Agv zinc ingot carrying path planning method, system and medium

Abstract

The application discloses an AGV zinc ingot carrying path planning method and system and a medium. When a zinc ingot carrying instruction is received, initial position information, terminal position information and an environment model corresponding to the zinc ingot carrying instruction are acquired. The state identifiers of all AGVs are acquired, and idle AGVs in an idle state are determined from all the AGVs according to the state identifiers. The AGV position information of all the idle AGVs is acquired, and a carrying path corresponding to the zinc ingot carrying instruction is planned according to the initial position information, the terminal position information, the environment model and the AGV position information. The application plans the shortest reasonable path that effectively avoids obstacles by identifying idle AGVs, combining their own displacement, the position where the zinc ingot needs to be carried and the environment model of the production workshop, thereby improving the efficiency and safety of zinc ingot carrying.

Description

Technical Field

[0001] This invention relates to the field of industrial control technology, and in particular to an AGV zinc ingot handling path planning method, system and medium. Background Technology

[0002] Power transmission towers are an important component of power transmission lines in power systems, and are manufactured from angle steel. To protect the angle steel from corrosion in outdoor environments, it is typically treated with zinc ingots for corrosion protection. Through hot-dip galvanizing, a zinc coating is formed on the surface of the angle steel, effectively preventing rust and extending the service life of the tower.

[0003] The processing and anti-corrosion treatment of angle steel are usually carried out in the production workshop. Zinc ingots are stacked in one location in the production workshop and then transported to the hot-dip galvanizing processing area using traditional methods such as forklifts and cranes for galvanizing. This method of handling zinc ingots is inefficient and poses safety hazards.

[0004] With the development of smart warehouses, Automated Guided Vehicles (AGVs) have been widely used in warehousing. AGVs are characterized by automation, intelligence, and high efficiency, effectively reducing labor costs, improving transportation efficiency, and playing a vital role in material handling and warehouse management. However, compared to warehousing, the application environment in zinc ingot production workshops is more complex. How to rationally plan the paths of AGVs when transporting zinc ingots in the production workshop, avoiding conflicts, and improving efficiency and safety remains a challenging problem. Summary of the Invention

[0005] The main objective of this invention is to provide an AGV zinc ingot handling path planning method, system, and medium, aiming to solve the technical problem of how to rationally plan the zinc ingot handling path in the prior art, avoid conflicts, and improve efficiency and safety.

[0006] To achieve the above objectives, the present invention provides an AGV zinc ingot handling path planning method, the AGV zinc ingot handling path planning method comprising: When a zinc ingot handling instruction is received, the initial position information, the destination position information, and the environment model corresponding to the zinc ingot handling instruction are obtained. Obtain the status identifier of each AGV, and determine the idle AGV that is in an idle state from among the AGVs based on the status identifier; Obtain the AGV position information of each of the idle AGVs, and plan the transportation path corresponding to the zinc ingot transportation instruction based on the initial position information, the destination position information, the environment model, and the position information of each AGV.

[0007] Preferably, the step of planning the transport path corresponding to the zinc ingot transport instruction based on the initial position information, the destination position information, the environmental model, and the position information of each AGV includes: Based on the initial location information, the destination location information, the environment model, and the location information of each AGV, a walking path corresponding to each of the idle AGVs is generated; For each walking path, a correction time factor corresponding to the walking path is determined, and the walking time corresponding to the walking path is determined based on the correction time factor. Based on the travel time corresponding to each of the travel paths, the transport path is determined from each of the travel paths, and the target AGV corresponding to the transport path is determined from each of the idle AGVs.

[0008] Preferably, the step of determining the transport path from the walking paths based on the walking time corresponding to each walking path includes: Based on the walking time, determine the unoccupied walking paths in each of the walking paths; For each of the aforementioned idle walking paths, find the path length and the corresponding length-time weight value for that idle walking path; Based on the length-time weight value, the walking time corresponding to the idle walking path and the path length are weighted and calculated to generate the path factor of the idle walking path; After generating a corresponding path factor for each of the idle walking paths, the path factors are compared among themselves, and the transport path is determined based on the comparison results.

[0009] Preferably, the step of generating a walking path corresponding to each of the idle AGVs based on the initial position information, the destination position information, the environment model, and the position information of each AGV includes: Based on the initial location information, the destination location information, and the environment model, a fixed path corresponding to the zinc ingot handling instruction is generated. For each idle AGV, a change path corresponding to the idle AGV is generated based on the AGV position information of the idle AGV, the initial position information, and the environment model; Based on the fixed path and each of the variable paths, a walking path corresponding to each of the idle AGVs is generated.

[0010] Preferably, the step of determining the target AGV corresponding to the transport path from each of the idle AGVs includes: Control the target AGV to move along the transport path, and receive environmental data collected by the target AGV during its movement; Based on the environmental data, it is determined whether there are obstacles in the transport path. If there are obstacles, the current position information of the target AGV is obtained, and the transport path is updated based on the current position information, the destination position information, and the environmental model.

