Intelligent manufacturing digital twinning whole-process management and control method and platform based on swan gap system

Abstract

The invention provides an intelligent manufacturing digital twinning whole process management and control method and platform based on a swan gap system, and the method comprises the steps: constructing a process structure chart according to a manufacturing process, and collecting operation parameters through a state collection port; and constructing a current state vector of each node according to the operation parameters, and obtaining an operability mark of each node according to the precursor node set, the current state vector and a state linkage rule. The node of the operable mark is defined as a target node. And generating a plurality of candidate control strategies of each target node according to the strategy parameter template, the current state vector and the current equipment margin. And obtaining a cost evaluation value of each candidate control strategy according to the cost function. The candidate control strategy corresponding to the minimum cost evaluation value is defined as a target control strategy. And sending the target control strategy to a controller for execution. The whole-process modeling capability with the process dependency relationship as the framework is achieved, and when the production line rhythm or the equipment load changes, the scheduling logic can also be automatically adapted accordingly.

Description

TECHNICAL FIELD

[0001] The application belongs to the field of intelligent manufacturing, and particularly relates to an intelligent manufacturing digital twin whole-process management and control method and platform based on a Hongmeng system. BACKGROUND

[0002] With the evolution of intelligent manufacturing towards multi-terminal collaboration, flexible production and high-speed rhythm, digital twin is regarded as a core enabling technology throughout the whole life cycle of design, production and operation. In related technologies, most manufacturing systems are limited to device-level or process section-level monitoring. Once the production line rhythm or device load changes, the scheduling logic cannot automatically adapt, and can only rely on manual experience to redevelop rules. Therefore, how to enable the manufacturing system to remain efficient, stable and self-adaptive in the rhythm fluctuation, device switching or production capacity adjustment scenarios has become a technical problem to be solved. SUMMARY

[0003] The purpose of the present application is to design an intelligent manufacturing digital twin whole-process management and control method and platform based on a Hongmeng system, which can enable the manufacturing system to remain efficient, stable and self-adaptive in the rhythm fluctuation, device switching or production capacity adjustment scenarios.

[0004] To achieve the above purpose, in a first aspect of the present application, an intelligent manufacturing digital twin whole-process management and control method based on a Hongmeng system is provided, which comprises:

[0005] constructing a process structure diagram according to a manufacturing process; wherein the process structure diagram comprises nodes and sequential relationships between the nodes; wherein the nodes comprise a state acquisition port and a set of predecessor nodes, and running parameters are acquired through the state acquisition port;

[0006] constructing a current state vector of each node according to the running parameters, and obtaining an operability mark of each node according to the set of predecessor nodes, the current state vector and a preset state linkage rule; wherein the operability mark comprises an operable mark, and a node with the operable mark is defined as a target node;

[0007] obtaining a current device margin, and generating a plurality of candidate control strategies for each target node according to a preset strategy parameter template, the current state vector and the current device margin;

[0008] evaluating the candidate control strategies according to a preset cost function to obtain a cost evaluation value of each candidate control strategy; wherein the candidate control strategy corresponding to the minimum cost evaluation value is defined as a target control strategy;

[0009] sending the target control strategy to a controller for execution.

[0010] As preferred, the sending the target control strategy to the controller for execution comprises:

[0011] judging whether the target control strategy is confirmed to be executed according to a preset judging rule;

[0012] if not, generating a candidate control strategy for each target node again to obtain an updated candidate control strategy, obtaining a cost evaluation value of the updated candidate control strategy to obtain an updated cost evaluation value, selecting an updated target control strategy based on the updated cost evaluation value, and sending the updated target control strategy to the controller for execution;

[0013] if yes, sending the target control strategy to the controller for execution.

[0014] As preferred, the judging whether the target control strategy is confirmed to be executed according to a preset judging rule comprises:

[0015] obtaining a reference state vector of the target node when the target control strategy is generated;

[0016] if the difference between the current state vector and the reference state vector is less than a preset state fluctuation threshold, judging that the target control strategy is confirmed to be executed, otherwise judging that the target control strategy is not confirmed to be executed.

[0017] As preferred, a manufacturing node state map is constructed according to the current state vector, and after the sending the target control strategy to the controller for execution, the method further comprises:

[0018] obtaining an updated running parameter according to the state acquisition port;

[0019] judging whether the target control strategy is successfully executed according to a preset judging rule to obtain a success or failure flag;

[0020] updating the updated running parameter and the success or failure flag to the manufacturing node state map.

