Intelligent task allocation method based on multi-device state coupling analysis

Abstract

The invention provides a laser cutting production line task automatic allocation control method based on multi-device state coupling analysis, and belongs to the technical field of intelligent manufacturing and industrial automation. The method comprises the following steps: constructing a dynamic closed-loop control process through a central control scheduling system: receiving a task information packet of an MES; constructing an equipment state vector based on a real-time station state, and introducing a dynamic weight factor set to generate a weighted state vector; performing task triggering judgment through a coupling triggering judgment function in combination with the task dependency graph and historical task records; when the conditions are met, a task instruction is issued to the target station; and feeding back the state and updating the historical record after the task is completed. The invention further relates to AGV intelligent scheduling, visual positioning compensation, process parameter dynamic adjustment, predictive conflict detection, weight self-optimization and the like. According to the method, the production line cooperation efficiency is remarkably improved, manual intervention and system delay are reduced, and the method is suitable for an intelligent laser processing scene in which multiple devices run in parallel.

Description

Technical Field

[0001] This invention relates to the field of intelligent manufacturing and industrial automation technology, and in particular to an intelligent task allocation method based on multi-device state coupling analysis. Background Technology

[0002] Currently, laser cutting technology is widely used in the precision processing of metal sheets. However, most existing production lines adopt semi-automation or distributed control modes, and the collaborative scheduling between various processes mainly relies on manual task instructions or static process control, which presents the following technical problems:

[0003] (1) Low task dispatch efficiency: Traditional MES systems have difficulty dynamically dispatching tasks based on the real-time status of equipment, resulting in process waiting, resource waste and unstable processing cycle;

[0004] (2) Weak system perception capability: The operating status of each device cannot be uniformly perceived, and the status feedback is delayed or inaccurate, which can easily lead to task conflicts or execution interruptions.

[0005] (3) Poor task coordination and coupling: There is a lack of automatic judgment mechanism based on state coupling between processes such as inkjet printing, cutting, and sorting, which makes it impossible to achieve seamless task connection;

[0006] (4) Material flow is disconnected from AGV scheduling: The existing AGV system lacks linkage with the task execution status, and the path planning relies on preset logic, making it difficult to adapt to changes in the production line rhythm.

[0007] Therefore, there is an urgent need for an intelligent task allocation method with multi-device status perception and collaborative analysis capabilities to achieve automated task issuance and material flow control, thereby improving the overall efficiency and intelligence level of the laser production line. Summary of the Invention

[0008] In view of the technical defects and drawbacks existing in the prior art, the embodiments of the present invention provide an intelligent task allocation method based on multi-device state coupling analysis to overcome the above problems or at least partially solve the above problems. The specific solution is as follows;

[0009] A smart task allocation method based on multi-device state coupling analysis is applied to a laser cutting production line including a coding station, a cutting station, a sorting station, a grinding station, and an AGV transportation system. The method is executed by a central control scheduling system and is implemented by constructing a dynamic closed-loop control process driven by task information packages. This process includes the following steps in sequence:

[0010] Task Information Reception and Parsing: The central control and scheduling system receives and parses task information packets from the Manufacturing Execution System (MES), extracting key information including at least the task number, process document path, and material parameters; the task information packet is the sole basis for subsequent workstation equipment to perform processing operations and for the central control and scheduling system to perform task scheduling;

[0011] Weighted Equipment State Vector Construction: Based on the real-time status information collected from each workstation on the production line, the central control and scheduling system constructs an equipment state vector S = [S1, S2, ..., S...]. n Simultaneously, a dynamically adjustable set of state weight factors W = [w1, w2, ..., w] is introduced. n ], generate a weighted device state vector S' = W · S, which is used to quantify the real-time resource matching capability required by the system to execute the current task;

[0012] Coupling Trigger Determination: The central control and scheduling system inputs the weighted equipment state vector S', the predefined task dependency graph T determined by the current task information package, and the historical task execution record H which records equipment performance and past task execution status into the coupling trigger determination function F(S', T, H), and outputs a Boolean determination result on whether a specific process task in the current task information package is triggered.

[0013] Task instruction issuance: If the output of the coupled trigger determination function F(S', T, H) for a certain process task is a trigger, and the operation permission flag of the target station is true, then the central control scheduling system issues a task instruction containing the corresponding process file path and parameters to the target station.

[0014] Status feedback and historical record update: After the task is completed, the executing device will send the task completion status feedback to the central control and scheduling system; the central control and scheduling system will update the task execution information to the historical task execution record H, forming a dynamic learning closed loop based on execution feedback.

[0015] This invention significantly improves the automation level of production line scheduling and reduces manual intervention through a closed-loop process of "receiving tasks → state perception → coupling judgment → instruction issuance → feedback update"; it uses weighted equipment state vector S' to quantify resource matching capabilities, making task triggering more accurate and reducing the risk of resource idleness or conflict; and it forms a learning loop through real-time updates of historical records H, providing a data foundation for system self-optimization.

[0016] Furthermore, in the coupling trigger determination step, the coupling trigger determination function F(S′,T,H) outputs a Boolean determination result indicating whether a specific process task in the current task information package is triggered, through a hierarchical determination model integrating static rules and dynamic analysis; the determination logic of this model covers the following levels:

[0017] Rule-based decision layer based on mapping table: The function internally maintains a state combination mapping table, which combines the system state represented by the weighted device state vector S' with the preorder and successor dependencies between tasks defined by the task dependency graph T, and maps them to the following specific task triggering conditions:

[0018] a) Triggering conditions for inkjet printing tasks: If the task in the task information package is the first task, the triggering condition is: the inkjet printing station status is represented as an acceptable task in S' after weight calculation, and its operation permission flag is true; if it is not the first task, the triggering condition also needs to be added: the inkjet printing status corresponding to the previous task is represented as completed in S'.