[0011] Preferably, the environmental data includes at least ultrasonic signals and environmental images, and the step of determining whether there are obstacles in the transport path based on the environmental data includes: Based on the ultrasonic signal, it is determined whether there is an obstacle to be identified in the transport path. If there is an obstacle to be identified, the environmental image is subjected to grayscale conversion, image correction and edge detection to generate a target image. The target image is identified based on a preset recognition model to determine the size of the obstacle to be identified; Based on the size of the object, it is determined whether the obstacle to be identified obstructs the movement of the target AGV. If it obstructs the movement of the target AGV, it is determined that there is an obstacle in the transport path.

[0012] Preferably, the step of obtaining the initial position information, destination position information, and environmental model corresponding to the zinc ingot handling instruction when the zinc ingot handling instruction is received includes the following steps before: Acquire a large number of historical environmental images corresponding to the transport path, and divide the historical environmental images into environmental training data and environmental verification data; The preset initial model is trained based on the environmental training data, and when the training reaches a preset duration, the preset initial model is validated based on the environmental validation data, generating a loss function value corresponding to the preset initial model. Determine whether the loss function value is less than a preset threshold. If it is less than the preset threshold, then generate the preset initial model as a preset recognition model.

[0013] Preferably, the step of determining whether the loss function value is less than a preset threshold is followed by: If the loss function value is greater than or equal to a preset threshold, the model parameters of the preset initial model are updated, and the preset initial model is trained based on the environmental training data for the updated preset initial model.

[0014] Furthermore, to achieve the above objectives, the present invention also provides an AGV zinc ingot handling path planning system, the AGV zinc ingot handling path planning system including a storage device, a processor, a communication bus, and a control program stored in the storage device: The communication bus is used to enable communication between the processor and the memory. The processor is used to execute the control program to implement the steps of the AGV zinc ingot handling path planning method described above.

[0015] Furthermore, to achieve the above objectives, the present invention also provides a medium, which is a readable storage medium storing a control program. When the control program is executed by a processor, it implements the steps of the AGV zinc ingot handling path planning method described above.

[0016] This invention discloses an AGV zinc ingot handling path planning method, system, and medium. Upon receiving a zinc ingot handling instruction, it acquires the initial position information, destination position information, and environmental model carried within the instruction. Simultaneously, it acquires the status identifiers of each AGV and, based on these identifiers, identifies idle AGVs. Furthermore, it acquires the AGV position information of each idle AGV and combines the initial position information, destination position information, and environmental model with the AGV position information to plan a reasonable handling path corresponding to the zinc ingot handling instruction. In this way, by identifying idle AGVs and combining their displacement with the location of the zinc ingot to be handled and the environmental model of the production workshop, it plans the shortest reasonable path that effectively avoids obstacles, improving the efficiency and safety of zinc ingot handling. Attached Figure Description

[0017] Figure 1 This is a flowchart illustrating the first embodiment of the AGV zinc ingot handling path planning method of the present invention; Figure 2 This is a flowchart illustrating the second embodiment of the AGV zinc ingot handling path planning method of the present invention; Figure 3 This is a flowchart illustrating the third embodiment of the AGV zinc ingot handling path planning method of the present invention; Figure 4 This is a schematic diagram of the hardware operating environment involved in an embodiment of the AGV zinc ingot handling path planning system of the present invention.

[0018] The realization of the objective, functional features and advantages of the present invention will be further explained in conjunction with the embodiments and with reference to the accompanying drawings. Detailed Implementation

[0019] It should be understood that the specific embodiments described herein are for illustrative purposes only and are not intended to limit the scope of the invention.

[0020] This invention provides an AGV zinc ingot handling path planning method, please refer to... Figure 1 , Figure 1 This is a flowchart illustrating the first embodiment of the AGV zinc ingot handling path planning method of the present invention.

[0021] This invention provides an embodiment of an AGV zinc ingot handling path planning method. It should be noted that although the logical order is shown in the flowchart, in some cases, the steps shown or described may be performed in a different order than that shown here. Specifically, the AGV zinc ingot handling path planning method in this embodiment includes:

[0022] Step S10: When a zinc ingot handling instruction is received, the initial position information, the destination position information, and the environment model corresponding to the zinc ingot handling instruction are obtained.

[0023] In this embodiment, the AGV zinc ingot handling path planning method is applied to the backend server of the path planning system.

