A 12-axis PCB drilling machine AGV material flow scheduling system

Abstract

The present invention discloses a 12-axis PCB drilling machine AGV material flow scheduling system, which relates to the technical field of material scheduling. By collecting AGV task completion signals and workstation material perception information, extracting the time difference ΔT and matching degree M, and constructing joint features for accuracy analysis, the authenticity of the AGV task status can be effectively judged. When the judgment is inaccurate, statistical analysis is performed based on the mean and standard deviation of abnormal data over a period of time, and the scheduling strategy is dynamically adjusted. This technical solution significantly improves the status recognition accuracy and the stability of the scheduling system under multi-AGV concurrency, effectively avoids material dislocation, repeated handling and process confusion, and ensures the continuity, accuracy and traceability of the intelligent manufacturing process.

Description

Technical Field

[0001] The present invention relates to the technical field of material scheduling, and in particular to a 12-axis PCB drilling machine AGV material flow scheduling system. Background Art

[0002] 12-axis PCB drilling machine AGV material flow scheduling is a system solution that combines efficient production with intelligent logistics in the electronics manufacturing process. This system utilizes a 12-axis PCB drilling machine to achieve multi-station, high-precision circuit board drilling operations, while also incorporating AGVs (Automated Guided Vehicles) for automated material handling and scheduling. The scheduling system intelligently assigns AGV paths and tasks based on production plans and process requirements, enabling efficient connection between the drilling machine and materials. This improves production efficiency, reduces labor costs, and enhances the automation and intelligence of the entire manufacturing process.

[0003] However, in existing AGV scheduling systems, when an AGV completes material delivery and transmits its status back to trigger the next process, communication delays or concurrency control flaws can cause the scheduling system to misjudge the task's completion and prematurely initiate downstream processes, leading to material misplacement, duplicate handling, or detection errors. This problem is particularly insidious in high-speed, multi-AGV concurrent operations. Once it occurs, it can cause process disruptions, equipment misdetection, production line stalls, and even risk batch product quality and disrupt the data traceability chain, seriously impacting the stability and reliability of intelligent manufacturing systems. Summary of the Invention

[0004] The purpose of the present invention is to provide a 12-axis PCB drilling machine AGV material flow scheduling system to solve the shortcomings of the background technology.

[0005] In order to achieve the above-mentioned object, the present invention provides the following technical solutions: a 12-axis PCB drilling machine AGV material flow scheduling system, comprising a drilling control unit, an AGV scheduling control unit, a data acquisition module, a state accuracy assessment module, and an exception handling module;

[0006] Drilling control unit, used to control the 12-axis PCB drilling machine to perform multi-station parallel drilling operations and output drilling progress status information in real time;

[0007] An AGV scheduling control unit is used to generate dynamic transport tasks based on the drilling progress status information, AGV operation status information and process task requirements, and perform path planning and task allocation;

[0008] The data acquisition module is used to collect AGV task completion signals, AGV positions, unique identification information of transported materials, and material sensing information of target workstations in real time;

[0009] The status accuracy assessment module extracts the time difference ΔT between the AGV task completion signal and the arrival of the target workstation material from the collected data, as well as the matching degree M between the workstation sensed material and the AGV transport task material. By jointly analyzing ΔT and M, it evaluates whether the task completion status returned by the AGV is accurate.

[0010] The exception handling module is used to dynamically optimize the task execution and scheduling strategy when the state accuracy assessment module determines that the AGV state feedback is inaccurate.

[0011] Preferably, the AGV scheduling control unit includes: receiving processing completion information, AGV status information and downstream process task requirements output by the drilling control unit; judging whether there are materials to be transferred; screening idle AGVs for resources, judging their current location, load capacity, power and traffic conflict; calculating the transportation path for the AGV based on the A* or Dijkstra algorithm; generating transportation tasks and issuing execution instructions; monitoring the AGV execution progress in real time, and updating the scheduling queue.