[0019] b) Cutting task triggering conditions: If the task in the task information package is the first task, the triggering conditions are: the status of the preceding inkjet printing task is represented as completed, and the status of the cutting station is represented as an acceptable task in S' after weight calculation, and its running permission flag is true; if it is not the first task, the triggering conditions also need to be added: the cutting status corresponding to the preceding task is represented as completed in S'.

[0020] c) Sorting task triggering conditions: If the task in the task information package is the first task, the triggering conditions are: the status of the preceding inkjet printing and cutting tasks are both represented as completed, and the sorting station status is represented as an acceptable task in S' after weight calculation, and its operation permission flag is true; if it is not the first task, the triggering conditions also need to be added: the status of all station tasks corresponding to the preceding task are represented as completed in S'.

[0021] The predictive conflict determination layer based on historical data: For tasks that meet the rules of the mapping table, the function further integrates the historical task execution records H for forward-looking analysis. Based on the equipment load curve and AGV scheduling log, it predicts whether there is a risk of resource conflict in the future task execution cycle. If the prediction model indicates that there is a high probability risk such as the target workstation is about to fail, the preceding task may time out, or there is a conflict in the AGV path, then the positive determination of the mapping table layer is rejected, and a negative determination (or false determination) is finally output to trigger delayed scheduling. Otherwise, a positive determination (or true determination) is finally output.

[0022] This hierarchical decision-making model combines explicit business rules with data-driven predictive analysis to ensure the correctness of process logic and further avoid potential conflicts, thereby achieving highly reliable and intelligent task triggering decisions.

[0023] This invention employs a layered model consisting of a "static rule layer + predictive conflict determination layer" to ensure the correctness of business process logic while proactively mitigating risks such as equipment failure and AGV path conflicts. By dynamically rejecting high-risk task triggers, it avoids chain shutdowns caused by potential failures and enhances system robustness.

[0024] Furthermore, in the weighted device state vector construction step, the central control scheduling system uses its integrated predictive conflict detection module to make a forward-looking judgment on task execution risks, specifically including:

[0025] The predictive conflict detection module uses the historical task execution record H as the core analysis data source, performs time series analysis and pattern recognition on the historical task execution time distribution, equipment load curve and AGV scheduling log contained therein, and constructs a short-term prediction model to predict potential task risks in the next scheduling cycle.

[0026] If the prediction model identifies any of the following potential abnormal trends that reach a preset risk threshold, the predictive conflict detection module will output a warning signal to the coupling trigger determination step, thereby triggering task blocking or scheduling delay logic in advance to achieve proactive fault avoidance:

[0027] a) Based on abnormal fluctuation patterns in equipment vibration, temperature, or power data, predict the probability of a target workstation failing within future task cycles.

[0028] b) Based on the execution history and current progress of the preceding task type, determine the probability that the currently executing preceding task is expected to time out;

[0029] c) Based on the path planning and real-time location information in the AGV scheduling log, predict the risk of traffic conflicts or scheduling bottlenecks in the AGV transportation path in the future period.

[0030] d) Based on the system communication logs, determine whether the network communication delay or data packet loss rate may deteriorate to a level that affects the reliability of the real-time decision-making of the central control and scheduling system in the next cycle.

[0031] This invention constructs a short-term prediction model based on historical data (such as equipment vibration and AGV logs) to proactively block high-risk tasks, realizing a shift from "passive response" to "proactive prevention"; it also provides early warnings of hidden risks such as communication delays and load timeouts, maintaining the reliability of decision-making.

[0032] Furthermore, between the task instruction issuance step and the status feedback step, the scheduling strategy of the AGV transportation system, as a key material flow link, is closely linked to the result of the coupling trigger determination step; specifically, after receiving a request for transporting a fully loaded pallet triggered by the sorting station completing sorting, the central control scheduling system executes the following intelligent scheduling actions:

[0033] a) Direct transfer logic: The central control scheduling system first queries the status representation of the grinding station in the weighted equipment status vector S′; if there is a grinding station with a status representation of "idle" and a running permission flag of true, then an AGV scheduling instruction is immediately generated to transport the full-load pallet directly from the sorting station to the idle grinding station.

[0034] b) Cache temporary storage logic: If the weighted equipment state vector S′ shows that all grinding stations are in a "busy" or "faulty" state, an AGV scheduling instruction is generated to transport the fully loaded pallet to the temporary buffer area for temporary storage and record the cache status of the pallet.

[0035] c) Secondary scheduling logic: The central control scheduling system continuously monitors the status of the grinding station; when the status of any grinding station changes from "busy" to "idle", the system immediately searches the temporary buffer area. If there is a waiting pallet, the secondary scheduling instruction is immediately triggered to dispatch the AGV to transport the pallet from the buffer area to the new idle grinding station.

[0036] The path planning of the AGV scheduling instructions is based on a multi-vehicle collaborative path optimization algorithm. This algorithm comprehensively considers the AGV scheduling log in the historical task execution record H, real-time path occupancy information, and physical layout between devices to dynamically calculate the optimal transportation path and timing in order to avoid conflicts and maximize transportation efficiency.

[0037] This invention dynamically switches between "direct transfer → buffer storage → secondary scheduling" strategies based on the status of the grinding station to avoid AGV blockage or station waiting, thereby improving material flow efficiency. Combined with a multi-vehicle collaborative algorithm, it dynamically plans routes to reduce traffic conflicts and maximize transportation throughput.