[0021] As preferred, the judging whether the target control strategy is successfully executed according to a preset judging rule to obtain a success or failure flag comprises:

[0022] obtaining an actual state vector after the target control strategy is executed by the controller;

[0023] when the actual state vector is within a preset expected state space range, judging that the target control strategy is successfully executed, and the success or failure flag is a success flag;

[0024] When the actual state vector is out of the expected state space range, it is judged that the target control strategy is not successfully executed, and the success or failure flag is a failure flag.

[0025] Preferably, the precursor node set comprises a plurality of precursor nodes, the current state vector of each node is constructed according to the operation parameter, and the operability mark of each node is obtained according to the precursor node set, the current state vector and a preset state linkage rule, comprising:

[0026] It is judged whether the precursor node is completed operation;

[0027] It is judged whether the node satisfies an independent operation preparation condition according to a preset state evaluation function and the current state vector;

[0028] The operability mark of each node is obtained according to the state linkage rule, whether the precursor node is completed operation, whether the node satisfies the independent operation preparation condition, a preset process path weight coefficient, a preset state influence weight and a preset start threshold bias.

[0029] Preferably, the candidate control strategy is evaluated according to a preset cost function, and a cost evaluation value of each candidate control strategy is obtained, comprising:

[0030] The unit task energy consumption in the execution process of the candidate control strategy is obtained;

[0031] The structural difference degree between the candidate control strategy and a preset reference standard template is obtained;

[0032] The multi-node conflict suppression item corresponding to the candidate control strategy is obtained;

[0033] The candidate control strategy is evaluated according to the unit task energy consumption, the structural difference degree and the multi-node conflict suppression item, and a cost evaluation value of each candidate control strategy is obtained.

[0034] In a second aspect of the present application, a smart manufacturing digital twin full-process management and control platform based on a Hongmeng system is provided, and the platform comprises:

[0035] A construction unit is configured to construct a process structure diagram according to a manufacturing process, wherein the process structure diagram comprises nodes and a sequential relationship between the nodes, and the nodes comprise a state acquisition port and a precursor node set, and operation parameters are acquired through the state acquisition port.

[0036] A marking unit is configured to construct a current state vector of each node according to the operation parameters, and obtain an operational mark of each node according to the set of predecessor nodes, the current state vector and a preset state linkage rule; wherein the operational mark comprises an operable mark, and a node with the operational mark as the operable mark is defined as a target node.

[0037] A policy unit is configured to obtain a current device margin, and generate a plurality of candidate control policies of each target node according to a preset policy parameter template, the current state vector and the current device margin.

[0038] An evaluation unit is configured to evaluate the candidate control policies according to a preset cost function, and obtain a cost evaluation value of each candidate control policy; wherein a candidate control policy corresponding to a minimum cost evaluation value is defined as a target control policy.

[0039] A sending unit is configured to send the target control policy to a controller for execution.

[0040] To achieve the above object, a third aspect of embodiments of the present application provides an electronic device, which comprises a memory and a processor, the memory stores a computer program, and the processor implements the method of the first aspect when executing the computer program.

[0041] To achieve the above object, a fourth aspect of embodiments of the present application provides a computer readable storage medium, which stores a computer program, and the computer program is executed by a processor to implement the method of the first aspect.

[0042] The present application has at least the following beneficial technical effects:

[0043] To solve the above problems, the application provides an intelligent manufacturing digital twin whole-process management and control method and platform based on a Hongmeng system, the core of which is to realize complete technical closure of manufacturing links from structure table to dynamic regulation and control through a four-level link of "atomic service modeling-state atlas synchronization-strategy generation verification-execution closed loop update". First, with the aid of the Hongmeng atomic service mechanism, each process is split into nodes that can be independently registered and have control and collection interfaces, and the process dependence is accurately described in the form of a directed graph. Second, a lightweight twin is deployed at each node end side to collect key states in real time and gather them into an indexable time series atlas, and a "maneuverability" label is derived using the process topology to form a state expression for scheduling. Third, the platform generates candidate control strategies in real time at maneuverable nodes through template parameter combination and atlas constraint linkage, and uses regular terms such as energy consumption, process deviation degree and multi-node conflict suppression to construct a verification function to screen the optimal instructions that meet physical and beat constraints. Finally, the control strategy is issued to the equipment through the Hongmeng distributed channel, and the execution result is written back to the state atlas in an expected-actual difference matching manner, closing the twin evolution loop. This scheme not only eliminates the communication and computing bottlenecks of centralized architecture, but also guarantees control safety and process consistency through a structure-aware strategy verification mechanism; the four-level link completes state synchronization, strategy generation and instruction issuance in milliseconds, enabling the manufacturing system to maintain efficient, stable and adaptive operation capability in the scenarios of beat fluctuation, equipment switching or capacity adjustment. BRIEF DESCRIPTION OF DRAWINGS

[0044] The application is further described with reference to the accompanying drawings, but the embodiments in the drawings do not constitute any limitation on the application, and other drawings can be obtained by those of ordinary skill in the art without creative labor on the basis of the following drawings.