[0024] The path planning system corresponds to the intelligent production workshop that produces angle steel for power transmission towers. Through a backend server connected to the system, it plans the paths for AGVs to transport zinc ingots within the workshop, ensuring the AGVs efficiently and safely move the ingots from their initial positions to the processing locations. When the intelligent production workshop needs to move zinc ingots, it sends a zinc ingot moving instruction to the backend server. This instruction carries at least the initial and final position coordinates. The backend server receives the instruction, identifying the initial position coordinates as the initial position information and the final position coordinates as the final position information. Simultaneously, the backend server searches a pre-established environmental model for the intelligent production workshop as the corresponding environment model for the zinc ingot moving instruction.

[0025] Step S20: Obtain the status identifier of each AGV, and determine the idle AGV that is in an idle state from among the AGVs based on the status identifier.

[0026] Understandably, a smart production workshop typically has multiple AGVs, each controlled by a backend server and in different states. These include at least a busy state operating within the smart production workshop, an energy storage state where they are charging in a charging area due to insufficient power, and an idle state where they are fully charged but waiting in a standby area. Furthermore, AGVs in different states are distinguished by different status identifiers. The backend server retrieves the status identifiers of each AGV, filters out the idle status identifiers, and then locates the AGVs corresponding to each idle status identifier. These idle AGVs are the ones in the smart production workshop that are currently idle.

[0027] Step S30: Obtain the AGV position information of each of the idle AGVs, and plan the transportation path corresponding to the zinc ingot transportation instruction based on the initial position information, the destination position information, the environment model, and the position information of each AGV.

[0028] Furthermore, the coordinates of the current location of each idle AGV are acquired as AGV position information. Based on this acquired AGV position information, combined with initial position information, destination position information, and the environmental model, the transport path from each AGV's current location to the initial location of the zinc ingot is determined. Then, according to the environmental model, the zinc ingot is transported from the initial location to the destination location via the transport path. Each AGV corresponds to one transport path, and different transport paths have different routes and require different transport times. Therefore, by combining the path length and transport time of each transport path, the most reasonable transport path can be determined.

[0029] The AGV zinc ingot handling path planning method implemented in this paper acquires the initial position information, destination position information, and environmental model carried in the zinc ingot handling instruction upon receiving it. Simultaneously, it acquires the status identifiers of each AGV and identifies idle AGVs based on these identifiers. Then, it acquires the AGV position information of each idle AGV and combines the initial position information, destination position information, and environmental model with the AGV position information to plan a reasonable handling path corresponding to the zinc ingot handling instruction. In this way, by identifying idle AGVs and combining their own displacement with the location of the zinc ingot to be handled and the environmental model of the production workshop, the shortest reasonable path that effectively avoids obstacles is planned, improving the efficiency and safety of zinc ingot handling.

[0030] Further, please refer to Figure 2 Based on the first embodiment of the AGV zinc ingot handling path planning method of the present invention, a second embodiment of the AGV zinc ingot handling path planning method of the present invention is proposed.

[0031] The difference between the second embodiment of the AGV zinc ingot handling path planning method and the first embodiment of the AGV zinc ingot handling path planning method is that the step of planning the handling path corresponding to the zinc ingot handling instruction based on the initial position information, the destination position information, the environmental model, and the position information of each AGV includes: Step S31: Generate a walking path corresponding to each of the idle AGVs based on the initial position information, the destination position information, the environment model, and the position information of each AGV. Furthermore, the intelligent production workshop can be divided into multiple areas for parking idle AGVs. Different parking areas correspond to different parking positions, and the paths from these different positions to the initial position are different. In addition, once an AGV completes its transport task, its status changes from busy to idle, resulting in some idle AGVs being on their way back to the parking area. The paths from these idle AGVs on their way to the initial position differ depending on their parking position. Moreover, even for the same AGV, there are multiple paths to the initial position. Based on this, the AGV location information of each idle AGV, along with the initial position information, the destination position information, and the environment model, are used to plan all possible travel paths for each idle AGV until a corresponding travel path is generated for each idle AGV. Specifically, the step of generating a travel path corresponding to each idle AGV based on the initial position information, the destination position information, the environment model, and the location information of each AGV includes:

[0032] Step S311: Generate a fixed path corresponding to the zinc ingot handling instruction based on the initial position information, the destination position information, and the environment model; Step S312: For each idle AGV, generate a change path corresponding to the idle AGV based on the AGV position information of the idle AGV, the initial position information, and the environment model; Step S313: Generate a walking path corresponding to each of the idle AGVs based on the fixed path and each of the variable paths.

[0033] Furthermore, the initial location of the zinc ingot and its destination are fixed. Although there are multiple paths from the initial location to the destination, these paths are all fixed for an idle AGV. Therefore, based on the operable sections of the intelligent production workshop as depicted in the environmental model, and combining the initial and destination location information, a fixed path corresponding to the zinc ingot handling instruction is generated. This fixed path encompasses multiple routes from the initial location to the destination.