[0038] Furthermore, in the task instruction issuance step, after the coupling trigger determination function F(S′,T,H) determines that the inkjet printing task triggering condition is met, the task instruction issued to the inkjet printing station drives its integrated vision positioning compensation system to perform the following precise positioning process. This process is a key preprocessing operation to ensure that the graphic pointed to by the inkjet printing file path in the task information package can match the actual board position with high precision:

[0039] The visual positioning compensation system first acquires images of the board material located at the inkjet printing station using an industrial camera, and automatically identifies the edge contour of the board material in the image and the preset positioning mark. Then, it calculates the positional deviation between the identified actual feature position and the preset theoretical coordinates in the task information package using an image processing algorithm. Based on the calculated deviation value, it corrects the output coordinates of the inkjet printing file in real time to compensate for the cumulative error generated during the transmission and positioning of the board material, thereby achieving high-precision matching between the inkjet printing graphic and the actual position of the board material.

[0040] This invention compensates for the positioning error of the board material through image recognition and coordinate correction, solves the problem of inkjet misalignment caused by cumulative transmission error, and improves processing accuracy.

[0041] Furthermore, in the task instruction issuance step, after the coupling trigger determination function F(S′,T,H) determines that the cutting task triggering condition is met, the task instruction issued to the cutting station drives the corresponding CNC laser cutting machine to perform the following key preprocessing operations:

[0042] The CNC laser cutting machine parses the material and thickness parameters of the sheet metal contained in the task information package from the received task instructions. Based on the parsed sheet metal characteristics, it dynamically adjusts the key process parameters on which the cutting operation depends through a built-in process parameter library or adaptive algorithm. These parameters include at least cutting power, cutting frequency, and cutting focal length. This dynamic adjustment mechanism aims to ensure that the energy output and focusing state of the laser cutting are precisely matched with the physical characteristics of the current sheet metal, thereby optimizing the cutting effect from the beginning of task execution and improving the cross-sectional quality of the cutting edge and the dimensional accuracy of the parts.

[0043] This invention dynamically adjusts parameters such as power and focal length according to the material and thickness of the sheet metal to ensure that the cutting energy matches the material properties, thereby improving the cross-sectional quality and dimensional consistency.

[0044] Furthermore, in the task instruction issuance step, after the coupling trigger determination function F(S′,T,H) determines that the sorting task trigger condition is met, the task instruction issued to the sorting station drives the corresponding automatic sorting machine to execute the following vision-based intelligent sorting process. This process is the core technical link for achieving accurate identification and classification of various irregular parts after cutting:

[0045] The automated sorting machine activates its industrial vision recognition system, using an industrial camera to capture images of the cut sheets of board located at the sorting station. Image recognition technology is then used to process the images, automatically extracting the contour information of each cut part. A crucial coordinate transformation operation is then performed, converting the two-dimensional pixel coordinates of the identified part contours in the image into operational coordinates for controlling the sorting robot to grasp and move in three-dimensional physical space, based on pre-calibrated transformation parameters. This transformation process ensures that the sorting robot can accurately identify and reliably grasp scattered or irregular parts based on the spatial positioning information provided by the vision system, and then classify and place them according to their type or order requirements.

[0046] This invention extracts the outline of parts and transforms their coordinates using a vision system, guiding a robotic arm to pick up scattered parts, reducing sorting error rates, and adapting to scenarios with irregular parts.

[0047] Furthermore, in the status feedback and historical record update steps, after the grinding equipment corresponding to the grinding station completes the deburring of the part edges, it performs the following quality assessment and data feedback process. This process is a key step in realizing closed-loop monitoring of processing quality and providing data support for the optimization of subsequent processes:

[0048] The grinding equipment collects the status data of the processed parts through an integrated high-frequency vibration sensor or industrial vision inspection system; by analyzing the characteristic spectrum of the vibration signal or the edge contour features in the visual image, it automatically determines whether the grinding effect meets the standard, and generates an evaluation result containing the grinding completion signal and the processing quality level, and uploads the evaluation result to the central control and scheduling system.

[0049] Upon receiving the result, the central control and scheduling system executes the following: It uses the grinding completion signal as key status information to drive downstream processes (such as triggering an AGV to transport materials to the storage area); and it updates the historical task execution record H with information such as the task information package number, the corresponding processing quality level, and the execution duration of this grinding task. This data provides direct data support for analyzing production line processing quality bottlenecks, optimizing grinding process parameters, and even adjusting the state weight factor set W to more accurately reflect equipment processing capabilities, thereby strengthening the self-optimization capability of the dynamic closed-loop control process.

[0050] This invention utilizes vibration sensors or visual inspection to automatically evaluate grinding quality, forming a closed loop of "processing-evaluation-optimization," providing a direct basis for adjusting process parameters; quality data is stored in historical records H for analyzing bottlenecks, optimizing weighting factors, and enhancing the system's self-optimization capabilities.

[0051] Furthermore, in the coupling trigger determination step, the central control scheduling system is equipped with a task priority adjustment mechanism to resolve resource contention when multiple tasks simultaneously meet the triggering conditions as determined by the coupling trigger determination function F(S′,T,H). The specific operation logic includes:

[0052] When the central control scheduling system detects that the triggering conditions for multiple tasks are met simultaneously, the priority adjustment mechanism will be activated immediately to dynamically sort these tasks according to a set of predefined priority rules; the sorting rules comprehensively consider at least one of the following factors:

[0053] The task sequence number defined in the task information package;

[0054] The processing urgency implied by the type of sheet material specified in the task information package;

[0055] Process urgency labels set by the Manufacturing Execution System (MES) or operators;

[0056] The central control scheduling system dynamically determines the order in which tasks are issued based on the sorting results, prioritizing the issuance of the highest priority task instructions to the workstations. This mechanism ensures that even when system resources (such as key equipment workstations) are limited, tasks can be intelligently allocated according to preset strategies (such as ensuring production line cycle time and prioritizing the processing of valuable or urgent materials), thereby optimizing the overall efficiency of the production line and the achievement of key performance indicators.

[0057] When multiple tasks are triggered, this invention dynamically sorts them according to rules such as task number and material urgency, prioritizing critical tasks and ensuring overall production line efficiency and KPI achievement.

