follows:

[0122] The master robot sends the controlled commands along with the slave robot's unique communication address (such as MAC address or IP address) to the backend. The communication address is automatically assigned and bound by the system during the pairing phase.

[0123] The backend database matches the corresponding master and slave scheduling instructions based on the controlled instructions, and constructs two independent communication paths based on the slave robot's communication address:

[0124] Main path: Send the main scheduling command to the main robot;

[0125] From the path: The dispatch command is sent directly to the slave robot via the workshop LAN (such as industrial Ethernet).

[0126] The master and slave robots synchronously receive their respective scheduling instructions and start execution synchronously according to preset timestamps (such as UTC time). The background continuously monitors the execution status of both robots and ensures that the time error of instruction execution is less than 10ms through a timestamp alignment algorithm (such as NTP protocol).

[0127] By standardizing the command interaction process between master and slave robots, the reliability of collaborative operations is enhanced. Both command transmission methods rely on a backend database to achieve accurate command matching, ensuring the consistency of master and slave robot actions. Specifically, both the master robot relay mode and the backend direct distribution mode enable efficient command transmission, effectively avoiding command delays or mismatches, ensuring accurate and error-free collaborative actions between master and slave robots, thereby improving transportation stability and response efficiency.

[0128] In the step of collecting load-bearing sensor data from the AGV robot and identifying the product height value from the load-bearing sensor data, this embodiment provides two accurate identification methods to adapt to different workshop environments and the characteristics of the load-bearing products.

[0129] Method 1: Image Sensing Recognition Method

[0130] This method uses an image sensor to acquire images of the product being carried and combines this with algorithmic analysis to achieve non-contact height measurement. The specific steps are as follows:

[0131] An industrial image sensor, such as a 2-megapixel industrial camera, equipped with a supplementary lighting module, is installed at a pre-defined position on the edge of the AGV robot's carrying platform, such as the center of the front end of the platform. Once the product is placed on the platform, the image sensor captures images of the stacked product from all angles at a preset frequency, such as 2 frames per second, obtaining complete image data including the top, bottom, and side contours of the product. During the acquisition process, the supplementary lighting module automatically adjusts the brightness according to the ambient light intensity to ensure image clarity, with pixel grayscale value fluctuations controlled within ±10.

[0132] The collected image data is input into a preset image recognition algorithm, which is based on a deep learning semantic segmentation model and has been trained with over 100,000 samples. The algorithm first preprocesses the image, including denoising and edge enhancement, and then identifies the pixel boundaries of the product stacking area; that is, the top edge pixel coordinate Y1 and the bottom edge pixel coordinate Y0, and automatically filters out interfering pixels from non-product areas such as trays and platforms.

[0133] The AGV robot uses a built-in positioning module, such as BeiDou + UWB fusion positioning, to obtain the real-time three-dimensional position coordinates (Xs, Ys, Zs) of the image sensor; at the same time, it uses an angle sensor to collect the sensor's shooting angle (including horizontal deflection angle α and vertical pitch angle β). The parameter acquisition frequency is synchronized with the image acquisition frequency to ensure spatiotemporal matching.

[0134] The product height is calculated based on the mapping relationship between pixel coordinates and physical dimensions, combined with sensor parameters.

[0135] Calculate the pixel height difference in the areas carrying stacked products in the image: ΔY = Y1 - Y0 (unit: pixels);

[0136] Based on the image sensor focal length f, the shooting distance (calculated from the real-time position), and the preset pixel physical size conversion factor k, k = 0.1 mm / pixel, the pixel height difference is converted into an initial physical height value through pre-calibration using a calibration board: H_initial = ΔY × k;

[0137] Introducing a shooting angle correction factor: When the vertical pitch angle β≠0, the product height value is finally obtained by correcting it through trigonometric functions, Hactual = Hinitial × cosβ.

