Cloud-edge multi-agent predictive control method and system based on digital twinning

Abstract

The application discloses a kind of communication restriction under cloud edge multi-agent digital twin control method and system, by updating the real-time state data of the physical agent of corresponding digital twin in cloud, incremental digital twin mapping model is constructed, and dynamic event trigger mechanism is introduced, when meeting the condition, incremental digital twin mapping model is updated;According to the updated incremental digital twin mapping model, the predicted state data of digital twin is obtained;According to each digital twin itself predicted state data and neighbor digital twin predicted state data, control instruction is generated, compensates time delay effect, while combining the output difference between physical agent to realize coordination consistency control.The application can significantly improve the collaborative control precision and communication efficiency of large-scale cloud edge multi-agent system, guarantee system stability and consistency convergence, and has strong applicability.

Description

Technical Field

[0001] This invention relates to the fields of industrial internet and automatic control technology, and in particular to a predictive control method and system based on digital twins and cloud-edge multi-agent systems. Background Technology

[0002] With the rapid development of the Industrial Internet and distributed control technology, multi-agent systems under cloud-edge architecture have become the core support for intelligent manufacturing, intelligent scheduling, and other scenarios. However, existing systems face many technical bottlenecks in actual deployment: on the one hand, industrial field communication resources are limited, and traditional periodic data transmission and model update methods will cause a large amount of redundant communication, exacerbate network load, and result in insufficient real-time status information and significant delays in control commands; on the other hand, physical agents generally have problems such as heterogeneous nonlinearity, unknown dynamic models, and variable operating conditions, making it difficult to establish accurate mathematical models, and traditional control methods cannot achieve accurate state tracking and coordination.

[0003] Although digital twins can achieve real-time mapping between physical entities and virtual models, existing solutions do not effectively combine event-triggered mechanisms to balance communication overhead and control performance, and lack predictive compensation strategies for communication latency. The collaborative consistency control among multiple agents is not effective, making it difficult to meet the requirements of large-scale cloud-based multi-agent systems in terms of high real-time performance, high reliability, and efficient collaboration. Summary of the Invention

[0004] To address the shortcomings of existing technologies, this invention provides a cloud-edge multi-agent digital twin control method and system under communication constraints.

[0005] On the one hand, a cloud-edge multi-agent digital twin control method under communication constraints is provided, including: Acquire real-time status data from multiple physical agents and synchronously update the corresponding digital twins in the cloud; Based on the real-time status data of the corresponding digital twin, an incremental digital twin mapping model is constructed; dynamic event triggering conditions are preset, and the incremental digital twin mapping model is updated when the conditions are met. Based on the updated incremental digital twin mapping model, the predicted state data of the digital twin is obtained; Based on the predicted state data of each digital twin and the predicted state data of its neighbors, control commands are generated to control the physical intelligent agent.

[0006] Furthermore, physical intelligent agents The corresponding digital twin received the following data: Physical intelligent agents The corresponding digital twin received the following data: ; in, express The expected output sequence at time 1. , It is a reference signal; It is a physical intelligent agent The information of the corresponding digital twin is transmitted to the physical intelligent agent. The delay between corresponding digital twins; , and These represent the forward communication channel delay and the backward communication channel delay between the physical agent and the cloud node, respectively.

[0007] Furthermore, the constructed incremental digital twin mapping model is as follows: ; ; in, Represents pseudo-partial derivatives The estimate, For backward difference operators, and These represent the output and input of the digital twin, respectively.

[0008] Furthermore, the preset dynamic event triggering conditions are as follows:

[0009] in, Indicates the trigger variable of the design , , and , For the designed adaptive trigger threshold, , .

[0010] Furthermore, physical intelligent agents The corresponding predicted state data of the digital twin is:

[0011] in ,in and These are physical intelligent agents Network latency of the forward and feedback channels to the cloud node; It is a physical intelligent agent The corresponding digital twin is based on System information prediction at time The predicted output value at time 10:00. .

[0012] Furthermore, based on the predicted state data of each digital twin and the predicted state data of its neighboring digital twins, Substitute The control commands obtained are: .