[0012] Preferably, the time difference ΔT between the AGV task completion signal and the arrival of the target station material is extracted by: constructing an AGV task completion timestamp sequence , n is the total number of AGV task completion timestamps, and the target station material arrival timestamp sequence is constructed , m is the total number of arrival timestamps of materials at the target workstation;

[0013] Set the time window W, sort the event sequences F and A, and ensure that they are arranged in ascending time order; for each , build a candidate set: ; That is, filter out all candidate material arrival events within the window W from A; filter out the best match: ; Calculate the time difference ΔT, the expression is: .

[0014] Preferably, the matching degree M between the workstation perception material and the AGV transport task material is extracted. The extraction method is as follows: both the workstation perception system and the AGV task system can provide or extract a set of material attribute features, including product model, batch number, number of board layers, processing technology, and time tag; these attributes are converted into two sets for similarity matching, where A is set as the attribute set of the material in the AGV transport task; B is set as the attribute set of the material identified by the workstation perception system; the matching degree M is defined by the Jaccard similarity, which is expressed as: ;in: is the number of elements in the intersection of A and B; is the number of elements in the union of A and B.

[0015] Preferably, the time difference ΔT and the matching degree M are converted into comprehensive feature vectors, and the comprehensive feature vectors are used as inputs of the machine learning model. The machine learning model uses each group of comprehensive feature vectors to predict the task completion status accuracy analysis value label returned by the AGV as the prediction target, and uses minimizing the sum of the prediction errors of the task completion status accuracy analysis value labels returned by all AGVs as the training target. The machine learning model is trained until the sum of the prediction errors reaches convergence, and the model training is stopped. The task completion status accuracy analysis value returned by the AGV is determined according to the model output results, wherein the machine learning model is a polynomial regression model.

[0016] Preferably, the obtained accuracy analysis value of the task completion status returned by the AGV is compared with the preset accuracy threshold. If the accuracy analysis value of the task completion status returned by the AGV is greater than or equal to the preset accuracy threshold, it means that the task completion status returned by the AGV is accurate; if the accuracy analysis value of the task completion status returned by the AGV is less than the preset accuracy threshold, it means that the task completion status returned by the AGV is inaccurate.

[0017] Preferably, the exception handling module is used to dynamically optimize the task execution and scheduling strategy when the state accuracy assessment module determines that the AGV state feedback is inaccurate, specifically including:

[0018] When the accuracy analysis value of the task completion status returned by the AGV is less than the preset accuracy threshold, it means that the task completion status returned by the AGV is inaccurate. The accuracy analysis values of the task completion status generated within the subsequent fixed time period T that are less than the preset accuracy threshold are collected, and a data set is established. The mean and standard deviation of the data set are calculated, and after analysis, the task execution and scheduling strategies are dynamically optimized based on the analysis results.

[0019] Preferably, if the mean of the task completion status accuracy analysis values in the data set is greater than or equal to the reference threshold of the task completion status accuracy analysis value mean, and the standard deviation of the task completion status accuracy analysis value is less than the reference threshold of the task completion status accuracy analysis value standard deviation, it is necessary to increase the task retry threshold and enable the intervention rule;

[0020] If the mean value of the task completion status accuracy analysis value is greater than or equal to the reference threshold value of the mean value of the task completion status accuracy analysis value, and the standard deviation of the task completion status accuracy analysis value is greater than or equal to the reference threshold value of the standard deviation of the task completion status accuracy analysis value, check the specific AGV / station log and disable the abnormal device;

[0021] If the mean value of the task completion status accuracy analysis value is less than the reference threshold value of the task completion status accuracy analysis value mean value, and the standard deviation of the task completion status accuracy analysis value is greater than or equal to the reference threshold value of the task completion status accuracy analysis value standard deviation, the global task buffer mechanism is activated to reduce the scheduling frequency;

[0022] If the mean value of the task completion status accuracy analysis value is less than the reference threshold value of the mean value of the task completion status accuracy analysis value, and the standard deviation of the task completion status accuracy analysis value is less than the reference threshold value of the standard deviation of the task completion status accuracy analysis value, the emergency scheduling strategy is immediately triggered, and manual review or full network suspension is enabled.