[0138] Method 2: Distance sensing scanning method:

[0139] This method uses a distance sensor to dynamically scan the surface of the product and measures its height by utilizing signal feedback differences. It is suitable for workshops with high dust levels or complex lighting conditions. The specific steps are as follows:

[0140] A laser distance sensor is installed on the top edge or side column of the AGV robot's carrying platform, at a height of approximately 1.5m above the platform surface. The laser distance sensor is equipped with a rotary drive module, allowing it to rotate and scan vertically (0-90°). Once the product is placed, the laser distance sensor rotates from the bottom of the stacked product (close to the platform surface) towards the top at a speed of 5° / s, while continuously emitting a continuous laser detection signal (wavelength 650nm, power ≤1mW, compliant with safety standards).

[0141] The sensor has a built-in signal receiving module that receives feedback signals reflected from the surface of the product in real time. The system analyzes the feedback signals in real time.

[0142] Initial feedback signal: When the sensor scans to the bottom of the product, the amplitude and phase of the feedback signal tend to stabilize due to the flatness of the platform surface. Record the signal reception time t0 and phase value φ0 at this time.

[0143] Abrupt feedback signal: When the sensor scans to the top edge of the product, the amplitude of the feedback signal suddenly attenuates because it exceeds the surface of the product, and the phase value jumps (fluctuation > 5°). Record the signal reception time t1 and the phase value φ1 at this time.

[0144] Calculate the product height based on the time difference and phase difference of signal feedback:

[0145] Calculate the sensor rotation angle: Δθ = rotational speed × (t1 - t0), to obtain the vertical angle corresponding to the scanning path;

[0146] Based on the laser signal propagation speed (c=3×10) 8 Given the phase difference (Δφ = φ1 - φ0) and the phase difference (Δφ = φ1 - φ0), calculate the straight-line distance from the sensor to the top of the product: L = (Δφ / 360°) × λ / 2 (λ is the laser wavelength);

[0147] Based on the sensor installation height (Han, pre-calibrated) and geometric relationships, the product height value is obtained as: Hactual = Han - L × cosΔθ.

[0148] In the step of calculating the temporary scheduling value based on the first and second scheduling values, this embodiment uses a two-dimensional sorting rule and a weighted algorithm to quantitatively evaluate the temporary robot scheduling priority. The specific steps are as follows:

[0149] Using the number of tasks in the temporary robot's task list as the core evaluation metric, a positive correlation between the number of tasks and the sorting sequence number is established. The specific implementation is as follows:

[0150] The system obtains the list of tasks to be executed for each temporary robot through the workshop task management module, and counts the number of tasks, including the number of transportation instructions received but not executed, but excluding tasks currently being executed. For example, temporary robot A has 2 tasks in its task list, and temporary robot B has 5 tasks in its task list.

[0151] The temporary robots are sorted according to the rule that "the more tasks in the task list, the higher the initial sorting number":

[0152] Sort all temporary robots by the number of tasks in their task lists in ascending order, from fewest to most.

[0153] Assign a sequence number to the sorting results, i.e., the first sorting sequence number: the robot ranked 1st is numbered 1, the 2nd is numbered 2, and so on.

[0154] Example: If temporary robot A (2 tasks) ranks 3rd, its first ranking number is 3; if temporary robot B (5 tasks) ranks 8th, its first ranking number is 8.

[0155] The first sorting number is directly used as the first scheduling value. The smaller the value, the less negative the impact of the task volume on the scheduling priority. For example, the first scheduling value of robot A is 3, and the first scheduling value of robot B is 8.

[0156] Using the temporary distance between the temporary robot and the main robot as the core evaluation metric, a positive correlation mapping relationship is established between this distance and the ranking sequence number. The specific implementation is as follows:

[0157] The AGV robot's positioning system obtains the real-time 3D coordinates of each temporary robot and the main robot, and calculates the straight-line distance as the temporary distance value. For example, the distance between temporary robot A and the main robot is 3.2m, and the distance between temporary robot B and the main robot is 6.5m.

[0158] The temporary robots are sorted according to the rule that "the smaller the temporary distance value, the smaller the second sorting sequence number":

[0159] Sort the temporary distance values ​​of all temporary robots in ascending order, from nearest to farthest;

[0160] Assign a sequence number to the sorting results, i.e., the second sorting sequence number: the robot ranked 1st is numbered 1, the 2nd is numbered 2, and so on.