[0013] On the other hand, a cloud-edge multi-agent digital twin control system under communication constraints is provided, including: The acquisition module acquires real-time status data from multiple physical agents and synchronously updates the corresponding digital twins in the cloud. Establish a module to build an incremental digital twin mapping model based on real-time status data; preset dynamic event triggering conditions, and update the incremental digital twin mapping model when the conditions are met; The prediction module obtains the predicted state data of the digital twin based on the updated incremental digital twin mapping model. The control module generates control commands based on the predicted state data of each digital twin and the predicted state data of its neighboring digital twins, and controls the physical intelligent agents.

[0014] Furthermore, an electronic device is also provided, including: Memory, used for non-transitory storage of computer-readable instructions; and Processor, for executing the computer-readable instructions, When the computer-readable instructions are executed by the processor, they perform the method described in the first aspect above.

[0015] In another aspect, a storage medium is also provided for non-transitory storage of computer-readable instructions, wherein when the non-transitory computer-readable instructions are executed by a computer, the method described in the first aspect is performed.

[0016] In another aspect, a computer program product is also provided, including a computer program that, when run on one or more processors, is used to implement the method described in the first aspect above.

[0017] The above technical solution has the following advantages or beneficial effects: This invention discloses a cloud-edge multi-agent digital twin control method and system under communication constraints. Through incremental digital twin modeling and online updating of pseudo-partial derivatives, it can accurately map heterogeneous nonlinear physical agents. It adopts a dynamic event triggering mechanism to significantly reduce redundant data transmission and computational overhead. It relies on the twin model to achieve state advance prediction and control compensation during communication delay. It also combines the output difference coordination strategy between physical agents to achieve consistent control. This effectively improves the collaborative control accuracy, stability and convergence of large-scale cloud-edge multi-agent systems. The overall solution has strong adaptability and has outstanding practical value and application prospects in the fields of industrial internet and automatic control. Attached Figure Description

[0018] The accompanying drawings, which form part of this invention, are used to provide a further understanding of the invention. The illustrative embodiments of the invention and their descriptions are used to explain the invention and do not constitute an improper limitation of the invention.

[0019] Figure 1 This is a flowchart of a cloud-edge multi-agent digital twin control method under communication constraints, as described in Embodiment 1. Figure 2 This is a framework diagram of the cloud-edge multi-agent system described in Embodiment 1; Figure 3 This is a diagram of the digital twin predictive control system described in Example 1; Figure 4 The reference signal r(t) and the control output signal described in Example 1 The system control input response curve; Figure 5 The system control input signal described in Example 1 The contrast curve; Figure 6 This is the pseudo-partial derivative estimation update curve described in Example 1; Figure 7 This is the event trigger interval distribution curve described in Example 1. Detailed Implementation

[0020] It should be noted that the following detailed descriptions are exemplary and intended to provide further illustration of the invention. Unless otherwise specified, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this invention pertains.

[0021] It should be noted that the terminology used herein is for the purpose of describing particular embodiments only and is not intended to limit the exemplary embodiments of the invention. The terms "comprising" and "having," and any variations thereof, are intended to cover non-exclusive inclusion, for example, a process, method, system, product, or apparatus that comprises a series of steps or units is not necessarily limited to those steps or units explicitly listed, but may include other steps or units not explicitly listed or inherent to such processes, methods, products, or apparatus.

[0022] In this embodiment of the invention, "and / or" is merely a description of the relationship between related objects, indicating that three relationships can exist. For example, A and / or B can represent: A existing alone, A and B existing simultaneously, and B existing alone. Furthermore, in the description of this invention, "multiple" refers to two or more.

[0023] Furthermore, to facilitate a clear description of the technical solutions of the embodiments of the present invention, the terms "first" and "second" are used in the embodiments of the present invention to distinguish identical or similar items with essentially the same function and effect. Those skilled in the art will understand that the terms "first" and "second" do not limit the quantity or execution order, and the terms "first" and "second" are not necessarily different.

[0024] Where there is no conflict, the embodiments and features in the embodiments of the present invention can be combined with each other.

[0025] All data acquisition in this embodiment is carried out in accordance with laws and regulations and with user consent, and the data is used legally.