[0034] Furthermore, each idle AGV is located in a different position, and each idle AGV takes a different route from its own position to the initial position of the zinc ingot. Moreover, there may be multiple routes for the same idle AGV from its starting position to the initial position of the zinc ingot. Therefore, for each idle AGV, based on the operable routes of the intelligent production workshop as represented by the environmental model, and combining its AGV position information and initial position information, a variable path is generated for the AGV to travel to its initial position. This variable path encompasses multiple paths from the AGV's current position to the initial position via different routes.

[0035] Furthermore, for each idle AGV, its variable path is merged with its fixed path, thus generating multiple travel paths corresponding to the idle AGV. The merging method involves merging each variable path with each fixed path individually, generating a travel path corresponding to each variable path, until each variable path is merged with all the fixed paths. For example, if an AGV has two variable paths, A1 and A2, and three fixed paths, B1, B2, and B3, then A1 is merged with B1, B2, and B3 respectively to generate three travel paths corresponding to A1, and A2 is merged with B1, B2, and B3 respectively to generate three travel paths corresponding to A2, resulting in a total of six travel paths.

[0036] Step S32: For each walking path, determine the correction time factor corresponding to the walking path, and determine the walking time corresponding to the walking path based on the correction time factor. Step S33: Based on the travel time corresponding to each travel path, determine the transport path from each travel path, and determine the target AGV corresponding to the transport path from each idle AGV; Understandably, different travel paths correspond to different routes, and different routes correspond to different turning angles. When the AGV reaches a turning angle, it needs to turn, and the turning speed of the AGV is usually slower than its straight-line speed. Therefore, routes with more turning angles require more turns, resulting in longer AGV travel time and hindering the efficiency of zinc ingot handling. Based on this, in this embodiment, the time lost by the AGV due to turning is pre-tested as a time factor. For each travel path, the number of turning angles in the route is counted, and then the time factor and torque quantity are multiplied to obtain the corrected time factor corresponding to the travel path. Simultaneously, for the straight-line routes excluding turning angles, the straight-line time is calculated based on the total route length and the AGV's travel speed. This straight-line time is then corrected according to the corrected time factor, and the two are summed to obtain the travel time corresponding to the travel path.

[0037] Furthermore, after generating the corresponding travel time for each travel path, the travel path with the shortest travel time can be selected as the transport path based on the length of each travel time. At the same time, the idle AGV corresponding to the travel path with the shortest time is identified as the target AGV, so that the target AGV can be controlled to transport zinc ingots according to the selected transport path.

[0038] Understandably, the target AGV transports zinc ingots according to a determined transport path, and the travel time for transporting these ingots is the travel duration. However, within this travel duration, other AGVs may occupy a segment of the transport path, resulting in route congestion and extending the travel time. Therefore, this embodiment determines the transport path based on the occupancy status of each travel path. Specifically, the step of determining the transport path from each travel path based on the travel time corresponding to each travel path includes:

[0039] Step S331: Based on the walking time, determine the unoccupied walking paths in each walking path. Step S332: For each of the idle walking paths, find the path length and the corresponding length-time weight value of the idle walking path; Step S333: Based on the length-time weight value, perform a weighted calculation on the walking time corresponding to the idle walking path and the path length to generate the path factor of the idle walking path; Step S334: After generating a corresponding path factor for each of the idle walking paths, compare the path factors and determine the transport path based on the comparison results.

[0040] Furthermore, the usage status of each segment within each travel path during future time periods is identified in the path planning system. Therefore, for each travel path, it is checked whether any segment is identified as being in use during the travel time. If such a segment exists, it indicates that other AGVs are traveling along that path during its travel time, and the path is considered occupied. Conversely, if no segment is identified as being in use, it indicates that other types of AGVs are traveling along that path during its travel time, and the path is considered unoccupied and free. Therefore, for the merged travel paths, the usage status of each path during its respective travel time is determined by the presence or absence of usage indicators, and unoccupied and free travel paths are then selected.

[0041] Furthermore, for each idle travel path, the system retrieves its statistically calculated path length and a pre-set length-time weight value. This length-time weight value includes at least a length weight value and a time weight value, reflecting the relative importance of path length and travel time. Based on this length-time weight value, the travel time and path length are weighted and calculated to generate a path factor for the idle travel path. This path factor indicates the efficiency of the AGV traveling along the idle travel path. The smaller the path factor value, the higher the AGV's travel efficiency and the more rational the path planning.

[0042] Furthermore, after generating corresponding path factors for each idle walking path, the path factors are compared, and the path factor with the smallest value is determined to generate a comparison result. Then, the idle walking path that generated this comparison result is found, and the found idle walking path is determined as the transport path, so that the target AGV can transport zinc ingots according to this transport path, thereby improving transport efficiency.