[0161] Example: If robot A (3.2m) ranks 2nd, its second ranking number is 2; if robot B (6.5m) ranks 7th, its second ranking number is 7.

[0162] The second sorting number is directly used as the second scheduling value. The smaller the value, the less negative the distance has on the scheduling priority. For example, the second scheduling value of robot A is 2, and the second scheduling value of robot B is 7.

[0163] A weighted average algorithm is used to fuse the first and second scheduling values ​​to quantify the overall priority of the temporary robot, as follows:

[0164] Based on the requirements of the workshop transportation scenario, the weights ω1 of the first scheduling value (task volume dimension) and ω2 of the second scheduling value (distance dimension) are preset, satisfying ω1 + ω2 = 1. For example, in a typical scenario, ω1 = 0.4 and ω2 = 0.6, prioritizing the impact of distance on response speed.

[0165] The temporary scheduling value for each temporary robot is calculated based on the following formula:

[0166] Temporary scheduling value = first scheduling value × ω1 + second scheduling value × ω2;

[0167] Example:

[0168] Robot A: Temporary scheduling value = 3 × 0.4 + 2 × 0.6 = 1.2 + 1.2 = 2.4;

[0169] Robot B: Temporary scheduling value = 8 × 0.4 + 7 × 0.6 = 3.2 + 4.2 = 7.4;

[0170] Sort all temporary scheduling values ​​of the robots in ascending order and select the robot with the smallest temporary scheduling value as the optimal robot. For example, if the temporary scheduling value of robot A (2.4) is less than that of robot B (7.4), then robot A is selected as the optimal robot.

[0171] If there are temporary robots with the same number of tasks or the same temporary distance value, they are sorted a second time by "task urgency" (e.g., whether they include urgent tasks) or "number of collaborations in the past" (prioritizing robots that have collaborated before) to avoid sequence number conflicts.

[0172] The method allows modification of the values ​​of ω1 and ω2 through the backend configuration interface:

[0173] When the workshop robots are generally idle, i.e. the workload is generally low, increase ω2, such as ω2 = 0.7, to strengthen the impact of distance on priority;

[0174] When the temporary robot load in the workshop is high, that is, when the workload is generally large, increase ω1, such as ω1=0.5, and give priority to selecting robots with less workload.

[0175] By ranking and weighting the tasks based on both workload and distance, the scheduling priority of temporary robots can be scientifically quantified. This allows robots with fewer tasks and shorter distances to receive lower scheduling values, thus prioritizing their participation in collaboration. This approach ensures efficient progress of the current transportation task while minimizing interference with other tasks. By dynamically balancing resource utilization and response efficiency, it enhances the overall scheduling flexibility and system stability, achieving optimized configuration for multi-robot collaboration.

[0176] In the process of calculating the temporary scheduling value based on the weighted average algorithm, this embodiment introduces a dynamic weight adjustment mechanism. By analyzing the task volume distribution characteristics and distance dispersion in real time, the weight allocation of the first scheduling value and the second scheduling value is adaptively adjusted. The specific implementation steps are as follows:

[0177] Count the number of tasks in the task lists of all temporary robots and calculate the arithmetic mean: Where Ni is the number of tasks for the i-th temporary robot, and n is the total number of temporary robots. (Based on the average number of tasks...) The weight ω1 of the first scheduling value is adjusted inversely based on the comparison result with the preset threshold TN:

[0178]

[0179] Wherein, ω1base is the base weight (e.g., 0.4), ω1max and ω1min are the upper and lower limits of the weight (e.g., 0.7 and 0.2); TNbase is the baseline threshold of the task load (e.g., 3 items), TNmin and TNmax are the threshold intervals (e.g., 1 and 5); k1 is the adjustment coefficient (e.g., 0.1), which controls the rate at which the weight changes with the task load.

[0180] When the workload is generally small When this happens, increase the task weight to the maximum value and prioritize selecting the robot with the fewest tasks;

[0181] When the workload is generally large At this time, reduce the weight of the task load to the minimum value to reduce the impact of the task load factor.

[0182] Calculate the standard deviation σ of the temporary distance values ​​for all temporary robots. D As an indicator of dispersion: Where Di is the temporary distance value of the i-th temporary robot. This represents the average distance.