[0023] In the above technical solution, the technical effects and advantages provided by the present invention are:

[0024] 1. This invention addresses the existing problem of AGV status misjudgment caused by communication delays or concurrent control defects. It establishes a closed-loop mechanism for the entire process, from drilling task completion and transport task generation to status collection, accuracy assessment, and exception handling. By introducing time difference ΔT and matching degree M as multidimensional feature inputs and combining them with a polynomial regression model to accurately predict AGV task status, this significantly improves the intelligence and reliability of scheduling decisions, effectively avoiding serious consequences such as process disruptions, repeated handling, and detection errors caused by misjudgment.

[0025] 2. This invention uses an exception handling module to dynamically monitor the accuracy of AGV task status during system operation. When anomalies are concentrated or system fluctuations are abnormal, it can intelligently classify and assess system operational stability based on the mean and standard deviation of the outliers, and automatically adjust the scheduling strategy, including disabling abnormal devices, task buffering, or pausing the entire network. This solution significantly enhances the AGV scheduling system's ability to respond to uncertainties in multi-AGV concurrent, high-paced manufacturing environments, improving production line stability, material traceability accuracy, and the operational safety of the entire intelligent manufacturing system. BRIEF DESCRIPTION OF THE DRAWINGS

[0026] In order to more clearly illustrate the embodiments of the present application or the technical solutions in the prior art, the following briefly introduces the drawings required for use in the embodiments. Obviously, the drawings described below are only some embodiments described in the present invention. For ordinary technicians in this field, other drawings can also be obtained based on these drawings.

[0027] Figure 1 This is a mind map of the system modules of the present invention. DETAILED DESCRIPTION

[0028] To make the objectives, technical solutions, and advantages of the embodiments of the present invention more clear, the technical solutions in the embodiments of the present invention will be clearly and completely described below in conjunction with the accompanying drawings of the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. All other embodiments obtained by ordinary technicians in this field based on the embodiments of the present invention without making any creative efforts shall fall within the scope of protection of the present invention.

[0029] For examples, see Figure 1 As shown, the 12-axis PCB drilling machine AGV material flow scheduling system described in this embodiment includes a drilling control unit, an AGV scheduling control unit, a data acquisition module, a state accuracy assessment module, and an exception handling module;

[0030] Drilling control unit, used to control the 12-axis PCB drilling machine to perform multi-station parallel drilling operations and output drilling progress status information in real time;

[0031] An AGV scheduling control unit is used to generate dynamic transport tasks based on the drilling progress status information, AGV operation status information and process task requirements, and perform path planning and task allocation;

[0032] The data acquisition module is used to collect AGV task completion signals, AGV positions, unique identification information of transported materials, and material sensing information of target workstations in real time;

[0033] The status accuracy assessment module extracts the time difference ΔT between the AGV task completion signal and the arrival of the target workstation material from the collected data, as well as the matching degree M between the workstation sensed material and the AGV transport task material. By jointly analyzing ΔT and M, it evaluates whether the task completion status returned by the AGV is accurate.

[0034] The exception handling module is used to dynamically optimize the task execution and scheduling strategy when the state accuracy assessment module determines that the AGV state feedback is inaccurate.

[0035] The drilling control unit is the core control module used to control the 12-axis PCB drilling machine to complete high-precision multi-station drilling tasks. It has the following key components and functional features:

[0036] The main control module is equipped with an industrial-grade controller (such as a PLC, motion control card, or embedded real-time system). It is used to receive process task parameters issued by the MES system or scheduling system, such as drilling drawings, drilling positions, hole sizes, board layer numbers, etc., and to control the 12 drilling axes in parallel, time-sharing, or synchronous manner.