[0043] Understandably, during the process of transporting zinc ingots along the transport path, unexpected situations may occur, such as items accidentally falling into the transport path, which constitute obstacles and affect the AGV's movement. To address this, each AGV is equipped with a device to collect data on its surrounding environment. This device determines whether obstacles exist in the transport path and then corrects the transport path accordingly. Specifically, after determining the target AGV corresponding to the transport path from the available AGVs, the process includes:

[0044] Step S335: Control the target AGV to move along the transport path and receive environmental data collected by the target AGV during its movement; Step S336: Determine whether there are obstacles in the transport path based on the environmental data. If there are obstacles, obtain the current position information of the target AGV, and update the transport path based on the current position information, the destination position information, and the environmental model.

[0045] Furthermore, when the target AGV receives control commands and moves along the transport path, the environmental data acquisition device installed on the target AGV collects environmental data along the transport path and transmits this data to the backend server. Upon receiving this environmental data, the backend server determines whether there are obstacles in the transport path. If obstacles are found, the server locates the current position of the target AGV, using this as its current position information. Then, based on the currently operable sections of the intelligent production workshop as represented by the environmental model, and combining the current position information with the destination position information, the transport path is updated to generate a new, more efficient transport path.

[0046] Furthermore, the environmental acquisition device includes at least an ultrasonic device and an imaging device, and the acquired environmental data includes at least ultrasonic signals and environmental images. The step of determining whether there are obstacles in the transport path based on the environmental data includes: Step S3361: Determine whether there is an obstacle to be identified in the transport path based on the ultrasonic signal. If there is an obstacle to be identified, perform grayscale conversion, image correction and edge detection on the environmental image to generate a target image. Step S3362: Based on a preset recognition model, the target image is recognized to determine the size of the obstacle to be recognized; Step S3363: Based on the size of the object, determine whether the obstacle to be identified obstructs the movement of the target AGV. If it obstructs the movement of the target AGV, then it is determined that there is an obstacle in the transport path.

[0047] Furthermore, the ultrasonic device emits ultrasonic signals toward the front of the transport path and determines whether there is an obstacle to be identified in the transport path based on the ultrasonic signals. If there is an obstacle to be identified, an environmental image captured by the imaging device is acquired, and the environmental image is converted to grayscale. The grayscale image is then corrected to correct the distortion generated during the image acquisition process. Finally, edge detection is performed on the corrected image to segment out the image of the obstacle to be identified, which is then used as the target image. Radial distortion and tangential distortion may occur during the image acquisition process, so the grayscale image is corrected for radial distortion and tangential distortion respectively according to preset formulas. For the specific formulas for radial distortion correction, please refer to the following formula (1), and for the formulas for tangential distortion correction, please refer to the following formula (2).

[0048] (1); Where (xr, yr) represents the coordinates of a point with radial distortion, (x, y) represents the coordinates of a point without distortion after normalization, k1, k2, and k3 represent the radial distortion coefficients, and r 2 =x 2 +y 2 ; (2); Where (xt, yt) represents the coordinates of a point with tangential distortion, (x, y) represents the coordinates of a point without distortion after normalization, p1 and p2 represent the tangential distortion coefficients, and r 2 =x 2 +y 2 ; Furthermore, a pre-trained recognition model is used to identify the target image, determine the size of the obstacle to be identified, and combine this size with the footprint of the target AGV and the passage distance of the target AGV to determine whether the obstacle obstructs the movement of the target AGV. If it does obstruct, the transport path is considered to have an obstacle and needs to be updated. Conversely, if it does not obstruct, the transport path is considered to be free of obstacles, and the target AGV continues to move mechanically along the original transport path.

[0049] This embodiment generates multiple travel paths using AGV location information, initial location information, destination location information, and an environmental model. It then adjusts the travel time of each path by incorporating a correction time factor related to turning angles. Finally, it determines a high-efficiency transport path by combining the adjusted travel time, path occupancy, and path length. Additionally, it sets up an obstacle prevention update mechanism for the transport path, ensuring the rationality and efficiency of the planned transport path from multiple dimensions.

[0050] Further, please refer to Figure 3 Based on the first and second embodiments of the AGV zinc ingot handling path planning method of the present invention, a third embodiment of the AGV zinc ingot handling path planning method of the present invention is proposed.

[0051] The difference between the third embodiment of the AGV zinc ingot handling path planning method and the first and second embodiments of the AGV zinc ingot handling path planning method is that, before the step of obtaining the initial position information, destination position information, and environmental model corresponding to the zinc ingot handling instruction when the zinc ingot handling instruction is received, the method includes: Step S40: Obtain a large number of historical environmental images corresponding to the transport path, and divide the historical environmental images into environmental training data and environmental verification data; Step S50: Train the preset initial model based on the environmental training data, and when the training reaches a preset duration, validate the preset initial model based on the environmental validation data to generate a loss function value corresponding to the preset initial model. Step S60: Determine whether the loss function value is less than a preset threshold. If it is less than the preset threshold, generate the preset initial model as a preset recognition model.