[0183] Based on the standard deviation of distance σ D The weight ω2 of the second scheduling value is adjusted in a positive correlation with the comparison result with the preset threshold TD.

[0184]

[0185] Wherein: ω2base is the base weight (e.g., 0.6), ω2max and ω2min are the upper and lower limits of the weight (e.g., 0.8 and 0.3); TDbase is the distance to the discrete baseline threshold (e.g., 2.0m), TDmin and TDmax are the threshold intervals (e.g., 1.0m and 3.0m); k2 is the adjustment coefficient (e.g., 0.1), which controls the rate at which the weight changes with the degree of dispersion.

[0186] When the distance distribution is relatively concentrated (σ D When ≤TDmin), reduce the distance weight to avoid overemphasizing distance factors;

[0187] When the distance distribution is dispersed (σ) D When the distance is greater than or equal to TDmax, the distance weight is increased, and the robot with the closest distance is selected first.

[0188] The temporary scheduling value is calculated using the dynamically adjusted weights ω1 and ω2 (satisfying ω1+ω2=1): Temporary scheduling value i=ω1×first scheduling value i+ω2×second scheduling value i;

[0189] Threshold filtering is applied to the calculation results to avoid scheduling deviations caused by extreme values:

[0190] If a robot's temporary scheduling value exceeds twice the standard deviation of the average value of all robots, then its scheduling value is set to the average value plus twice the standard deviation.

[0191] Ensure all scheduling values ​​are positive; negative values ​​are automatically corrected to 0.1.

[0192] Intelligent weighting is achieved through dynamic adjustment of weights: if the overall workload of temporary robots is low, the workload weight is increased to prioritize the use of idle resources; if the temporary robots are widely distributed, the distance weight is increased to accurately select neighboring robots. This adaptive adjustment mechanism can dynamically balance resource utilization and response speed based on the real-time scheduling environment, preventing decision-making biases caused by a single indicator.

[0193] In the step of controlling the optimal robot (slave robot) to move to the AGV robot (master robot) carrying the product and jointly carrying the product, this embodiment achieves stable support for stacked products through symmetrical position layout and precise movement control. The specific sub-steps are as follows:

[0194] Based on the pallet size model, the default pallet center coordinates are set as the origin (0, 0), and the major axis is along the transport route. The first position is marked next to the pallet center with coordinates (+L / 2, 0), where L is the safety distance from the pallet center to the side. The safety distance is dynamically set according to the pallet width. For example, when the pallet width is 1.2m, L = 0.4m.

[0195] The robot uses laser navigation to locate its position, with the positioning error controlled within ±2cm.

[0196] Once the robot reaches the first position, the top support mechanism, such as an electric lifting platform, is activated, lifting the robot at a preset speed (5cm / s) until it contacts the bottom of the tray. A pressure sensor detects the load signal; when the pressure reaches 30% of the total weight of the product, the lifting stops and the height is locked, completing the load-bearing preparation.

[0197] The main robot, initially located at the center of the pallet (0, 0), moves to a second position (-L / 2, 0) along the transport route after receiving a signal indicating that the robot has completed its load-bearing task. This second position is symmetrical about the center of the pallet to the first position. During the relocation, the movement distance is fed back in real time by a displacement sensor to ensure that the final position deviation is ≤ ±1cm.

[0198] The system obtains the real-time coordinates of the master and slave robots through their positioning modules and calculates the symmetry deviation value: if the deviation between the midpoint of the line connecting the first position and the second position and the center of the tray is >3cm, a fine-tuning command is triggered, and the master robot or slave robot moves slightly until the symmetry accuracy requirement is met.

[0199] The midpoint of the line connecting the first and second positions coincides with the central axis of the transport route, ensuring that both the master and slave robots are located on the transport route:

[0200] Route recognition sensors, such as vision sensors, can be used to identify the deviation between the robot's center and the route's central axis by recognizing ground navigation lines.

[0201] If the deviation is greater than 5cm, start the lateral fine adjustment at a speed of ≤2cm / s until it fits the route.