[0058] Furthermore, the set of state weight factors W = [w1, w2, ..., w n The dynamic adjustment of [the weight] is achieved through a self-optimization module that integrates a periodic triggering mechanism and multi-dimensional performance index analysis. The self-optimization module completes the calculation and updating of weights through the following steps:

[0059] Periodic triggering and data acquisition steps: The module sets an update cycle based on the number of tasks (e.g., every N task cycles completed, where N is a preset positive integer). When the cycle condition is met, the optimization process is automatically started. After triggering, the module accesses the historical task execution record H and collects the running data of each workstation in the most recent cycle.

[0060] Multi-dimensional performance index quantification and analysis steps: The module performs quantitative analysis on the collected data and calculates multi-dimensional performance indices for each workstation i. These indices include at least the following:

[0061] Historical task efficiency index Ei Calculated based on the deviation between the actual execution time of the corresponding workstation task and the standard working time;

[0062] Equipment load rate index L i Calculated based on the ratio of the actual running time of the corresponding workstation to the total running time within the statistical period;

[0063] Equipment failure rate index F i Calculated based on the ratio of the frequency of failures to the runtime of the corresponding workstation within the statistical period;

[0064] The dynamic calculation steps for weighting factors are as follows: The module sets a basic weight for each workstation and dynamically adjusts it based on its quantified performance indicators; for workstation i, its new weighting factor w i The calculation of (new) follows the following optimization principles:

[0065] If E i If the display efficiency is higher than the system average, then the efficiency will be increased according to preset rules. i (new) to prioritize task allocation;

[0066] If L i If the display shows a consistently high load, then reduce the load according to the preset rules. i (new) to achieve load balancing;

[0067] If F i If the reliability of the display decreases, then w will be reduced according to the preset rules. i (new) to mitigate production risks;

[0068] Normalization and update implementation steps: Normalize all calculated new weight factors to ensure their sum is 1, generating a new generation of state weight factor set W. new The system uses W new Atomically replace the current set of weight factors W current This is then immediately applied to the construction of the weighted device state vector S' for subsequent tasks.

[0069] This self-optimization module uses periodic data-driven decision-making to enable the weight factor set W to adaptively reflect the latest performance status of the production line equipment, thereby ensuring that the judgment logic of the weighted equipment state vector S' and the coupled trigger judgment function F(S',T,H) remains optimal, ultimately achieving self-evolution and improvement of production line scheduling efficiency.

[0070] For example:

[0071] E i = (Total Standard Working Hours - |Total Actual Working Hours - Total Standard Working Hours|) / Total Standard Working Hours;

[0072] The closer the value is to 1, the higher the efficiency; a value greater than 1 indicates that the efficiency exceeds expectations.

[0073] L i = Actual runtime of workstation i within the statistical period / Total duration of the statistical period;

[0074] F i = Number of failures at workstation i during the statistical period / Total number of operating hours at workstation i during the statistical period;

[0075] If E i If the efficiency is greater than 1.05 (meaning the efficiency of workstation i is more than 5% higher than the standard level), then the efficiency will be increased proportionally.

[0076] Increase weight, for example, w i (new) = w i (current) * [1 + (E i - 1) * α], where α is the gain coefficient (e.g., set to 0.1), meaning that for every 1% increase in efficiency above the baseline, the weight increases by 0.1%.

[0077] If L i If the value is greater than 0.85 (i.e., the load factor exceeds 85%), the weight will be reduced proportionally, for example, w i (new)

[0078] = w i (current) * [1 - (L i [- 0.85) * β], where β is the attenuation coefficient (e.g., set to 0.5), meaning that for every 1% increase in load rate above the threshold, the weight decreases by 0.5% to achieve load balancing.

[0079] If F i If the value is > 0.005 (i.e., more than 1 failure per 200 hours of operation), the weight should be significantly reduced, for example, w i (new) = w i (current) * γ, where γ is a small penalty factor (such as 0.7 or lower), which directly reduces the weight of the workstation by 70% to significantly mitigate production risks.

[0080] Assuming a certain cutting station is in the current cycle:

[0081] E i = 1.08 (efficiency 8% higher than baseline), L i = 0.88 (load rate 88%), F i = 0.002 (Good reliability);

[0082] Current weight wi (current) = 0.25;

[0083] According to the rules, only efficiency improvements and load balancing adjustments are triggered:

[0084] Efficiency improvement adjustment: 0.25 * (1 + (1.08-1)*0.1) = 0.25 * 1.008 = 0.252

[0085] Load balancing adjustment: 0.252 * (1 - (0.88-0.85)*0.5) = 0.252 * (1 - 0.015) = 0.252 * 0.985 ≈ 0.248

[0086] After normalization, the new weight of this workstation may drop slightly to around 0.248, reflecting that load balancing has a slightly higher priority than efficiency rewards.

[0087] This invention periodically adjusts the weight factor set W based on indicators such as efficiency E, load rate L, and failure rate F, so that the scheduling strategy continuously matches the actual state of the equipment; through data-driven weight updates, the scheduling logic achieves self-iteration and improves production line efficiency in the long term.

[0088] The present invention has the following beneficial effects:

[0089] 1. By adopting a closed-loop process of "receiving tasks → status awareness → coupling judgment → instruction issuance → feedback update", the automation level of production line scheduling has been significantly improved, and manual intervention has been reduced.

[0090] 2. By utilizing the weighted device state vector S' to quantify resource matching capabilities, task triggering becomes more accurate, reducing the risk of resource idleness or conflict.