[0026] Example 1 This embodiment provides a cloud-edge multi-agent digital twin control method under communication constraints, such as... Figure 1 As shown, it includes the following steps: S1: Acquire real-time status data of multiple physical agents and synchronously update the corresponding digital twins in the cloud; S2: Construct an incremental digital twin mapping model based on the real-time status data of the corresponding digital twin; preset dynamic event triggering conditions, and update the incremental digital twin mapping model when the conditions are met; S3: Based on the updated incremental digital twin mapping model, obtain the predicted state data of the digital twin; S4: Based on the predicted state data of each digital twin and the predicted state data of its neighbors, generate control commands to control the physical intelligent agent.

[0027] The discrete-time multi-agent system studied in step S1 includes A heterogeneous nonlinear physical intelligent agent, such as Figure 2 As shown, the system model of the physical intelligent agent is: (1) in, It is an unknown nonlinear function. and These are physical intelligent agents exist I / O at any moment and These are physical intelligent agents The unknown order of the output and input, and .

[0028] Physical intelligent agents The corresponding digital twin in the cloud is: (2) in and These represent the output and input of the digital twin, respectively. Yes The data-driven estimation is based on the historical input-output data of the physical agent.

[0029] like Figure 3 As shown, in the digital twin predictive control system described in this embodiment, the physical intelligent agent... The sensor in The system continuously sends the data sequence from the next data packet to the corresponding digital twin in the cloud. The corresponding digital twin receives the following data: (3) in, express The expected output sequence at time step (i.e., the reference signal sequence). , It is a reference signal that the physical system needs to consistently follow; It is a physical intelligent agent The information of the corresponding digital twin is transmitted to the physical intelligent agent. The delay between corresponding digital twins; , and These represent the forward communication channel delay and the backward communication channel delay between the physical agent and the cloud node, respectively.

[0030] Based on received data, the cloud platform corrects pseudo-partial derivatives online and constructs an incremental digital twin mapping model. Constructing this incremental digital twin mapping model requires the physical intelligent agent system to meet the following conditions: function about The partial derivatives are continuous, and the physical intelligent agent... Satisfies the generalized Lipschitz: that is, for any time... , At that time, there must exist a positive constant. , making .

[0031] The incremental digital twin mapping model can be represented as: (4) (5) in, Represents pseudo-partial derivatives The estimate, It is a backward difference operator. The update formula has a saturation condition, which occurs when any of the following conditions are met. They will all be reset to their initial values. : (6) in, , , It is a small positive number. Represents a symbolic function.

[0032] The triggering mechanism for the preset dynamic event triggering condition in step S2 is as follows: (7) in, It is the trigger variable of the design. , , and , For the designed adaptive trigger threshold, , , This represents the adjustment parameter for the adaptive trigger threshold.

[0033] Step S3 involves using an incremental digital twin mapping model during the communication delay to predict the future state data of the physical agent, specifically including the physical agent's... The corresponding digital twin's predicted state data and neighboring physical agents The corresponding predicted state data of the digital twin.

[0034] Physical intelligent agents The corresponding predicted state data of the digital twin is: (8) in ,in and These are physical intelligent agents Network latency of the forward and feedback channels to the cloud node; It is a physical intelligent agent The corresponding digital twin is based on System information prediction at time The predicted output value at time 10:00. .

[0035] Neighbor physical intelligent agent The corresponding predicted state data of the digital twin is: (9) in , It is a physical intelligent agent Cloud nodes and physical intelligent agents Communication latency between cloud nodes.

[0036] In step S4, based on the predicted state data of each digital twin and the predicted state data of its neighboring digital twins, an output difference coordination strategy is designed, and control commands are calculated and sent to the physical agent. Specifically, this includes: (10) in, and If and only if hour It is assigned the value 1, otherwise ; ; The designed coordination mechanism employs a leader-follower strategy. (11).

[0037] Formula (10) is the incremental controller designed in this embodiment. Based on this controller, and considering communication transmission, the following will be implemented: d Substitute The data is obtained from the digital twin in the cloud and distributed to the physical intelligent agent. The control commands are: (12) This embodiment uses MATLAB simulation software to simulate and verify the designed controller. The simulation results show that the multi-agent system of this embodiment can effectively track a given signal and maintain a consistent state under communication constraints. Moreover, all signals are globally consistent and eventually bounded, and the consistent tracking error converges exponentially to a compact set near the origin that can be adjusted by changing parameters.