[0202] The master and slave robots collect real-time load data through pressure sensors. The system calculates the load ratio between the two robots, with a target ratio of 1:1 and an allowable deviation of ±5%. If the ratio is unbalanced, the system balances the force by finely adjusting the height of the load-bearing mechanism (each adjustment ≤1cm), and finally generates a "cooperative load-bearing ready" signal to trigger the subsequent transportation process.

[0203] The master and slave robots are symmetrically distributed on both sides of the center of the pallet and share the load. This layout ensures that the center of gravity of the stacked products falls precisely on the symmetrical center area supported by both robots, fundamentally reducing the possibility of center of gravity shift. At the same time, both robots are positioned close to the transport route, ensuring even force distribution during coordinated movement. Whether during acceleration, deceleration, or turning, the swaying of the carried products is significantly reduced, thereby minimizing the risk of falling and making the transportation process more stable and reliable.

[0204] Reference Figure 3 In the step of controlling the optimal robot (slave robot) to move to the AGV robot (master robot) carrying the product and jointly carrying the product, this embodiment achieves dynamic optimization of cooperative posture and support spacing through robot size parameter analysis. The specific sub-steps are as follows:

[0205] The system retrieves the rated dimensional parameters of the master and slave robots from the robot parameter database, including length (maximum dimension along the direction of movement) and width (maximum dimension perpendicular to the direction of movement). If the robots are heterogeneous (master and slave models are different), the actual measured dimensions are collected synchronously.

[0206] Main robot aspect ratio = Main robot length ÷ Main robot width (rounded to 2 decimal places);

[0207] Aspect Ratio = Robot Length ÷ Robot Width (rounded to 2 decimal places);

[0208] Average aspect ratio = (Main aspect ratio + Sub aspect ratio) ÷ 2.

[0209] Example: The main robot is 1.2m long and 0.8m wide (main aspect ratio 1.5), and the secondary robot is 1.0m long and 0.7m wide (secondary aspect ratio 1.43). The average aspect ratio is 1.46.

[0210] Preset aspect ratio reference value (set according to transportation stability test, such as 1.3): When the average aspect ratio is greater than the reference value, the robot is judged to have "significant elongated features" and its posture needs to be adjusted to reduce the risk of lateral swaying.

[0211] The master and slave robots receive attitude adjustment instructions and rotate the length direction of the robot body by 90° through the steering drive module so that it is perpendicular to the forward direction of the current transportation route. The original length direction is along the forward direction, and the adjusted width direction is along the forward direction.

[0212] During the adjustment process, the steering angle is monitored in real time using a gyroscope, and the steering mechanism is locked once the desired position is reached.

[0213] After the attitude adjustment is completed, the two are aligned in a straight line along the direction of travel, with the width of the fuselage parallel to the transport route.

[0214] Advantages: When the long, rectangular robot is placed horizontally, the lateral anti-overturning moment is increased by 40%, which can effectively counteract the centrifugal force during turning.

[0215] Calculate the average ratio of the aspect ratio to the average of the reference values: Average ratio = Average aspect ratio ÷ Preset aspect ratio reference value;

[0216] Example: If the average aspect ratio is 1.46 and the reference value is 1.3, then the average ratio = 1.12; the larger the ratio, the larger the support spacing is required.

[0217] Basic spacing setting: Preset the initial spacing based on the pallet width. For example, if the pallet width is 1.2m, the initial spacing is 0.8m.

[0218] Dynamic adjustment: The spacing is adjusted according to the average ratio; that is, the final spacing = the basic spacing × the average ratio; the maximum spacing shall not exceed 80% of the pallet width to avoid exceeding the pallet support range;

[0219] Execution control: The master and slave robots adjust the distance by lateral movement (speed ≤ 5cm / s), and confirm the position by infrared alignment sensor (distance error ≤ ±3cm).

[0220] Example: If the initial spacing is 0.8m and the average ratio is 1.12, then the final spacing = 0.8 × 1.12 = 0.9m, which is 0.1m more than the initial spacing, thus improving the stability of the support.