[0091] 3. By updating the historical record H in real time, a learning loop is formed, providing a data foundation for the system's self-optimization. Attached Figure Description

[0092] Figure 1 This is a system framework diagram of an automatic task allocation method for laser cutting production lines based on multi-device state coupling analysis provided by an embodiment of the present invention;

[0093] Figure 2 This is a schematic diagram of the interaction logic between the central control and dispatching system and the automatic conveyor line provided in an embodiment of the present invention;

[0094] Figure 3 This is a communication flowchart between the central control and scheduling system and the marking inkjet printer provided in this embodiment of the invention;

[0095] Figure 4This is a schematic diagram illustrating the interaction between the central control and scheduling system and the CNC laser cutting machine provided in an embodiment of the present invention;

[0096] Figure 5 This is a flowchart illustrating the interaction between the central control scheduling system and the automatic sorting machine provided in this embodiment of the invention.

[0097] Figure 6 This is a flowchart of the AGV material scheduling process provided in an embodiment of the present invention;

[0098] Figure 7 This is a flowchart illustrating the automatic distribution logic of inkjet printing tasks provided in this embodiment of the invention.

[0099] Figure 8 This is a flowchart illustrating the automatic distribution logic of cutting tasks provided in an embodiment of the present invention.

[0100] Figure 9 This is a flowchart of the automatic distribution of sorting tasks provided in an embodiment of the present invention. Detailed Implementation

[0101] To enable those skilled in the art to better understand the technical solutions of the present invention, exemplary embodiments of the present invention are described below in conjunction with the accompanying drawings, including various details of the embodiments of the present invention to aid understanding. These should be considered merely exemplary. Therefore, those skilled in the art should recognize that various changes and modifications can be made to the embodiments described herein without departing from the scope and spirit of the present invention. Similarly, for clarity and brevity, descriptions of well-known functions and structures are omitted in the following description.

[0102] Where there is no conflict, the various embodiments of the present invention and the features thereof may be combined with each other.

[0103] As used herein, the term “and / or” includes any and all combinations of one or more related enumerated entries.

[0104] The terminology used herein is for the purpose of describing particular embodiments only and is not intended to limit the invention. As used herein, the singular forms “a” and “the” are intended to include the plural forms as well, unless the context clearly indicates otherwise. It will also be understood that when the terms “comprising” and / or “made of” are used in this specification, the presence of the stated feature, integral, step, operation, element, and / or component is specified, but the presence or addition of one or more other features, integrals, steps, operations, elements, components, and / or groups thereof is not excluded. Terms such as “connected” or “linked” are not limited to physical or mechanical connections but can include electrical connections, whether direct or indirect.

[0105] Unless otherwise specified, all terms used herein (including technical and scientific terms) have the same meaning as commonly understood by one of ordinary skill in the art. It will also be understood that terms such as those defined in commonly used dictionaries should be interpreted as having the meaning consistent with their meaning in the context of the relevant art and the invention, and will not be interpreted as having an idealized or overly formal meaning unless expressly so defined herein.

[0106] In the technical solution of this invention, the collection, storage, use, processing, transmission, provision, and disclosure of user personal information all comply with relevant laws and regulations and do not violate public order and good morals. The use of user data in this technical solution follows relevant national laws and regulations (e.g., the "Information Security Technology - Personal Information Security Specification"). For example: appropriate measures are taken for personal information access control; restrictions are imposed on the display of personal information; the purpose of using personal information does not exceed the scope of direct or reasonable association; and explicit identity targeting is eliminated when using personal information to avoid precisely locating a specific individual.

[0107] The purpose of this invention is to provide an automatic task allocation method for laser processing production lines based on multi-device state coupling analysis, which solves the problems of task response delay, fragmented equipment scheduling, and poor process connection in existing laser production lines.

[0108] To achieve the above objectives, embodiments of the present invention provide an intelligent task allocation method, constructing a multi-station collaborative control system comprising a Manufacturing Execution System (MES), a central control scheduling system, an automated conveyor line, a marking and coding machine, a laser cutting machine, sorting equipment, an AGV transportation system, and a grinding equipment. In this system:

[0109] The central control and scheduling system receives task information packets from the MES and parses the task number, process path, and processing parameters.

[0110] By interacting with each workstation on the production line in real time, the system collects operational status information for workstations such as coding, cutting, sorting, and polishing.

[0111] Construct the device state vector S = [S1, S2, ..., S...] n ], and introduce a set of state weight factors W = [w1, w2,..., w n Construct a weighted device state vector S' = W · S;

[0112] The system constructs a task scheduling priority model based on the weighted device state vector S', task dependency graph T, and historical task execution records H, and dynamically determines the task triggering conditions by coupling the triggering judgment function F(S', T, H).

[0113] When multiple tasks meet the conditions simultaneously, scheduling and sorting are performed based on indicators such as material of the board, urgency of the task, and sequence number to achieve optimal allocation and distribution of multi-process tasks such as coding, cutting, and sorting.

[0114] After the task is issued, the corresponding equipment is driven to perform the processing operation, and the status is sent back after the task is completed to update the historical task execution record H;

[0115] The AGV system achieves intelligent path scheduling that matches the processing rhythm based on the workstation status and material flow requirements.

[0116] Figure 1 This is a system framework diagram of an automatic task allocation method for laser cutting production lines based on multi-device state coupling analysis provided by an embodiment of the present invention. It illustrates the functional relationships and signal flow paths between the MES system, central control scheduling system, automatic conveyor line, marking and coding machine, CNC laser cutting machine, automatic sorting machine, AGV transportation system and grinding equipment.

[0117] This invention forms a dynamic closed-loop control process of "equipment status perception -> status weighted calculation -> coupling judgment -> task issuance -> status feedback -> historical record update", which significantly improves the response speed and automation level of the production line, reduces the probability of manual intervention and task conflicts, and is suitable for intelligent laser processing scenarios of metal sheets in multiple processes.

[0118] The specific embodiments of the present invention will be described below with reference to the accompanying drawings, but the scope of protection of the present invention is not limited to the following embodiments.