[0037] The multi-axis motion control module enables independent motion trajectory control of 12 drilling spindles; control includes XY translation platform, Z-axis lifting system and spindle servo control; supports parallel processing of multi-station PCB boards, and has dynamic path compensation and avoidance mechanisms; and can support the parsing and execution of G-code or custom path instructions.

[0038] The drilling progress monitoring module records the current processing progress of each drilling axis in real time, including: the current drilling position number; the current drilling plate number and layer number; the current drilling completion rate (number of holes drilled / total number); and has a real-time status signal output function, which is transmitted to the scheduling system or upper-level MES system via the bus or IO signal.

[0039] Workpiece identification and workstation management module: Each workstation can be equipped with a material identification device (such as a QR code scanner or RFID reader) to identify the unique identifier of the PCB board to be processed; the processing status of each workstation is recorded, such as "to be processed", "processing", "processing completed", "abnormal pause", etc.

[0040] The fault detection and alarm module monitors drill wear, drilling deviation, shaft operation abnormality, excessive load, plate misalignment, etc. in real time; once an abnormality is detected, it is immediately reported through an alarm signal and the machine is shut down.

[0041] The data interface with the dispatch system provides REST API, OPC UA, Modbus, or custom communication protocol interfaces. It outputs drilling task execution status, drilling completion flag, workstation ID, plate ID, processing time, and more. It supports periodic (e.g., every 3 seconds) or event-triggered data reporting. When processing is completed at a workstation, a material-ready signal is sent to the AGV dispatch control unit. A timestamp and unique material ID can be embedded to ensure consistency in transportation and subsequent processes.

[0042] The AGV scheduling control unit is used to generate dynamic transportation tasks based on the drilling progress status information, AGV operation status information and process task requirements, and perform path planning and task allocation, specifically including:

[0043] The AGV dispatch control unit receives the following three types of core data input in real time:

[0044] The drilling control unit outputs drilling progress status information: including drilling task completion signal, workstation number, PCB board ID, processing completion timestamp, etc.

[0045] AGV status information: including the AGV's current coordinates, task status (idle / in progress / fault), current load, battery status, and estimated travel time from the current location to the destination.

[0046] Process task requirement information (provided by MES or task management module): including the next process type (AOI inspection, lamination, warehousing, etc.), target workstation number, priority level, material type requirements, etc.

[0047] Based on the acquired data, the AGV dispatch control unit executes the following logic:

[0048] Determine whether there are PCB boards that have been drilled but not yet removed. If so, generate a task item for transportation and associate the following key fields:

[0049] PCB board ID; starting location (drilling station); target process station; transportation time window and priority; material feature label (such as single-sided, double-sided, multi-layer board identification, etc.).

[0050] Retrieve a list of AGVs that are currently in the "idle" or "dispatchable" state; sort the available AGVs based on the following evaluation parameters:

[0051] The current distance to the target workstation; whether the load capacity matches; whether the current task priority conflicts; whether the battery is sufficient to complete the task; and the degree of overlap with other AGVs' current paths.

[0052] To generate a path plan for a candidate AGV, perform the following sub-steps:

[0053] Path generation algorithm: Use A*, Dijkstra, or D* Lite algorithms to calculate the shortest or optimal path on a preset map topology; taking into account obstacle avoidance areas, busy lanes, current dynamic traffic load, etc.

[0054] Path conflict detection: Determines whether the current path overlaps with other AGV paths in time or space; if so, automatically generates an alternative path or adjusts the order of travel. Generates a task execution package: This includes: starting point, destination point, path node sequence, speed limit, task ID, timeout, etc.

[0055] Issue transport task instructions to the selected AGV; monitor its task execution status in real time, including: path progress; whether it reaches the target point; whether the material is successfully delivered; and whether any anomalies occur (deviation from the path, communication interruption, equipment failure, etc.).