[0038] This embodiment employs a data-driven digital twin method that integrates event-triggered mechanisms to address the challenges posed by system latency and nonlinear dynamics. Figures 4-7 The curves corresponding to the physical agent's output and the reference signal, the system control input response curve, the pseudo-partial derivative estimation update curve, and the event trigger interval distribution curve respectively, can intuitively reflect the control effect and operating characteristics of the proposed method.

[0039] Depend on Figure 4 It is evident that the output trajectories of each group of physical agents can follow the reference signal, and the deviation between the output trajectories of different physical agents is small, exhibiting good cooperative consistency characteristics. Figure 5 The system control input response curve shows a smooth change without significant abrupt changes, consistent with the reasonable operating rules of the system control input. From... Figure 6 The pseudo-partial derivative estimation update curve shows that the parameter update process exhibits a stable convergence trend, reflecting the dynamic characteristics of the system. Although the system output oscillates slightly during the reference signal transition phase due to the intermittent update characteristics of the model, the overall parameter update remains stable and convergent, effectively capturing the dynamic changes of the system. Figure 7 The event trigger interval distribution shows that during the 1500-step simulation, the trigger events of each physical agent are sparsely distributed. The pseudo-partial derivatives are only updated when the trigger conditions are met. This feature significantly reduces the update frequency of the incremental twin model, thereby effectively saving the communication and computing resources of the system.

[0040] In summary, the technical solution proposed in this embodiment achieves good overall collaborative control of the system and can effectively compensate for random communication constraints in distributed cloud-edge multi-agent systems. Experimental results show that the proposed digital twin predictive control method can significantly improve system stability and control performance in complex network environments, while also possessing high computational efficiency.

[0041] Example 2 This embodiment provides a cloud-edge multi-agent digital twin control system under communication constraints, including: The acquisition module acquires real-time status data from multiple physical agents and synchronously updates the corresponding digital twins in the cloud. Establish a module to build an incremental digital twin mapping model based on real-time status data; preset dynamic event triggering conditions, and update the incremental digital twin mapping model when the conditions are met; The prediction module obtains the predicted state data of the digital twin based on the updated incremental digital twin mapping model. The control module generates control commands based on the predicted state data of each digital twin and the predicted state data of its neighboring digital twins, and controls the physical intelligent agents.

[0042] The descriptions of each embodiment in the above embodiments have different focuses. For parts not described in detail in a certain embodiment, please refer to the relevant descriptions in other embodiments.

[0043] The proposed system can be implemented in other ways. For example, the system embodiments described above are merely illustrative, and the division of modules described above is only a logical functional division. In actual implementation, there may be other division methods. For example, multiple modules may be combined or integrated into another system, or some features may be ignored or not executed.

[0044] Example 3 This embodiment also provides an electronic device, including: one or more processors, one or more memories, and one or more computer programs; wherein, the processor is connected to the memory, and the one or more computer programs are stored in the memory. When the electronic device is running, the processor executes the one or more computer programs stored in the memory to cause the electronic device to perform the method described in Embodiment 1.

[0045] It should be understood that in this embodiment, the processor can be a central processing unit (CPU), or it can be other general-purpose processors, digital signal processors (DSPs), application-specific integrated circuits (ASICs), programmable gate arrays (FPGAs), or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, etc. The general-purpose processor can be a microprocessor or any conventional processor, etc.

[0046] Memory may include read-only memory and random access memory, and provides instructions and data to the processor. A portion of memory may also include non-volatile random access memory. For example, memory may also store information about the device type.

[0047] In the implementation process, each step of the above method can be completed by the integrated logic circuits in the processor hardware or by software instructions.

[0048] The method in Embodiment 1 can be directly implemented by a hardware processor, or implemented by a combination of hardware and software modules within the processor. The software modules can reside in readily available storage media in the art, such as random access memory, flash memory, read-only memory, programmable read-only memory, electrically erasable programmable memory, or registers. This storage medium is located in memory; the processor reads information from the memory and, in conjunction with its hardware, completes the steps of the above method. To avoid repetition, a detailed description is not provided here.