[0119] 1. MES System Function Description

[0120] The MES system generates task information packages based on the production line plan and sends production tasks to the central control and scheduling system. The task information package includes: task number (uniquely identifying the task), task sequence number (controlling the production scheduling order), inkjet printing file path, cutting file path, sorting file path, nesting drawing number, start mode, working mode, calibration mode, sheet material, cutting mode, and cutting quality parameters. The MES system also monitors the execution status signals of each workstation on the production line. These signals form the basis for constructing the equipment state vector S, including:

[0121] Loading position status (idle, loading in progress, loading completed, fault occurred);

[0122] inkjet printing position status (idle, printing in progress, printing completed, malfunction occurred);

[0123] Cutting position status (idle, cutting in progress, cutting completed, malfunction occurred);

[0124] Sorting station status (idle, sorting in progress, sorting completed, malfunction occurred);

[0125] AGV status (idle, in transit, transit completed, malfunctioning);

[0126] Grinding position status (idle, grinding in progress, grinding completed, malfunction occurred).

[0127] 2. Central control and dispatch software logic

[0128] The central control and dispatch system receives and parses the task information packets sent by the MES. This is done using the device state vector S=[S1,S2,…,S…]. n The system uses a set of state weight factors W to form a weighted device state vector S' = W · S. The system dynamically adjusts the state weight factor set W based on historical task efficiency, device load, and failure rate data contained in historical task execution records H. The weights support periodic updates; every 20 task cycles, the W value is automatically recalculated based on the latest statistical results, thereby continuously optimizing the task judgment logic. The system constructs a coupled triggering decision function F(S',T,H) based on the weighted device state vector S', the task dependency graph T, and the historical task records H to make task triggering decisions.

[0129] 3. Automated conveyor line control logic

[0130] The conveyor line is responsible for transporting the sheet metal from the loading station, through the coding, cutting, and sorting stations, and back to the loading area, forming a closed loop. The central control unit interacts with the conveyor line through five types of points, collecting status information to update the equipment status vector S.

[0131] Loading point: Controls the loading status (allowed / loading in progress / loading completed);

[0132] Inkjet printing point: Controls whether inkjet printing tasks are allowed to be issued and the inkjet printing status is fed back;

[0133] Cutting point: Works with laser equipment to transmit allow / execute / complete signals;

[0134] Sorting point: The linked vision sorting system triggers sorting tasks and reads the status;

[0135] Heartbeat point: Ensure the conveyor line and central control system are continuously connected online.

[0136] Figure 2 This is a schematic diagram of the interaction logic between the central control scheduling system and the automatic conveyor line provided in the embodiment of the present invention. It shows how the central control system dynamically controls the material transmission process based on the status of the feeding, coding, cutting and sorting points fed back by the conveyor line.

[0137] 4. Marking inkjet printer

[0138] After receiving the inkjet file path, the central control system sends a start signal to the inkjet printer. Upon receiving the task reception signal, the device begins inkjet printing and sends real-time status updates such as "inkjet printing in progress" and "inkjet printing complete" to the central control system for updating the status vector. The device integrates a visual positioning compensation system to automatically identify the edges of the board material and the marking positions, fine-tuning the position of the inkjet graphic to improve inkjet printing accuracy.

[0139] Figure 3 This is a communication flowchart between the central control scheduling system and the marking inkjet printer provided in this embodiment of the invention. It shows the path of the task file sent by the central control system to the marking inkjet printer and the status signals such as task reception, marking in progress, and marking completion.

[0140] 5. CNC laser cutting machine

[0141] The cutting task issuance logic is coupled with the inkjet printing status. The cutting equipment begins operation after confirming that the inkjet printing is complete and the cutting status is permitted. Before cutting, the process file is automatically loaded and the material and thickness information of the sheet metal in the task information package is parsed. Parameters such as power, focal length, and frequency are dynamically adjusted to adapt to different cutting requirements. Feedback statuses such as "cutting in progress" and "cutting complete" are used to drive downstream processes.

[0142] Figure 4 This is a schematic diagram of the interaction between the central control scheduling system and the CNC laser cutting machine provided in this embodiment of the invention, showing how the central control system completes the signal triggering of the cutting task based on the inkjet printing and receives the cutting status changes.

[0143] 6. Automated sorting machine

[0144] After receiving the "cutting complete" status, the central control unit issues a sorting task to the sorting machine. The sorting equipment parses the nesting diagram and sorting path in the task information package, uses an industrial vision recognition system to identify the outline of the cut parts, and automatically maps the image coordinates to the robotic arm control coordinates to achieve accurate grasping and sorting. The entire process includes four stages: file parsing, status feedback, execution, and feedback completion signal.

[0145] Figure 5 This is a flowchart of the interaction between the central control scheduling system and the automatic sorting machine provided in this embodiment of the invention. It shows that after the task is issued, the sorting equipment parses the nesting file, performs the classification action, and provides feedback on the status.

[0146] 7. AGV Material Scheduling

[0147] After sorting is completed, the central control unit receives a request to replace a full-load pallet and queries the status of the grinding station in the weighted equipment state vector S':

[0148] If there is a vacant grinding station, an AGV will be dispatched directly to deliver the materials to that station.

[0149] If there are no available workstations, the workstation will be temporarily stored in the waiting area and will be reassigned when a workstation becomes available.

[0150] AGVs use path planning algorithms to evaluate transport routes, obstacle avoidance strategies, and work sequences.

[0151] Figure 6 This is a flowchart of the AGV material scheduling process provided in an embodiment of the present invention.

[0152] 8. Grinding equipment

[0153] Upon receiving the AGV pallet, the system automatically identifies and loads the material, then performs edge deburring. After polishing, sensors detect the part's status and send polishing completion information and processing quality level back to the central control system. This information is used to update the historical task execution record H and trigger the AGV to transport the part to the storage area.