[0045] Figure 1 is a flowchart of the intelligent manufacturing digital twin whole-process management and control method based on the Hongmeng system provided by the application.

[0046] Figure 2 is a flowchart of the intelligent manufacturing digital twin whole-process management and control method based on the Hongmeng system provided by another embodiment of the application.

[0047] Figure 3 is a flowchart of the intelligent manufacturing digital twin whole-process management and control method based on the Hongmeng system provided by a third embodiment of the application.

[0048] Figure 4 is a structural schematic diagram of the intelligent manufacturing digital twin whole-process management and control platform based on the Hongmeng system provided by the application. DETAILED DESCRIPTION

[0049] Embodiments of the present application are described below in detail with reference to the accompanying drawings, wherein the same or similar components or components having the same or similar functions are denoted by the same or similar reference numerals throughout the drawings. The embodiments described below with reference to the accompanying drawings are exemplary and are only used to explain the present application and cannot be understood as a limitation of the present application.

[0050] Please refer to Figure 1 , Figure 1 is a flowchart of a method for intelligent manufacturing digital twin whole-process management based on a Hongmeng system provided by an embodiment of the present application, Figure 1 The method in the embodiment can include but is not limited to steps S101 to S105.

[0051] Step S101, constructing a process structure diagram according to a manufacturing process; wherein the process structure diagram includes nodes and sequential relationships between the nodes; wherein the nodes include a state acquisition port and a set of predecessor nodes, and running parameters are acquired through the state acquisition port;

[0052] Step S102, constructing a current state vector of each node according to the running parameters, and obtaining an operability mark of each node according to the set of predecessor nodes, the current state vector and a preset state linkage rule; wherein the operability mark includes an operable mark, and a node with the operability mark as the operable mark is defined as a target node;

[0053] Step S103, acquiring a current device margin, and generating a plurality of candidate control strategies for each target node according to a preset strategy parameter template, the current state vector and the current device margin;

[0054] Step S104, evaluating the candidate control strategies according to a preset cost function, and obtaining a cost evaluation value of each candidate control strategy; wherein a candidate control strategy corresponding to the smallest cost evaluation value is defined as a target control strategy;

[0055] Step S105, sending the target control strategy to a controller for execution.

[0056] In step S101 of some embodiments, the goal is to convert each process unit in the manufacturing process from a traditional physical description to a system that can be programmed, deployed and scheduled. In order to achieve this goal, the entire manufacturing process needs to be deconstructed into a number of atomic process nodes with clear control logic. Each node must bind an actual control channel of a physical device and clearly define its upstream and downstream process dependency relationship. These nodes will eventually be deployed as independent atomic services under the distributed operation architecture of the Hongmeng system, with real-time control interface and state acquisition capability, thereby laying a foundation for subsequent twin expression, state mapping and control generation.

[0057] Specifically, the process flow structure diagram derived from the manufacturing site or MES (Manufacturing Execution System) and the equipment configuration list. Among them, the process flow structure diagram represents the logical order of the entire process chain, such as "raw material handling → positioning → heating → compression → cooling → unloading". And the equipment configuration list includes the control hardware (such as PLC controller, motor driver) and acquisition devices (such as thermocouple, displacement encoder) used in each station.

[0058] The manufacturing process modeling work starts with process decomposition. Take an automatic press assembly line as an example, its main processes include:

[0059] Process 1: The raw material is handled by the feeding robot;

[0060] Process 2: The raw material is positioned by the conveying device;

[0061] Process 3: The heating furnace is heated to the set temperature;

[0062] Process 4: The hydraulic mechanism completes the press assembly operation;

[0063] Process 5: The cooling air duct is ventilated;

[0064] Process 6: The finished product is transported by the unloading robot.

[0065] Each process is abstracted as a node , and the structure of the node is uniformly defined as shown in the following formula (1):

[0066] (1);

[0067] Among them, represents the control interface set of the node, which specifically includes device start-stop, parameter issuance, etc. Control signals are connected to lower devices (such as PLC) through the distributed control module of the Hongmeng system. represents the state acquisition port, such as temperature, position, running flag, etc. These quantities are connected to sensors through the gateway and collected by the drive interface provided by the Hongmeng system. and respectively represent the predecessor node set and the successor node set of the node, which are the directed dependency set obtained from the BOM structure or engineering configuration file derived from the MES. Each node has been deployed in the atomic service of the Hongmeng system and connected to the physical device through the distributed interface. In this step, the state acquisition port of each node will collect its running parameters in real time.