[0221] The robot's collaborative posture and spacing are dynamically adjusted based on its aspect ratio: if the aspect ratio is large, the robot's length direction is adjusted to be perpendicular to the forward direction to reduce the possibility of lateral swaying; at the same time, the spacing is increased according to this ratio to enhance the stability of the support. This adaptive adjustment mechanism optimizes the spatial distribution of the master and slave robots and improves the control of the center of gravity.

[0222] This application also discloses an AGV robot scheduling system for a smart factory, including a processor, wherein the processor executes the steps of the AGV robot scheduling method for a smart factory as described in any of the above embodiments.

[0223] This application also discloses a storage medium storing a program that, when executed by a processor, implements the steps of the AGV robot scheduling method for a smart factory as described in any of the above embodiments.

[0224] Although embodiments of this application have been shown and described above, it is understood that the above embodiments are exemplary and should not be construed as limiting this application. Those skilled in the art can make changes, modifications, substitutions and variations to the above embodiments within the scope of this application.

Claims

1. A method for scheduling AGV robots in a smart factory, characterized in that, Includes the following steps: Based on product handling instructions and transportation routes, the AGV robot's load-bearing sensor data is collected, and the product height value is identified from the load-bearing sensor data; if the product height value is higher than the preset reference height value, an auxiliary scheduling instruction is generated. In response to the auxiliary scheduling command, search for other AGV robots whose shortest distance to the transportation route is within a preset first distance range; read the working status of the other AGV robots, and mark the AGV robots whose working status is unloaded as temporary robots; Obtain the task list of the temporary robot, sort the temporary robots according to the number of tasks in the task list to obtain a first scheduling value; calculate the temporary distance value between the temporary robot and the AGV robot carrying the product, sort the temporary robots according to the temporary distance value to obtain a second scheduling value; calculate the temporary scheduling value according to the first scheduling value and the second scheduling value, and sort the robots according to the temporary scheduling value to obtain the optimal robot; The optimal robot is paired and bound to the AGV robot carrying the product. The optimal robot is controlled to move to the AGV robot carrying the product and jointly carry the product. The AGV robot carrying the product is the master robot and the optimal robot is the slave robot. According to the controlled instructions of the master robot, the master scheduling instructions and slave scheduling instructions are matched from the preset scheduling instruction library. The master robot responds to the master scheduling instructions, and the slave robot responds to the slave scheduling instructions.

2. The AGV robot scheduling method for a smart factory according to claim 1, characterized in that, The steps of the master robot responding to the master scheduling command and the slave robot responding to the slave scheduling command further include the following sub-steps: Based on the main scheduling command, the main robot calculates motion data, which includes motion direction, motion speed, and motion acceleration. The direction coefficient is matched from a preset direction symbol library according to the direction of motion; The speed coefficient is calculated based on the movement speed and the preset reference speed; The acceleration coefficient is calculated based on the motion acceleration and the preset reference acceleration; The height adjustment coefficient is calculated based on the direction coefficient, the velocity coefficient, and the acceleration coefficient. The height of the master robot can be adjusted inversely based on the height adjustment coefficient, or the height of the slave robot can be adjusted in a positive correlation based on the height adjustment coefficient, or the height of the master robot can be adjusted inversely based on the height adjustment coefficient while the height of the slave robot is adjusted in a positive correlation.

3. The AGV robot scheduling method for a smart factory according to claim 1, characterized in that, The step of matching the master scheduling instruction and the slave scheduling instruction from the preset scheduling instruction library according to the master robot's controlled instructions further includes the following sub-steps: The master robot sends the controlled instruction to the background database and queries the background database for the master scheduling instruction and the slave scheduling instruction corresponding to the controlled instruction; the master robot receives the master scheduling instruction and the slave scheduling instruction sent by the background database; The master robot executes the master scheduling instruction and sends the slave scheduling instruction to the slave robot; The robot receives and executes the scheduling instructions. Alternatively, the master robot sends the controlled instruction and the communication address of the slave robot to the backend. The backend queries the backend database for the master scheduling instruction and the slave scheduling instruction corresponding to the controlled instruction. The backend returns the master scheduling instruction to the master robot and sends the slave scheduling instruction to the slave robot. The master robot receives and executes the master scheduling instruction, and the slave robot receives and executes the slave scheduling instruction.