[0049] Those skilled in the art will recognize that the units and algorithm steps described in connection with the various examples of this embodiment can be implemented in electronic hardware or a combination of computer software and electronic hardware. Whether these functions are implemented in hardware or software depends on the specific application and design constraints of the technical solution. Those skilled in the art can use different methods to implement the described functions for each specific application, but such implementation should not be considered beyond the scope of this invention.

[0050] Example 4 This embodiment also provides a storage medium for storing computer instructions, which, when executed by a processor, complete the method described in Embodiment 1.

[0051] Example 5 This embodiment also provides a computer program product, including a computer program that, when run on one or more processors, implements the method described in Embodiment 1.

[0052] The above description is merely a preferred embodiment of the present invention and is not intended to limit the invention. Various modifications and variations can be made to the present invention by those skilled in the art. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the scope of protection of the present invention.

Claims

1. A method for cloud-edge multi-agent digital twin control under communication restriction, characterized in that, include: Acquire real-time status data from multiple physical agents and synchronously update the corresponding digital twins in the cloud; Based on the real-time status data of the corresponding digital twin, an incremental digital twin mapping model is constructed; dynamic event triggering conditions are preset, and the incremental digital twin mapping model is updated when the conditions are met. Based on the updated incremental digital twin mapping model, the predicted state data of the digital twin is obtained; Based on the predicted state data of each digital twin and the predicted state data of its neighbors, control commands are generated to control the physical intelligent agent.

2. The cloud-edge multi-agent digital twin control method under communication constraints according to claim 1, characterized in that, Physical agent The corresponding digital twin receives the data as: ； wherein, denotes the expected output sequence at the time instant, , is a reference signal; is a physical agent information of the corresponding digital twin is transferred to the physical agent the latency between the corresponding digital twins; , and denote the forward and backward communication channel latency of the physical agent and the cloud node, respectively.

3. The cloud-edge multi-agent digital twin control method under communication constraints according to claim 1, characterized in that, The constructed incremental digital twin mapping model is as follows: ； ； wherein, denotes an estimate of the pseudo partial derivative , is a backward difference operator, and denotes the output and input of the digital twin, respectively.

4. The cloud-edge multi-agent digital twin control method under communication constraints according to claim 1, characterized in that, The preset dynamic event triggering conditions are: wherein, represents a trigger variable of the design , , and , is an adaptive trigger threshold of the design, , .

5. The cloud-edge multi-agent digital twin control method under communication constraints according to claim 1, characterized in that, Physical agent The predicted state data of the corresponding digital twin is: in ,in and These are physical intelligent agents Network latency of the forward and feedback channels to the cloud node; It is a physical intelligent agent The corresponding digital twin is based on System information prediction at time The predicted output value at time 10:

00. .

6. The cloud-edge multi-agent digital twin control method under communication constraints according to claim 1, characterized in that, Based on the predicted state data of each digital twin and the predicted state data of its neighbors, Substitute The control commands obtained are: 。 7. A cloud-edge multi-agent digital twin control system under communication constraints, characterized in that, include: The acquisition module acquires real-time status data from multiple physical agents and synchronously updates the corresponding digital twins in the cloud. Establish a module to build an incremental digital twin mapping model based on real-time status data; preset dynamic event triggering conditions, and update the incremental digital twin mapping model when the conditions are met; The prediction module obtains the predicted state data of the digital twin based on the updated incremental digital twin mapping model. The control module generates control commands based on the predicted state data of each digital twin and the predicted state data of its neighboring digital twins, and controls the physical intelligent agents.

8. An electronic device, characterized in that, include: Memory is used to store computer-readable instructions in a non-transitory manner. as well as Processor, for executing the computer-readable instructions, When the computer-readable instructions are executed by the processor, they perform the cloud-edge multi-agent digital twin control method under communication constraints as described in any one of claims 1-6.

9. A storage medium, characterized in that, Non-transitory storage of computer-readable instructions, wherein, when executed by a computer, the non-transitory computer-readable instructions perform the cloud-edge multi-agent digital twin control method under communication constraints as described in any one of claims 1-6.

10. A computer program product, characterized in that, Includes a computer program, which, when running on one or more processors, implements the cloud-edge multi-agent digital twin control method under communication constraints as described in any one of claims 1-6.

Images