[0068] The set composed of all nodes is represented as , and the process structure diagram is constructed as shown in the following formula (2):

[0069] (2);

[0070] in, This indicates the sequential relationship between all process steps, such as This indicates that "Process 3 will only be executed after Process 2 is completed." The process structure diagram is used downstream for state diagram construction and strategy generation context judgment.

[0071] After each node is modeled, a service instance is generated in the system by calling the atomic service registration interface of the HarmonyOS system, and control capabilities, status items, and node dependency information are published to form a unified distributed service topology. This process can be completed automatically by deployment tools, or it can be manually built by engineers using a UI modeling method.

[0072] Taking the pressing station as an example, its nodes include:

[0073] The PLC receives control signals such as "pressurization start", "stop", and "pressure target setting".

[0074] Data is collected from pressure sensors and limit switches, including "current pressure value", "action completion flag", and "position status".

[0075] : Heating node;

[0076] Cooling node.

[0077] Process structure diagram Ultimately, it is stored in a structured configuration file (such as JSON) and called through system interfaces in subsequent modules for indexing and scheduling of states and policies. All elements in the expression are logically consistent, and the nodes and edges form a complete topology, which helps with the logical recursion of subsequent state evolution simulation and process control decisions.

[0078] In step S102 of some embodiments, each node The current state vector at time t It is expressed as shown in the following formula (3):

[0079] (3);

[0080] in, Indicates the first Each state variable originates from the actual sensor modules deployed on the equipment. For example, for a hydraulic pressing node, its... It can collect real-time pressure values ​​(accessed to the HarmonyOS edge node via the CAN interface). It can output position signals for displacement sensors (acquired via the SPI channel). Boolean flag (feedback by digital quantity in PLC) can be completed for the action.

[0081] The manufacturing node state atlas is constructed according to the current state vector of all nodes, as shown in the following formula (4):

[0082] (4);

[0083] Atlas is stored as a snapshot in a structured state database, supports structured query with node ID and timestamp as index, and is used for data calling of subsequent decision modules.

[0084] Further, in order to realize the dynamic controllability identification of the manufacturing link, an operational marking mechanism with "process dependence awareness" is designed. Unlike traditional "state threshold judgment", this mechanism introduces upstream and downstream state linkage rules, and combines the predecessor node set to define whether each node has the state of "can be dispatched" at the current time. The operational marking is defined as shown in the following formula (5):

[0085] (5);

[0086] Among them, represents the operational marking, when is 1, the operational marking is the operational marking, is 0, the operational marking is the inoperational marking. represents whether the predecessor node has completed operation, which can be judged by the controller feedback Boolean signal whether the predecessor node has completed operation. represents the process path weight coefficient (set according to the process importance, usually 1 or 0.5). is the current node state evaluation function, which represents whether the independent operation preparation condition of the node is met. is the state influence weight (set according to the sensitivity of the node to the process quality). is the starting threshold bias, is the step function, which outputs 1 only when the input is non-negative, otherwise 0. Among them, does not use a single numerical judgment, but a structure evaluation function, for example:

[0087] For the heating node, represents whether the current temperature is in the set interval (such as 180~200°C);

[0088] For the conveying node, represents that the last movement is completed and the position is zeroed;

[0089] For the assembly node, indicates whether the positioning fixture is locked in place.

[0090] The current state vector is updated simultaneously with the operational flag, and the period can be set to 1s or the manufacturing beat corresponding period. This composite logic rule runs in the Hongmeng service process through embedded scripts or rule engines, supporting personalized configuration by node.

[0091] It should be noted that formula (5) introduces a structure-dependent weighting term and the linkage mechanism of the state evaluation function , rather than simply relying on whether a state variable meets the condition. Especially in a manufacturing process with multiple branches and multiple paths executing concurrently, this method can effectively avoid the scheduling entry problem of misjudgment under the traditional "only look at this node" strategy. For example, in a certain parallel assembly scenario, if node depends on and , the traditional method may execute as soon as is completed, but this method can strictly determine whether all critical paths meet the requirements according to the path weight, ensuring the accuracy of the scheduling timing.

[0092] For example, in a heating-pressing process, assuming that the heating node has a state collection result of temperature 185°C, heating Boolean flag 1, and action completion flag 1 at t=22, then , = the current pressure is within the allowed range, if only depends on , and , then . This judgment can be used as a scheduling entry basis directly used by the strategy generation module.

[0093] Further, the node with the operational flag as the target node is defined as the target node, and the target is to generate a control strategy that meets the physical constraints, safety rules, and production beat based on the current state of the manufacturing system, and to perform real-time verification and scoring on each strategy, and finally to select the optimal control instruction with executable performance and issue it to the equipment for execution.