[0052] This example trains a preset recognition model. A large number of historical environmental images corresponding to the transportation path are acquired and divided into training data and environmental verification data according to a preset division ratio. The preset initial model is then trained using the environmental training data, and the training time is counted. Once the training time reaches the preset time, the preset initial model is verified using the environmental verification data, and the loss function value of the preset initial model is generated using the preset facility loss function. This loss function represents the difference between the processing result obtained by the preset initial model on the environmental verification data and the reference result corresponding to the environmental verification data. The smaller the difference, the better the processing performance of the preset initial model on the environmental verification data; conversely, if the difference is large, the processing performance of the preset initial model on the environmental verification data is poor, and continuous iterative training is required. The specific preset loss function can be found in the following formula (3);

[0053] (3); Where L represents the loss function value, n represents the number of processing results, yi0 represents the i-th reference result, yi1 represents the i-th processing result, s1 represents the regularization coefficient, W represents the model parameter set, s2 represents the hyperparameter, m represents the number of model parameters, and wj represents the j-th model parameter.

[0054] Furthermore, to reflect the magnitude of the difference, a preset threshold is set. The loss function value is compared with this preset threshold to determine whether the loss function value is less than the preset threshold. If it is less, it indicates that the difference between the processed result and the reference result is small, and the performance of the preset initial model is good. Therefore, the preset initial model is generated as the preset recognition model. Otherwise, the preset initial model is iteratively trained. Specifically, after the step of determining whether the loss function value is less than the preset threshold, the following steps are included:

[0055] Step S70: If the loss function value is greater than or equal to a preset threshold, the model parameters of the preset initial model are updated, and the preset initial model is trained based on the environmental training data for the updated preset initial model.

[0056] Furthermore, if the comparison determines that the loss function value is greater than or equal to the preset threshold, the model parameters of the preset initial model are updated according to the pre-set update formula, and the updated preset initial model is trained again using environmental training data to generate a new loss function value for comparison, until the generated loss function value is less than the preset threshold.

[0057] This embodiment trains a preset initial model to generate a preset recognition model by using a large number of historical environmental images related to the transport path. This enables the preset recognition model to more accurately recognize environmental images related to the transport path, thereby improving the recognition effect of obstacles in the transport path.

[0058] Furthermore, embodiments of the present invention also provide an AGV zinc ingot handling path planning system. Please refer to... Figure 4 , Figure 4 This is a schematic diagram of the hardware operating environment of the equipment involved in the embodiment of the AGV zinc ingot handling path planning system of the present invention.

[0059] like Figure 4As shown, the AGV zinc ingot handling path planning system may include: a processor 1001, such as a CPU; a communication bus 1002; a user interface 1003; a network interface 1004; and a storage device 1005. The communication bus 1002 is used to establish communication between these components. The user interface 1003 may include a display screen or an input unit such as a keyboard; optionally, the user interface 1003 may also include a standard wired interface or a wireless interface. The network interface 1004 may optionally include a standard wired interface or a wireless interface (such as a Wi-Fi interface). The storage device 1005 may be a high-speed RAM or a stable, non-volatile memory, such as a disk drive. Optionally, the storage device 1005 may also be a storage device independent of the aforementioned processor 1001.

[0060] Those skilled in the art will understand that Figure 4 The hardware structure of the AGV zinc ingot handling path planning system shown in the figure does not constitute a limitation on the AGV zinc ingot handling path planning system. It may include more or fewer components than shown, or combine certain components, or have different component arrangements.

[0061] like Figure 4 As shown, the storage device 1005, which serves as a storage medium, may include an operating system, a network communication module, a user interface module, and a control program. The operating system is a program that manages and controls the AGV zinc ingot handling path planning system and software resources, supporting the operation of the network communication module, user interface module, control program, and other programs or software. The network communication module manages and controls the network interface 1004; the user interface module manages and controls the user interface 1003.

[0062] exist Figure 4 In the hardware structure of the AGV zinc ingot handling path planning system shown, the network interface 1004 is mainly used to connect to other system servers and communicate data with them; the user interface 1003 is mainly used to connect to the client (user end) and communicate data with it; the processor 1001 can call the control program stored in the memory 1005 and perform the following operations: When a zinc ingot handling instruction is received, the initial position information, the destination position information, and the environment model corresponding to the zinc ingot handling instruction are obtained. Obtain the status identifier of each AGV, and determine the idle AGV that is in an idle state from among the AGVs based on the status identifier; Obtain the AGV position information of each of the idle AGVs, and plan the transportation path corresponding to the zinc ingot transportation instruction based on the initial position information, the destination position information, the environment model, and the position information of each AGV.