[0154] 9. Logic for automatically assigning tasks

[0155] Task issuance is controlled by the coupled trigger decision function F(S′,T,H), and its core logic is as follows:

[0156] Inkjet printing task triggering logic: If it is the first task, the inkjet printing / cutting / sorting status is empty and inkjet printing is allowed; if it is not the first task, it is necessary to ensure that the current task status is empty, the previous task inkjet printing has been completed, and inkjet printing is allowed.

[0157] Figure 7 This is a flowchart illustrating the automatic dispatch logic of inkjet printing tasks provided in this embodiment of the invention, showing how the combination of the current task and the previous task status determines the triggering conditions of the inkjet printing task.

[0158] Cutting task triggering logic: If it is the first task, the inkjet printing must be completed, the current cutting and sorting status must be empty, and cutting is allowed. If it is not the first task, the inkjet printing must be completed, the current cutting and sorting status must be empty, and cutting is allowed when the previous task's inkjet printing / cutting is completed.

[0159] Figure 8 This is a flowchart illustrating the automatic distribution logic of cutting tasks provided in this embodiment of the invention, showing the coupling relationship between the inkjet printing and cutting states in the state vector and the corresponding triggering conditions.

[0160] Sorting task triggering logic: If it is the first task, both inkjet printing and cutting must be completed, the current task sorting status must be empty, and sorting must be allowed to trigger; if it is not the first task, both inkjet printing and cutting must be completed, the current task sorting status must be empty, all previous tasks at all workstations must be completed, and sorting must be allowed to trigger.

[0161] Figure 9This is an automatic task distribution flowchart provided by an embodiment of the present invention, which describes how to determine and trigger sorting execution based on the equipment status after the previous processing is completed.

[0162] This implementation method uses a closed-loop logic of "state perception -> weighted calculation -> coupling judgment -> condition triggering" to form a task-driven closed loop for the inkjet printing, cutting, sorting, and polishing processes. This significantly reduces intermediate waiting time, improves automation and task accuracy, and has high engineering feasibility and promotional value.

[0163] The laser production line task scheduling method based on multi-device state coupling analysis provided by this invention can be adapted to complex production line structures with multiple types of processing units. It is suitable for application in sheet metal processing, new energy equipment manufacturing, intelligent warehousing and sorting, and other fields, and has good cross-platform portability and commercial realization value.

[0164] The system and method described in this invention have been deployed and applied in actual production lines, covering the entire process control links such as inkjet printing, cutting, sorting, AGV scheduling, and polishing. Practical testing has verified that the method of this invention significantly improves task response efficiency and equipment utilization, demonstrating good engineering feasibility and industrialization potential.

[0165] Example embodiments have been disclosed herein, and while specific terminology has been used, it is for illustrative purposes only and should be construed as such, and is not intended to be limiting. In some instances, it will be apparent to those skilled in the art that features, characteristics, and / or elements described in conjunction with particular embodiments may be used alone, or in combination with features, characteristics, and / or elements described in conjunction with other embodiments, unless otherwise expressly indicated. Therefore, those skilled in the art will understand that various changes in form and detail may be made without departing from the scope of the invention as set forth in the appended claims.

Claims

1. A smart task allocation method based on multi-device state coupling analysis, applied to a laser cutting production line including a coding station, a cutting station, a sorting station, a grinding station, and an AGV transportation system, characterized in that... The method is executed by the central control and dispatch system and includes the following steps: Task information reception and parsing: The central control scheduling system receives task information packets from the manufacturing execution system and parses out the task number, process document path and material parameters; Weighted Equipment State Vector Construction: Based on the real-time status information collected from each workstation on the production line, the central control and scheduling system constructs an equipment state vector S = [S1, S2, ..., S...]. n ], and introduces a dynamically adjustable set of state weight factors W = [w1, w2,..., w n Generate a weighted device state vector S' = W · S; Coupling Trigger Determination: The central control and scheduling system inputs the weighted equipment state vector S', the predefined task dependency graph T determined by the current task information package, and the historical task execution record H which records equipment performance and past task execution status into the coupling trigger determination function F(S', T, H), and outputs a Boolean determination result on whether a specific process task in the current task information package is triggered; Task instruction issuance: If the output of the coupled trigger determination function for a certain process task is trigger, and the operation permission flag of the target workstation is true, then the central control scheduling system issues a task instruction containing the corresponding process file path and parameters to the corresponding target workstation. Status feedback and historical record update: After the task is completed, the executing device will send the task completion status feedback to the central control and scheduling system; the central control and scheduling system will update the task execution information to the historical task execution record H.

2. The method according to claim 1, characterized in that, In the coupling trigger determination step, the coupling trigger determination function F(S', T, H) outputs a Boolean determination result through the following hierarchical determination model: Rule-based decision layer based on mapping table: The coupled trigger decision function maintains a state combination mapping table, which combines the system state represented by the weighted device state vector S' with the preorder and successor dependencies between tasks defined by the task dependency graph T, and maps them to the following task triggering conditions: a) Triggering conditions for inkjet printing tasks: If the task in the task information package is the first task, the triggering condition is: the inkjet printing station status is represented as an acceptable task in S' after weight calculation, and its operation permission flag is true. If it is not the first task, the triggering condition needs to be added: the coding status of the previous task is represented as completed in S'; b) Cutting task triggering conditions: If the task in the task information package is the first task, the triggering conditions are: the status of the preceding inkjet printing task is represented as completed, and the status of the cutting station is represented as an acceptable task in S' after weight calculation, and its running permission flag is true. If it is not the first task, the triggering condition needs to be added: the cutting state corresponding to the previous task is represented as completed in S'; c) Sorting task triggering conditions: If the task in the task information package is the first task, the triggering conditions are: the status of the preceding inkjet printing and cutting tasks are both represented as completed, and the sorting station status is represented as an acceptable task in S' after weight calculation, and its operation permission flag is true; if it is not the first task, the triggering conditions also need to be added: the status of all station tasks corresponding to the preceding task are represented as completed in S'. The predictive conflict determination layer based on historical data: For tasks that meet the rules of the mapping table, the coupled trigger determination function also predicts whether there is a risk of resource conflict in the future task execution cycle based on the equipment load curve and AGV scheduling log in the historical task execution record H; if the prediction model indicates that the target workstation is about to fail, the preceding task may time out, or there is a conflict in the AGV path, then a negative determination is output; otherwise, a positive determination is output.