[0094] In step S103 of some embodiments, considering that manufacturing processes generally have the characteristics of reusable operation templates, clear process requirements, and low error tolerance, a general AI decision model is not used, but a parameter generation module guided by structural constraints is used. For each node , a set of strategy parameter templates is predefined, and each template represents a typical process operation strategy. For example, in the "pressing" node, It can represent "fast forward + pressurization + hold + rewind". It can represent "slow advance + pressure + slow retreat". Each strategy parameter template includes a combination of parameters for multiple action segments, such as speed, stroke, and hold time.

[0095] The system obtains the current device margin and generates multiple candidate control strategies for each target node based on the strategy parameter template, the current state vector, and the current device margin. In other words, strategy generation combines the current state vector... With strategy parameter template Candidate control strategies are dynamically adjusted and generated while satisfying equipment boundary constraints. As shown in the following formula (6):

[0096] (6);

[0097] in, Indicates the first The first target node One candidate control strategy Indicates the first A strategy parameter template. Indicates the first The current state vector of each target node, such as the current pressing position, equipment temperature, and execution time. Indicates the first The current equipment margins of each target node, such as remaining travel, current temperature rise rate, and safe shutdown distance, are provided in real time by the controller feedback interface. The function adjust(...) is a fine-tuning function for the structure template parameters. It is executed by performing linear interpolation or switching jump conditions on individual parameters in the strategy parameter template based on the current state vector.

[0098] In some embodiments, in steps S104 to S105, each time a candidate control strategy is generated... Real-time verification is performed immediately to ensure that the strategy will not cause physical conflicts, process deviations or safety hazards under the current system state and structural conditions. This verification is evaluated by designing a set of cost functions for the structural characteristics of the manufacturing system. The cost functions include not only energy consumption and the structural difference between the candidate control strategy and the preset reference standard template, but also introduce a multi-node conflict suppression term, defined as shown in the following formula (7):

[0099] (7);

[0100] in, Indicate candidate control strategies The cost assessment value, Indicate candidate control strategies The energy consumption of each unit task in the execution process comes from the device power model or historical execution data. The structural difference between the candidate control strategy and the reference standard template measures whether the control action deviates from the typical beat path, using segmented cosine similarity or key parameter difference aggregation. The candidate control strategy corresponds to the multi-node conflict suppression term, which determines whether the strategy deviates from its adjacent nodes The current state conflict (such as mechanical arm trajectory overlap, simultaneous pressing request, etc.) is based on the manufacturing node state atlas The possible execution window of the adjacent node is calculated in real time. 、 and are cost weight coefficients, which are adjusted and configured according to different production targets (energy saving, precision, safety).

[0101] In formula (7), by ensuring that the generated strategy does not deviate from the standard process behavior, and enhancing consistency when the control action has uncertainty. By integrating atlas structure information and adjacent node behavior, a dynamic conflict suppression mechanism is formed for structural scenarios, which is rarely seen in traditional strategy verification processes.

[0102] When the cost evaluation value of multiple candidate control strategies is less than the threshold value set by the node , the strategy with the smallest cost evaluation value is selected as the target control strategy , which is sent to the device controller, or multiple candidate control strategies are retained according to the redundancy requirement for system dynamic switching. Strategies with cost evaluation values not less than the threshold value are all excluded, if there are no feasible items after exclusion, a conservative candidate is generated using the conservative template of the node, and then evaluated. If there are still no feasible items, no instructions are issued in this period, and the node is marked as currently unschedulable, and the new manufacturing node state atlas is recalculated in the next period.

[0103] For example, in the heating node , the strategy parameter template corresponds to the strategy "medium power heating for 8 seconds + temperature maintenance", if the current temperature has reached the critical point, the system automatically adjusts the power curve according to the temperature rise margin in ; the generated candidate control strategy has a small difference compared to the reference standard template , and the current adjacent node If the (press-fit) is not schedulable yet, then = 0, the final strategy is selected and sent to the controller for execution.