[0063] Furthermore, the step of planning the transport path corresponding to the zinc ingot transport instruction based on the initial position information, the destination position information, the environmental model, and the position information of each AGV includes: Based on the initial location information, the destination location information, the environment model, and the location information of each AGV, a walking path corresponding to each of the idle AGVs is generated; For each walking path, a correction time factor corresponding to the walking path is determined, and the walking time corresponding to the walking path is determined based on the correction time factor. Based on the travel time corresponding to each of the travel paths, the transport path is determined from each of the travel paths, and the target AGV corresponding to the transport path is determined from each of the idle AGVs.

[0064] Further, the step of determining the transport path from the walking paths based on the walking time corresponding to each walking path includes: Based on the walking time, determine the unoccupied walking paths in each of the walking paths; For each of the aforementioned idle walking paths, find the path length and the corresponding length-time weight value for that idle walking path; Based on the length-time weight value, the walking time corresponding to the idle walking path and the path length are weighted and calculated to generate the path factor of the idle walking path; After generating a corresponding path factor for each of the idle walking paths, the path factors are compared among themselves, and the transport path is determined based on the comparison results.

[0065] Further, the step of generating a walking path corresponding to each of the idle AGVs based on the initial position information, the destination position information, the environment model, and the position information of each AGV includes: Based on the initial location information, the destination location information, and the environment model, a fixed path corresponding to the zinc ingot handling instruction is generated. For each idle AGV, a change path corresponding to the idle AGV is generated based on the AGV position information of the idle AGV, the initial position information, and the environment model; Based on the fixed path and each of the variable paths, a walking path corresponding to each of the idle AGVs is generated.

[0066] Further, after the step of determining the target AGV corresponding to the transport path from each of the idle AGVs, the processor 1001 can call the control program stored in the storage 1005 and perform the following operations: Control the target AGV to move along the transport path, and receive environmental data collected by the target AGV during its movement; Based on the environmental data, it is determined whether there are obstacles in the transport path. If there are obstacles, the current position information of the target AGV is obtained, and the transport path is updated based on the current position information, the destination position information, and the environmental model.

[0067] Furthermore, the environmental data includes at least ultrasonic signals and environmental images, and the step of determining whether there are obstacles in the transport path based on the environmental data includes: Based on the ultrasonic signal, it is determined whether there is an obstacle to be identified in the transport path. If there is an obstacle to be identified, the environmental image is subjected to grayscale conversion, image correction and edge detection to generate a target image. The target image is identified based on a preset recognition model to determine the size of the obstacle to be identified; Based on the size of the object, it is determined whether the obstacle to be identified obstructs the movement of the target AGV. If it obstructs the movement of the target AGV, it is determined that there is an obstacle in the transport path.

[0068] Furthermore, before the step of obtaining the initial position information, destination position information, and environment model corresponding to the zinc ingot handling instruction upon receipt, the processor 1001 may call the control program stored in the memory 1005 and perform the following operations: Acquire a large number of historical environmental images corresponding to the transport path, and divide the historical environmental images into environmental training data and environmental verification data; The preset initial model is trained based on the environmental training data, and when the training reaches a preset duration, the preset initial model is validated based on the environmental validation data, generating a loss function value corresponding to the preset initial model. Determine whether the loss function value is less than a preset threshold. If it is less than the preset threshold, then generate the preset initial model as a preset recognition model.

[0069] Further, after the step of determining whether the loss function value is less than a preset threshold, the processor 1001 can call the control program stored in the memory 1005 and perform the following operations: If the loss function value is greater than or equal to a preset threshold, the model parameters of the preset initial model are updated, and the preset initial model is trained based on the environmental training data for the updated preset initial model.

[0070] The specific implementation of the AGV zinc ingot handling path planning system of the present invention is basically the same as the embodiments of the above-described AGV zinc ingot handling path planning method, and will not be repeated here.

[0071] This invention also proposes a medium. The medium is a readable storage medium storing a control program. When executed by a processor, the control program implements the steps of the AGV zinc ingot handling path planning method described above.

[0072] The storage medium of the present invention can be a computer-readable storage medium, and its implementation is basically the same as the embodiments of the above-described AGV zinc ingot handling path planning method, and will not be described again here.

[0073] The embodiments of the present invention have been described above with reference to the accompanying drawings. However, the present invention is not limited to the specific embodiments described above. The specific embodiments described above are merely illustrative and not restrictive. Those skilled in the art can make many modifications under the guidance of the present invention without departing from the spirit and scope of the claims. All equivalent structural or procedural transformations made using the content of the present invention specification and drawings, or direct or indirect applications in other related technical fields, are within the protection scope of the present invention.