[0056] After the task is completed, according to the completion status information returned by the AGV: the task queue status in the scheduling system is updated; the downstream process site is triggered to prepare for reception; and the execution log and timestamp are written to the system for subsequent traceability analysis; if the status assessment module identifies an inaccurate status, the exception handling process is initiated (see the previous status feedback mechanism).

[0057] The data acquisition module is used to collect AGV task completion signals, AGV positions, unique identification information of transported materials, and material sensing information of target workstations in real time, including:

[0058] The data acquisition module is used to collect core data in real time and accurately during the execution of AGV transportation tasks. It is the basis for the scheduling system to achieve dynamic judgment, abnormality identification and precise control. This module usually includes multiple data interface submodules and collection devices, including:

[0059] The AGV task completion signal acquisition unit is connected to the AGV controller or scheduling system interface to collect task completion flag signals (such as task ID + end status + timestamp); the acquisition method can adopt industrial protocols such as Modbus, CAN, MQTT, OPC UA or RESTAPI; and different completion statuses can be distinguished: normal completion, timeout completion, failure exit, etc.

[0060] The AGV real-time position acquisition unit obtains the AGV's current XY coordinates, heading angle, running speed, etc.; data sources include: laser SLAM positioning system; QR code / magnetic stripe landmark reading; inertial navigation system; and supports linkage with the path planning module to achieve dynamic correction of positioning accuracy (≤±10cm).

[0061] The transport material unique identification information collection unit automatically collects the unique identification of the transported material at the AGV loading or unloading station. Common methods include: RFID tag recognition; QR code / barcode scanning; image recognition-assisted ID confirmation; binding the material ID to the current AGV task for subsequent tracking and error identification.

[0062] The target station material sensing information collection unit is a sensor device installed at the target station (such as AOI inspection, storage station, etc.) to determine whether the material has been successfully delivered. Common sensing methods include: photoelectric / laser sensors: detecting the presence of pallets or materials; weight sensors: identifying whether the material loading status matches; RFID / QR code readers: comparing material identities; completing signal comparison with AGV to verify task delivery.

[0063] All collected signals are uniformly transmitted to the status accuracy assessment module and scheduling control unit; they include timestamps, multi-dimensional parameters and task associations; the data format is standardized as structured data (JSON / XML), supporting real-time message push or database writes; and they have caching and breakpoint resumption mechanisms to ensure high-reliability data synchronization.

[0064] The status accuracy assessment module extracts the time difference ΔT between the AGV's task completion signal and the arrival of the target workstation material from the collected data, as well as the matching degree M between the workstation's perceived material and the AGV's transport task material. By jointly analyzing ΔT and M, the module assesses the accuracy of the task completion status reported by the AGV. Specifically, the module:

[0065] Receives the following inputs from the data acquisition module:

[0066] AGV task completion signal (including task ID and completion timestamp); unique material identifier of the AGV corresponding to the task (such as RFID, QR code); material perception information of the target workstation (arrival time, perception status, material identification result); system preset normal time difference threshold ΔT0 and matching degree threshold M0.

[0067] Extract the time difference ΔT between the AGV task completion signal and the arrival of the target workstation material. The extraction method is: Construct the AGV task completion timestamp sequence , n is the total number of AGV task completion timestamps, and the target station material arrival timestamp sequence is constructed (collected by the perception sensor) , m is the total number of arrival timestamps of materials at the target workstation;

[0068] Set the time window W, sort the event sequences F and A, and ensure that they are arranged in ascending time order; for each , build a candidate set: ; That is, filter out all candidate material arrival events within the window W from A; filter out the best match: ; Calculate the time difference ΔT, the expression is: ;

[0069] Set an abnormal threshold δ (such as 3 seconds). If ΔT<δ, it may be reported in advance, resulting in a false completion signal. If ΔT>δ, it may be due to delayed material arrival or perception failure.