[0104] The steps S101 to S105 shown in the embodiments of the present application are to construct a process structure diagram according to a manufacturing process, wherein the process structure diagram includes nodes and sequential relationships between the nodes. The nodes include a state acquisition port and a set of predecessor nodes, and the running parameters are acquired through the state acquisition port. The current state vector of each node is constructed according to the running parameters, and the operability mark of each node is obtained according to the set of predecessor nodes, the current state vector and the preset state linkage rule. The operability mark includes an operable mark, and the node with the operable mark is defined as a target node. The current equipment margin is obtained, and a plurality of candidate control strategies of each target node are generated according to the preset strategy parameter template, the current state vector and the current equipment margin. The candidate control strategies are evaluated according to the preset cost function, and the cost evaluation value of each candidate control strategy is obtained. The candidate control strategy corresponding to the minimum cost evaluation value is defined as a target control strategy. The target control strategy is sent to the controller for execution. The full-process modeling capability with the process dependency relationship as the skeleton is realized, and when the production line beat or the equipment load changes, the scheduling logic can also automatically adapt. The information round-trip delay and bandwidth consumption problem can be solved by flexible deployment on the distributed terminal side. The series-parallel process structure can be finely expressed, and the real-time strategy generation and verification function is also provided.

[0105] Please refer to Figure 2 In some embodiments, after step S105, the intelligent manufacturing digital twin full-process management and control method based on the Hongmeng system can further include but is not limited to steps S201 to S203:

[0106] Step S201: determining whether the target control strategy is confirmed for execution according to a preset judgment rule;

[0107] Step S202: if not, the candidate control strategies of each node are regenerated to obtain updated candidate control strategies, the cost evaluation values of the updated candidate control strategies are obtained to obtain updated cost evaluation values, the updated target control strategy is selected based on the updated cost evaluation values, and the updated target control strategy is sent to the controller for execution;

[0108] Step S203: if yes, the target control strategy is sent to the controller for execution.

[0109] In step S201 of some embodiments, the reference state vector of the target node when generating the target control strategy is obtained. The determination rule is used to determine whether to issue. The determination rule is that the operational marker is marked as an operational marker, and the difference between the current state vector and the reference state vector is less than the preset state fluctuation threshold. Then it is determined that the target control strategy is confirmed to be executed, otherwise it is determined that the target control strategy is confirmed not to be executed. As shown in the following formula (8):

[0110] (8);

[0111] wherein, indicates whether the target control strategy is confirmed to be executed, indicates the operational marker, and indicates whether the current state of the node allows scheduling. is the reference state vector of the target node when generating the target control strategy, is the state fluctuation threshold allowed by the node. indicates the indicator function, which takes 1 when the condition is true, and 0 otherwise, when is 1, it is determined that the target control strategy is confirmed to be executed. When is 0, it is determined that the target control strategy is confirmed not to be executed.

[0112] The actual issuance of the target control strategy is communicated with the edge device through the distributed interface of the Hongmeng system, and the target control strategy is sent using a structured command package . The structured control package includes:

[0113] Node ID;

[0114] Control action type (such as "pressing start", "temperature control maintenance");

[0115] Action parameter set (such as target pressure, temperature, and holding time);

[0116] Check code and execution window timestamp.

[0117] After receiving the instruction, the device controller (such as PLC or embedded server) enters the task execution state, and the system simultaneously starts the execution monitoring module to collect the key state changes in the control process.

[0118] In steps S202 to S203 of some embodiments, when it is determined that the target control strategy is confirmed not to be executed, the candidate control strategy of each target node is regenerated to obtain an updated candidate control strategy. The cost evaluation value of the updated candidate control strategy is obtained to obtain an updated cost evaluation value. The updated target control strategy is selected based on the updated cost evaluation value, and the updated target control strategy is sent to the controller for execution. If yes, the target control strategy is sent to the controller for execution.

[0119] Through the above steps S201 to S203, before the target control strategy is issued, it is necessary to confirm again that the target control strategy is still effective in the current state. Considering that the state in the manufacturing system changes fast, there may be hundreds of milliseconds of delay between the generation of the target control strategy and the actual issuance, if some node state mutates, such as temperature change exceeding the limit, previous node not completed, etc., the current control request needs to be cancelled.

[0120] Please refer to Figure 3 In some embodiments, after step S105, the intelligent manufacturing digital twin whole-process management and control method based on the hyper-intelligent system can further include, but is not limited to, steps S301 to S303:

[0121] Step S301, collecting running parameters according to the state collection port to obtain updated running parameters;

[0122] Step S302, judging whether the target control strategy is executed successfully according to the preset judgment rule to obtain a success or failure flag;

[0123] Step S303, updating the updated running parameters and the success or failure flag to the manufacturing node state map.

[0124] In step S301 of some embodiments, after the equipment completes the control instruction action according to the target control strategy, the state collection port collects running parameters to obtain updated running parameters, and the updated running parameters are uploaded to the edge control unit.