4. The AGV robot scheduling method for a smart factory according to claim 1, characterized in that, The step of collecting the load-bearing sensor data of the AGV robot and identifying the product height value from the load-bearing sensor data further includes the following sub-steps: The carrying sensing data is obtained from the image sensor at a preset position on the AGV robot, and the image data of the product to be carried is obtained based on the image sensor. The image recognition algorithm identifies the pixel coordinates of the product in the image data. Obtain the real-time position and shooting angle of the image sensor; The product height is calculated based on the real-time location, shooting angle, and pixel coordinates. or, The load-bearing sensing data is from the distance sensor on the top or side of the AGV robot. The distance sensor emits continuous detection signals as it rotates from the bottom to the top of the stacked load-bearing products. The feedback signal of the continuous detection signal is analyzed to obtain the initial feedback signal and the abrupt feedback signal; The product height value is calculated based on the time and phase of the mutation feedback signal, as well as the time and phase of the initial feedback signal.

5. The AGV robot scheduling method for a smart factory according to claim 1, characterized in that, The step of calculating the temporary scheduling value based on the first scheduling value and the second scheduling value further includes the following steps: The sequence number of the temporary robot in the first sorting is positively correlated with the number of tasks in the task list. The more tasks there are, the larger the sequence number in the first sorting, and the fewer tasks there are, the smaller the sequence number in the first sorting. The sequence number of the temporary robot in the second sorting is positively correlated with the temporary distance value. The larger the temporary distance value, the larger the sequence number in the second sorting; the smaller the temporary distance value, the smaller the sequence number in the second sorting. The first scheduling value and the second scheduling value are used to calculate the temporary scheduling value using a weighted average algorithm; the optimal robot is the temporary robot with the smallest temporary scheduling value.

6. The AGV robot scheduling method for a smart factory according to claim 5, characterized in that, The step of calculating the temporary scheduling value using a weighted average algorithm based on the first scheduling value and the second scheduling value further includes the following steps: The weight of the first scheduling value is adjusted inversely based on the average number of tasks in the task lists of all the temporary robots; the smaller the average number, the greater the weight of the first scheduling value; the greater the average number, the smaller the weight of the first scheduling value. The weight of the second scheduling value is adjusted according to the positive correlation between the dispersion value of the temporary distance values ​​of all the temporary robots and the dispersion value; the smaller the dispersion value, the smaller the weight of the second scheduling value; the larger the dispersion value, the larger the weight of the second scheduling value.

7. The AGV robot scheduling method for a smart factory according to claim 1, characterized in that, The step of controlling the optimal robot to move to the AGV robot carrying the product and jointly carrying the product includes the following sub-steps: The lifting and load-bearing mechanism bears the load as the robot moves to the first position next to the center of the pallet carrying the product. The main robot moves from the center of the pallet carrying the product to a second position next to the center of the pallet. The first position and the second position are located on both sides of the center of the pallet and are centrally symmetrical. The first position and the second position are located on the transport route.

8. The AGV robot scheduling method for a smart factory according to claim 7, characterized in that, The step of controlling the optimal robot to move to the AGV robot carrying the product and jointly carrying the product includes the following sub-steps: Obtain the dimensions of the master robot and the slave robot; The aspect ratio of the main robot is calculated based on its dimensions, and the aspect ratio of the slave robot is calculated based on its dimensions. The average aspect ratio is calculated based on the main aspect ratio and the secondary aspect ratio; If the average aspect ratio is greater than the preset aspect ratio reference value, the main robot and the slave robot will adjust their length direction to be perpendicular to the forward direction of their current position on the transportation route, and move along the forward direction. Calculate the average ratio of the aspect ratio to the preset aspect ratio reference value, and adjust the spacing between the first position and the second position according to the positive correlation of the average ratio.

9. An AGV robot scheduling system for a smart factory, characterized in that, Includes a processor, wherein the steps of the AGV robot scheduling method for a smart factory as described in any one of claims 1-8 are executed.

10. A storage medium, characterized in that, The storage medium stores a program that, when executed by a processor, implements the steps of the AGV robot scheduling method for a smart factory as described in any one of claims 1-8.

Images