[0070] To extract the matching degree M between the workstation sensing material and the AGV transport task material, the following extraction method is used: both the workstation sensing system and the AGV task system can provide or extract a set of material attribute features, such as: product model (e.g., PCB-8L); batch number (e.g., LOT20250501); number of board layers (e.g., 8L); processing technology (e.g., HASL, ENIG); and time stamp (e.g., 14:03). These attributes are then converted into two sets for similarity matching. Let A be the attribute set of the material in the AGV transport task; let B be the attribute set of the material identified by the workstation sensing system. The matching degree M is defined by the Jaccard similarity, expressed as: ;in: is the number of intersection elements of A and B (number of common features); is the number of elements in the union of A and B (the total number of feature types). The matching degree M∈[0,1], the closer it is to 1, the higher the matching degree.

[0071] If M ≥ 0.85: High match (accurate status); if 0.6 ≤ M < 0.85: Suspicious match (requires secondary verification); if M < 0.6: Match failure (error status, downstream startup prohibited).

[0072] The time difference ΔT and the matching degree M are converted into comprehensive feature vectors, and the comprehensive feature vectors are used as the input of the machine learning model. The machine learning model uses each set of comprehensive feature vectors to predict the task completion status accuracy analysis value label returned by the AGV as the prediction target, and takes minimizing the sum of the prediction errors of the task completion status accuracy analysis value labels returned by all AGVs as the training target. The machine learning model is trained until the sum of the prediction errors reaches convergence, and the model training is stopped. The task completion status accuracy analysis value returned by the AGV is determined according to the model output results. The machine learning model is a polynomial regression model.

[0073] The obtained task completion status accuracy analysis value returned by the AGV is compared with the preset accuracy threshold. If the task completion status accuracy analysis value returned by the AGV is greater than or equal to the preset accuracy threshold, it means that the task completion status returned by the AGV is accurate; if the task completion status accuracy analysis value returned by the AGV is less than the preset accuracy threshold, it means that the task completion status returned by the AGV is inaccurate.

[0074] The exception handling module is used to dynamically optimize the task execution and scheduling strategy when the state accuracy assessment module determines that the AGV state feedback is inaccurate.

[0075] When the accuracy analysis value of the task completion status returned by the AGV is less than the preset accuracy threshold, it means that the task completion status returned by the AGV is inaccurate. The accuracy analysis values of the task completion status generated within the subsequent fixed time period T that are less than the preset accuracy threshold are collected, and a data set is established. The mean and standard deviation of the data set are calculated, and after analysis, the task execution and scheduling strategies are dynamically optimized based on the analysis results.

[0076] If the mean of the task completion status accuracy analysis values in the data set is greater than or equal to the reference threshold of the task completion status accuracy analysis value mean, and the standard deviation of the task completion status accuracy analysis value is less than the reference threshold of the task completion status accuracy analysis value standard deviation, it means that the anomaly is slight and concentrated, with small fluctuations, and the task retry threshold needs to be increased, and the soft intervention rule needs to be enabled;

[0077] If the mean of the task completion status accuracy analysis value is greater than or equal to the reference threshold of the mean of the task completion status accuracy analysis value, and the standard deviation of the task completion status accuracy analysis value is greater than or equal to the reference threshold of the standard deviation of the task completion status accuracy analysis value, the abnormality is slight but the fluctuation is large, and individual AGVs or workstations are unstable. Check the specific AGV / workstation logs and dynamically disable the abnormal equipment;

[0078] If the mean value of the task completion status accuracy analysis value is less than the reference threshold value of the task completion status accuracy analysis value mean value, and the standard deviation of the task completion status accuracy analysis value is greater than or equal to the reference threshold value of the task completion status accuracy analysis value standard deviation, the system status is low and fluctuates greatly, and there is widespread instability. The global task buffer mechanism is activated to reduce the scheduling frequency;

[0079] If the mean value of the task completion status accuracy analysis value is less than the reference threshold value of the mean value of the task completion status accuracy analysis value, and the standard deviation of the task completion status accuracy analysis value is less than the reference threshold value of the standard deviation of the task completion status accuracy analysis value, it means that the abnormalities are concentrated and serious, and the system is generally low. The emergency scheduling strategy is immediately triggered, and manual review or network-wide suspension is enabled.