Claims

1. A method for planning the transport path of zinc ingots using an AGV, characterized in that, The AGV zinc ingot handling path planning method includes: When a zinc ingot handling instruction is received, the initial position information, the destination position information, and the environment model corresponding to the zinc ingot handling instruction are obtained. Obtain the status identifier of each AGV, and determine the idle AGV that is in an idle state from among the AGVs based on the status identifier; Obtain the AGV position information of each of the idle AGVs, and plan the transportation path corresponding to the zinc ingot transportation instruction based on the initial position information, the destination position information, the environment model, and the position information of each AGV.

2. The AGV zinc ingot handling path planning method as described in claim 1, characterized in that, The step of planning the transport path corresponding to the zinc ingot transport instruction based on the initial position information, the destination position information, the environment model, and the position information of each AGV includes: Based on the initial location information, the destination location information, the environment model, and the location information of each AGV, a walking path corresponding to each of the idle AGVs is generated; For each walking path, a correction time factor corresponding to the walking path is determined, and the walking time corresponding to the walking path is determined based on the correction time factor. Based on the travel time corresponding to each of the travel paths, the transport path is determined from each of the travel paths, and the target AGV corresponding to the transport path is determined from each of the idle AGVs.

3. The AGV zinc ingot handling path planning method as described in claim 2, characterized in that, The step of determining the transport path from each of the walking paths based on the walking time corresponding to each walking path includes: Based on the walking time, determine the unoccupied walking paths in each of the walking paths; For each of the aforementioned idle walking paths, find the path length and the corresponding length-time weight value for that idle walking path; Based on the length-time weight value, the walking time corresponding to the idle walking path and the path length are weighted and calculated to generate the path factor of the idle walking path; After generating a corresponding path factor for each of the idle walking paths, the path factors are compared among themselves, and the transport path is determined based on the comparison results.

4. The AGV zinc ingot handling path planning method as described in claim 2, characterized in that, The step of generating a walking path corresponding to each of the idle AGVs based on the initial position information, the destination position information, the environment model, and the position information of each AGV includes: Based on the initial location information, the destination location information, and the environment model, a fixed path corresponding to the zinc ingot handling instruction is generated. For each idle AGV, a change path corresponding to the idle AGV is generated based on the AGV position information of the idle AGV, the initial position information, and the environment model; Based on the fixed path and each of the variable paths, a walking path corresponding to each of the idle AGVs is generated.

5. The AGV zinc ingot handling path planning method as described in claim 2, characterized in that, The step of determining the target AGV corresponding to the transport path from each of the idle AGVs includes: Control the target AGV to move along the transport path, and receive environmental data collected by the target AGV during its movement; Based on the environmental data, it is determined whether there are obstacles in the transport path. If there are obstacles, the current position information of the target AGV is obtained, and the transport path is updated based on the current position information, the destination position information, and the environmental model.

6. The AGV zinc ingot handling path planning method as described in claim 5, characterized in that, The environmental data includes at least ultrasonic signals and environmental images, and the step of determining whether there are obstacles in the transport path based on the environmental data includes: Based on the ultrasonic signal, it is determined whether there is an obstacle to be identified in the transport path. If there is an obstacle to be identified, the environmental image is subjected to grayscale conversion, image correction and edge detection to generate a target image. The target image is identified based on a preset recognition model to determine the size of the obstacle to be identified; Based on the size of the object, it is determined whether the obstacle to be identified obstructs the movement of the target AGV. If it obstructs the movement of the target AGV, it is determined that there is an obstacle in the transport path.

7. The AGV zinc ingot handling path planning method according to any one of claims 1-6, characterized in that, Before the step of obtaining the initial position information, destination position information, and environmental model corresponding to the zinc ingot handling instruction upon receiving the instruction, the following steps are included: Acquire a large number of historical environmental images corresponding to the transport path, and divide the historical environmental images into environmental training data and environmental verification data; The preset initial model is trained based on the environmental training data, and when the training reaches a preset duration, the preset initial model is validated based on the environmental validation data, generating a loss function value corresponding to the preset initial model. Determine whether the loss function value is less than a preset threshold. If it is less than the preset threshold, then generate the preset initial model as a preset recognition model.

8. The AGV zinc ingot handling path planning method as described in claim 7, characterized in that, The step of determining whether the loss function value is less than a preset threshold is followed by: If the loss function value is greater than or equal to a preset threshold, the model parameters of the preset initial model are updated, and the preset initial model is trained based on the environmental training data for the updated preset initial model.

9. An AGV zinc ingot handling path planning system, characterized in that, The AGV zinc ingot handling path planning system includes a storage device, a processor, a communication bus, and a control program stored in the storage device. The communication bus is used to enable communication between the processor and the memory. The processor is used to execute the control program to implement the steps of the AGV zinc ingot handling path planning method as described in any one of claims 1-8.

10. A medium, characterized in that, The medium is a readable storage medium, on which a control program is stored. When the control program is executed by a processor, it implements the steps of the AGV zinc ingot handling path planning method as described in any one of claims 1-8.

Images