[0125] In steps S302 to S303 of some embodiments, whether the target control strategy is executed successfully is judged according to the preset judgment rule to obtain a success or failure flag. The actual state vector after the controller executes the target control strategy is obtained. The judgment rule is that when the actual state vector is within the preset expected state space range, it is judged that the target control strategy is executed successfully, and the success or failure flag is a success flag. When the actual state vector is outside the expected state space range, it is judged that the target control strategy is not executed successfully, and the success or failure flag is a failure flag. That is, the judgment rule is shown in the following formula (9):

[0126] (9);

[0127] Wherein, represents the success or failure flag, represents the actual state vector. represents the expected state space range of the target control strategy . represents an indicator function, when is 1, it is judged that the target control strategy is executed successfully, and the success or failure flag is a success flag. When When the value of the temperature deviation is 0, it is determined that the target control strategy is not successfully executed, and the success flag is the unsuccessful flag.

[0128] Further, the updated running parameter and the success flag are updated to the manufacturing node state atlas For example, under a target control strategy of "heating to 180°C and maintaining for 30 seconds", if the actual state collected after execution is and the maintaining duration is 29.6 seconds, if the temperature deviation is set to be ±3°C and the duration deviation is not more than 1 second, then it is considered that the target control strategy is successfully executed, and the success flag is the successful flag.

[0129] In some process flows with synchronous coordination requirements, such as synchronous handling of dual-arm robots, synchronous triggering of press-fitting and positioning, the system also designs a "synchronous control window mechanism": a plurality of control strategies are combined into a synchronous package for issuing, and a synchronous window parameter is introduced to ensure that all actions are completed within a set time difference. In actual application, the value of the synchronous window parameter may be set to 100 ms to ensure assembly accuracy.

[0130] Through the above steps S301 to S303, the updated running parameter and the success flag are written to the manufacturing node state atlas, the consistency between the twin and the field can be maintained in real time, the trusted input is provided for strategy generation in the next cycle, the abnormality is quickly located and the failure cause is determined, the conflict avoidance and the beat reorganization are supported in the multi-node coordination, the traceable execution record is formed to facilitate quality and compliance audit, and the engineering calibration efficiency of the process parameters and the threshold is improved through the continuously accumulated data.

[0131] Referring to Figure 4 , the embodiment of the application further provides an intelligent manufacturing digital twin whole-process management and control platform based on a Hongmeng system, which can implement the intelligent manufacturing digital twin whole-process management and control method based on the Hongmeng system. The system comprises:

[0132] A construction unit 401 is configured to construct a process structure diagram according to a manufacturing process; wherein the process structure diagram comprises nodes and sequential relationships between the nodes; wherein the nodes comprise a state collection port and a set of predecessor nodes, and the running parameter is collected through the state collection port;

[0133] A marking unit 402 is configured to construct a current state vector of each node according to the running parameter, and obtain an operability mark of each node according to the set of predecessor nodes, the current state vector and a preset state linkage rule; wherein the operability mark comprises an operable mark, and the node with the operability mark as the operable mark is defined as a target node;

[0134] The policy unit 403 is configured to acquire a current device margin, and generate a plurality of candidate control strategies for each target node according to a preset policy parameter template, a current state vector and the current device margin;

[0135] The evaluation unit 404 is configured to evaluate the candidate control strategies according to a preset cost function, to obtain a cost evaluation value of each candidate control strategy; wherein the candidate control strategy corresponding to the minimum cost evaluation value is defined as a target control strategy;

[0136] The sending unit 405 is configured to send the target control strategy to a controller for execution.

[0137] The specific implementation of the power transformation equipment digital twin data processing system based on big data is basically the same as the specific embodiments of the above-mentioned intelligent manufacturing digital twin whole-process management and control method based on the Hongmeng system, and will not be repeated here.

[0138] The above is the preferred embodiment of the present application. It should be noted that for those skilled in the art, without departing from the principles of the present application, a number of improvements and refinements can be made, which are also considered within the scope of protection of the present application.

Claims

1. A method for intelligent manufacturing digital twin whole-process management and control based on a hyper system, characterized in that, The method comprises: constructing a process structure diagram according to a manufacturing process; wherein the process structure diagram comprises nodes and sequential relationships between the nodes; wherein the nodes comprise a state acquisition port and a set of predecessor nodes, and operating parameters are acquired through the state acquisition port; constructing a current state vector of each node according to the operating parameters, and obtaining an operability mark of each node according to the set of predecessor nodes, the current state vector and a preset state linkage rule; wherein the operability mark comprises an operable mark, and a node with the operable mark is defined as a target node; obtaining a current equipment margin, and generating a plurality of candidate control strategies of each target node according to a preset strategy parameter template, the current state vector and the current equipment margin; evaluating the candidate control strategies according to a preset cost function, and obtaining a cost evaluation value of each candidate control strategy; wherein the candidate control strategy corresponding to the minimum cost evaluation value is defined as a target control strategy; sending the target control strategy to a controller for execution.