[0080] Feed the above classification results back to the scheduling system and dynamically adjust the following policy parameters:

[0081] AGV task allocation priority; path scheduling interval; re-dispatching threshold and judgment rules; temporary exclusion of faulty AGVs / sensing stations; and startup delay time of downstream processes.

[0082] The above formulas are all dimensionless and numerical calculations. The formulas are obtained by collecting a large amount of data and performing software simulation to obtain the most recent real situation. The preset parameters in the formulas are set by technicians in this field according to actual conditions.

[0083] The above embodiments can be implemented in whole or in part by software, hardware, firmware or any other combination. When implemented using software, the above embodiments can be implemented in whole or in part in the form of a computer program product. The computer program product includes one or more computer instructions or computer programs. When the computer instructions or computer program are loaded or executed on a computer, the process or function described in the embodiment of the present application is generated in whole or in part. The computer can be a general-purpose computer, a special-purpose computer, a computer network, or other programmable device. The computer instructions can be stored in a computer-readable storage medium or transmitted from one computer-readable storage medium to another computer-readable storage medium. For example, the computer instructions can be transmitted from one website, computer, server or data center to another website, computer, server or data center via a wired (e.g., infrared, wireless, microwave, etc.) method. The computer-readable storage medium can be any available medium that can be accessed by a computer or a data storage device such as a server or data center that contains one or more available media sets. The available medium can be a magnetic medium (e.g., a floppy disk, a hard disk, a tape), an optical medium (e.g., a DVD), or a semiconductor medium. The semiconductor medium can be a solid-state drive.

[0084] The above is only a specific implementation method of the present application, but the scope of protection of the present application is not limited thereto. Any technician familiar with this technical field can easily think of changes or replacements within the technical scope disclosed in this application, which should be covered by the scope of protection of the present application.

Claims

1. A 12-axis PCB drilling machine AGV material flow scheduling system, characterized by: It includes drilling control unit, AGV scheduling control unit, data acquisition module, state accuracy assessment module and exception handling module; Drilling control unit, used to control the 12-axis PCB drilling machine to perform multi-station parallel drilling operations and output drilling progress status information in real time; An AGV scheduling control unit is used to generate dynamic transport tasks based on the drilling progress status information, AGV operation status information and process task requirements, and perform path planning and task allocation; The data acquisition module is used to collect AGV task completion signals, AGV positions, unique identification information of transported materials, and material sensing information of target workstations in real time; The status accuracy assessment module extracts the time difference ΔT between the AGV task completion signal and the arrival of the target workstation material from the collected data, as well as the matching degree M between the workstation sensed material and the AGV transport task material. By jointly analyzing ΔT and M, it evaluates whether the task completion status returned by the AGV is accurate. Specifically, it includes: extracting the time difference ΔT between the AGV task completion signal and the arrival of the target workstation material. The extraction method is: constructing the AGV task completion timestamp sequence , n is the total number of AGV task completion timestamps, and the target station material arrival timestamp sequence is constructed , m is the total number of timestamps of materials arriving at the target station; set the time window W, sort the event sequences F and A, and ensure that they are arranged in ascending order of time; for each , build a candidate set: ; That is, filter out all candidate material arrival events within the window W from A; filter out the best match: ; Calculate the time difference ΔT, the expression is: ; The exception handling module is used to dynamically optimize the task execution and scheduling strategy when the state accuracy assessment module determines that the AGV state feedback is inaccurate; Specifically, when the accuracy analysis value of the task completion status returned by the AGV is less than the preset accuracy threshold, it indicates that the task completion status returned by the AGV is inaccurate. The accuracy analysis values of the task completion status that are less than the preset accuracy threshold generated in the subsequent fixed time period T are collected, and a data set is established. The mean and standard deviation of the data set are calculated, and after analyzing the data set, the task execution and scheduling strategy is dynamically optimized based on the analysis results. If the mean of the task completion status accuracy analysis values in the data set is greater than or equal to the reference threshold of the task completion status accuracy analysis value mean, and the standard deviation of the task completion status accuracy analysis value is less than the reference threshold of the task completion status accuracy analysis value standard deviation, it is necessary to increase the task retry threshold and enable the intervention rule; If the mean value of the task completion status accuracy analysis value is greater than or equal to the reference threshold value of the mean value of the task completion status accuracy analysis value, and the standard deviation of the task completion status accuracy analysis value is greater than or equal to the reference threshold value of the standard deviation of the task completion status accuracy analysis value, check the specific AGV / station log and disable the abnormal device; If the mean value of the task completion status accuracy analysis value is less than the reference threshold value of the task completion status accuracy analysis value mean value, and the standard deviation of the task completion status accuracy analysis value is greater than or equal to the reference threshold value of the task completion status accuracy analysis value standard deviation, the global task buffer mechanism is activated to reduce the scheduling frequency; If the mean value of the task completion status accuracy analysis value is less than the reference threshold value of the mean value of the task completion status accuracy analysis value, and the standard deviation of the task completion status accuracy analysis value is less than the reference threshold value of the standard deviation of the task completion status accuracy analysis value, the emergency scheduling strategy is immediately triggered, and manual review or full network suspension is enabled.