3. The method according to claim 1, characterized in that, In the weighted device state vector construction step, the central control scheduling system performs the following operations through its integrated predictive conflict detection module: Using the historical task execution record H as the data source for analysis, time-series analysis and pattern recognition are performed on the historical task execution time distribution, equipment load curve and AGV scheduling log contained therein to construct a short-term prediction model; If the prediction model identifies any of the following potential abnormal trends that reach a preset risk threshold, then an early warning signal is output to the coupling trigger determination step: a) Based on abnormal fluctuation patterns in equipment vibration, temperature, or power data, predict the probability of a target workstation failing within future task cycles. b) Based on the execution history and current progress of the preceding task type, determine the probability that the currently executing preceding task is expected to time out; c) Based on the path planning and real-time location information in the AGV scheduling log, predict the risk of traffic conflicts or scheduling bottlenecks in the AGV transportation path in the future period. d) Based on the system communication logs, determine whether the network communication delay or data packet loss rate may deteriorate to a level that affects the reliability of the real-time decision-making of the central control and scheduling system in the next cycle.

4. The method according to claim 1, characterized in that, Between the task instruction issuance step and the status feedback step, the central control scheduling system performs the following AGV transportation scheduling actions: a) If a sorting station triggers a full-load pallet transport request, and there is a grinding station in the weighted equipment state vector S' whose state is represented as idle and whose operation permission flag is true, then an AGV scheduling instruction is generated to transport the full-load pallet directly from the sorting station to the corresponding grinding station. b) If all polishing stations are busy or malfunctioning, generate an AGV scheduling instruction to transport the fully loaded pallet to a temporary buffer area for storage. c) When the status of any polishing station changes from busy to idle, if there is a waiting pallet in the temporary buffer area, a secondary scheduling instruction is triggered to dispatch the AGV to transport the pallet from the buffer area to the corresponding polishing station.

5. The method according to claim 1, characterized in that, In the task instruction issuance step, when the inkjet printing task triggering condition is met, the inkjet printing station performs the following visual positioning operation: The industrial camera captures images of the sheet metal at the inkjet printing station, and identifies the edge contours of the sheet metal and the preset positioning marks. Calculate the positional deviation between the actual feature location and the theoretical coordinates in the task information package, and correct the output coordinates of the inkjet file in real time based on the deviation value.

6. The method according to claim 1, characterized in that, In the task instruction issuance step, when the cutting task triggering condition is met, the CNC laser cutting machine performs the following operations: Parse the board characteristics contained in the task information packet from the received task instructions, including the board material and board thickness parameters. Based on the analyzed characteristics of the sheet material, the key process parameters on which the cutting operation depends are dynamically adjusted through a built-in process parameter library or an adaptive algorithm. These parameters include at least cutting power, cutting frequency, and cutting focal length.

7. The method according to claim 1, characterized in that, In the task instruction issuance step, when the sorting task triggering condition is met, the automatic sorting machine performs the following operations: Start the industrial vision recognition system and use industrial cameras to capture images of the cut boards at the sorting station; Image recognition technology is used to extract the contour information of each cut part in the image; The two-dimensional pixel coordinates of the identified part outline are converted into operation coordinates to control the sorting robot to grasp and move in three-dimensional physical space.

8. The method according to claim 1, characterized in that, In the status feedback and historical record update step, the grinding equipment performs the following operations: The status data of the parts after processing is collected by high-frequency vibration sensors or industrial vision inspection systems. Analyze the characteristic spectrum of vibration signals or the edge contour features in visual images to automatically determine whether the polishing effect meets the standards. The system generates an evaluation result that includes a grinding completion signal and the processing quality level, and uploads it to the central control and dispatch system.

9. The method according to claim 1, characterized in that, In the coupling trigger determination step, when multiple tasks simultaneously meet the triggering conditions, the central control and scheduling system performs the following priority adjustment operation: Tasks that meet the triggering conditions are dynamically sorted according to predefined priority rules; The sorting rules take into account at least one of the following factors: the task number defined in the task information package, the processing urgency implied by the material type of the sheet metal, and the process urgency label set by the manufacturing execution system or the operator. Based on the sorting results, the highest priority task instructions are issued to the workstations first.

10. The method according to claim 1, characterized in that, The dynamic adjustment of the state weight factor set W is achieved through the self-optimization module performing the following operations: Set an update cycle based on the number of tasks, and automatically start the optimization process when a preset number N tasks are completed; Access the historical task execution record H and collect the operation data of each workstation in the most recent period; Calculate the multi-dimensional performance metrics for each workstation i, including: Historical task efficiency index E i Calculation based on the deviation between the actual execution time of workstation tasks and standard working hours; Equipment load rate index L i Calculated based on the ratio of actual runtime to total runtime of the workstation within the statistical period; Equipment failure rate index F i Calculated based on the ratio of the frequency of failures to the runtime of the workstation within the statistical period; The basic weight of each workstation is dynamically adjusted based on performance indicators: If E i If the display efficiency is higher than the system average, then the efficiency will be increased according to preset rules. i (new); If L i If the display shows a consistently high load, then reduce the load according to the preset rules. i (new); If F i If the reliability of the display decreases, then w will be reduced according to the preset rules. i (new); The calculated new weight factors are normalized to generate a new generation of state weight factor set W. new ; Use W new Atomically replace the current set of weight factors W current This is then applied to the construction of the weighted device state vector S' for subsequent tasks.

Images