2. The method according to claim 1, wherein, The sending of the target control strategy to the controller for execution comprises: judging whether the target control strategy is confirmed for execution according to a preset judgment rule; if not, regenerating the candidate control strategies of each target node, obtaining an updated cost evaluation value of the updated candidate control strategies, selecting an updated target control strategy based on the updated cost evaluation value, and sending the updated target control strategy to the controller for execution; if yes, sending the target control strategy to the controller for execution. 3.The method of claim 2, wherein, The judging of whether the target control strategy is confirmed for execution according to the preset judgment rule comprises: obtaining a reference state vector of the target node when the target control strategy is generated; if the operability mark is the operable mark and the difference between the current state vector and the reference state vector is less than a preset state fluctuation threshold, it is judged that the target control strategy is confirmed for execution, otherwise it is judged that the target control strategy is not confirmed for execution.

4. The method of claim 1, wherein the method further comprises: constructing a manufacturing node state map according to the current state vector, and after the sending of the target control strategy to the controller for execution, the method further comprises: acquiring operating parameters according to the state acquisition port to obtain updated operating parameters; judging whether the target control strategy is successfully executed according to a preset judgment rule to obtain a success or failure flag; updating the updated operating parameters and the success or failure flag to the manufacturing node state map.

5. The smart manufacturing digital twin whole-process management and control method based on the Mesosystem according to claim 4, characterized in that, The judging of whether the target control strategy is successfully executed according to the preset judgment rule to obtain a success or failure flag comprises: obtaining an actual state vector after the controller executes the target control strategy; when the actual state vector is within a preset expected state space range, it is judged that the target control strategy is successfully executed, and the success or failure flag is a success flag; when the actual state vector is outside the expected state space range, it is judged that the target control strategy is not successfully executed, and the success or failure flag is a failure flag. 6.The method of claim 1, wherein, The precursor node set comprises a plurality of precursor nodes, the current state vector of each node is constructed according to the operation parameter, the operability mark of each node is obtained according to the precursor node set, the current state vector and a preset state linkage rule, and comprises: determining whether the precursor node completes operation; determining whether the node satisfies an independent operation preparation condition according to a preset state evaluation function and the current state vector; obtaining the operability mark of each node according to the state linkage rule, whether the precursor node completes operation, whether the node satisfies the independent operation preparation condition, a preset process path weight coefficient, a preset state influence weight and a preset starting threshold bias.

7. The method of claim 1, wherein the method further comprises: The candidate control strategy is evaluated according to the preset cost function, and a cost evaluation value of each candidate control strategy is obtained, and comprises: obtaining unit task energy consumption in a candidate control strategy execution process; obtaining structural difference degrees between the candidate control strategy and a preset reference standard template; obtaining a plurality of node conflict suppression items corresponding to the candidate control strategy; the candidate control strategy is evaluated according to the unit task energy consumption, the structural difference degree and the plurality of node conflict suppression items, and a cost evaluation value of each candidate control strategy is obtained.

8. The intelligent manufacturing digital twin whole-process management and control platform based on the Hongmeng system, characterized in that, The platform comprises: a construction unit configured to construct a process structure diagram according to a manufacturing process; wherein the process structure diagram comprises nodes and sequential relationships between the nodes; wherein the nodes comprise a state acquisition port and a precursor node set, and operation parameters are acquired through the state acquisition port; a marking unit configured to construct a current state vector of each node according to the operation parameters, and obtain an operability mark of each node according to the precursor node set, the current state vector and a preset state linkage rule; wherein the operability mark comprises an operable mark, and a node with the operability mark as the operable mark is defined as a target node; a strategy unit configured to acquire a current device margin, and generate a plurality of candidate control strategies for each target node according to a preset strategy parameter template, the current state vector and the current device margin; an evaluation unit configured to evaluate the candidate control strategies according to a preset cost function, and obtain a cost evaluation value of each candidate control strategy; wherein a candidate control strategy corresponding to a minimum cost evaluation value is defined as a target control strategy; a sending unit configured to send the target control strategy to a controller for execution.

9. An electronic device, comprising: The electronic device comprises a memory and a processor, the memory stores a computer program, and the processor implements the intelligent manufacturing digital twin whole-process management and control method based on the Hongmeng system in any one of claims 1 to 7 when executing the computer program.

10. A computer-readable storage medium storing a computer program, the computer program comprising instructions that, when executed by a computer, cause the computer to perform the method of any one of claims 1 to 9. The computer program is executed by the processor to implement the intelligent manufacturing digital twin whole-process management and control method based on the Hongmeng system in any one of claims 1 to 7.

Images