2. The 12-axis PCB drilling machine AGV material flow scheduling system according to claim 1 is characterized by: The AGV scheduling control unit includes: receiving processing completion information, AGV status information and downstream process task requirements output by the drilling control unit; determining whether there is material to be transferred; screening idle AGV resources to determine their current location, load capacity, power and traffic conflict; calculating the transportation path for the AGV based on the A* or Dijkstra algorithm; generating transportation tasks and issuing execution instructions; monitoring the AGV execution progress in real time and updating the scheduling queue.

3. The 12-axis PCB drilling machine AGV material flow scheduling system according to claim 1 is characterized by: Extract the matching degree M between the workstation perception material and the AGV transport task material. The extraction method is as follows: both the workstation perception system and the AGV task system can provide or extract a set of material attribute features, including product model, batch number, number of board layers, processing technology, and time tag; convert these attributes into two sets for similarity matching, where A is the attribute set of the material in the AGV transport task; B is the attribute set of the material identified by the workstation perception system; the matching degree M is defined by the Jaccard similarity, expressed as: ;in: is the number of elements in the intersection of A and B; is the number of elements in the union of A and B.

4. The 12-axis PCB drilling machine AGV material flow scheduling system according to claim 3 is characterized by: The time difference ΔT and the matching degree M are converted into comprehensive feature vectors, and the comprehensive feature vectors are used as the input of the machine learning model. The machine learning model uses each set of comprehensive feature vectors to predict the task completion status accuracy analysis value label returned by the AGV as the prediction target, and takes minimizing the sum of the prediction errors of the task completion status accuracy analysis value labels returned by all AGVs as the training target. The machine learning model is trained until the sum of the prediction errors reaches convergence, and the model training is stopped. The task completion status accuracy analysis value returned by the AGV is determined according to the model output results. The machine learning model is a polynomial regression model.

5. The 12-axis PCB drilling machine AGV material flow scheduling system according to claim 4, characterized in that: The obtained task completion status accuracy analysis value returned by the AGV is compared with the preset accuracy threshold. If the task completion status accuracy analysis value returned by the AGV is greater than or equal to the preset accuracy threshold, it means that the task completion status returned by the AGV is accurate; if the task completion status accuracy analysis value returned by the AGV is less than the preset accuracy threshold, it means that the task completion status returned by the AGV is inaccurate.

Images