A digital twin control platform

Abstract

The present invention discloses a digital twin control platform, comprising: an edge intelligent gateway with multi-protocol adaptability; a dynamic twin modeling layer, including an online transfer learning engine and a federated learning coordinator; an intelligent decision control layer, integrating an MPC-DRL hybrid algorithm and an elastic resource scheduler; and a secure interaction layer, supporting RLWE homomorphic encryption and blockchain traceability. The edge intelligent gateway of the present invention significantly improves the efficiency of multimodal data acquisition and transmission through a wavelet noise reduction algorithm and an improved Huffman encoder, combined with 5G and TSN networks, solves the processing lag problem of traditional platforms, and meets industrial real-time requirements. By adopting the physical constraint loss function of the online transfer learning engine and the differential privacy mechanism of the federated learning coordinator in the dynamic twin modeling layer, real-time model updates and distributed collaborative privacy protection are achieved, overcoming the defects of static modeling and data leakage.

Description

Technical Field

[0001] The present invention relates to the technical field of Internet of Things, and in particular to a digital twin control platform. Background Art

[0002] With the advancement of Industry 4.0 and intelligent manufacturing, digital twin technology has been widely used in fields such as industrial control and new energy management. However, existing digital twin control platforms have significant shortcomings in data acquisition and processing. Traditional edge devices have high processing latency (often exceeding 10ms), low data compression rates (<70%), and support only a single protocol, making it difficult to accommodate multimodal and multi-source data acquisition requirements, resulting in poor real-time performance. Furthermore, static or offline modeling methods are often used, making it difficult to dynamically adapt to changes in physical entities. Data sharing in distributed collaboration is prone to privacy leaks and lacks effective privacy protection mechanisms. These issues limit the efficient application of digital twin systems in complex industrial scenarios.

[0003] On the other hand, existing platforms also have shortcomings in control decision-making and security. Traditional control algorithms (such as PID or single MPC) have limited accuracy and are unable to cope with dynamic system demands. Resource scheduling is often static and cannot be optimized based on real-time load, resulting in wasted or insufficient resources. Furthermore, data transmission and storage lack strong encryption support, making them vulnerable to attacks. The decision-making process is not traceable, making it difficult to meet the high security and compliance requirements of industrial scenarios. Therefore, a new digital twin control platform is urgently needed. Summary of the Invention

[0004] The present invention aims to address the shortcomings of existing technologies by proposing a digital twin control platform. This platform addresses these issues by improving data processing efficiency, modeling dynamics, control accuracy, and security, providing reliable support for industrial intelligence.

[0005] In order to achieve the above object, the present invention adopts the following technical solutions:

[0006] A digital twin control platform includes: a multi-protocol-adaptive edge intelligent gateway for real-time collection of multimodal data of physical entities; a dynamic twin modeling layer, including an online transfer learning engine and a federated learning coordinator; an intelligent decision control layer, integrating an MPC-DRL hybrid algorithm and an elastic resource scheduler; and a secure interaction layer, supporting RLWE homomorphic encryption and blockchain traceability.

[0007] Preferably, the edge intelligent gateway has a built-in wavelet noise reduction algorithm and an improved Huffman encoder. The data compression rate of the edge intelligent gateway is ≥85%, the end-side processing delay is ≤5ms, and it supports OPC UA, MQTT and Modbus protocol conversion.

[0008] Preferably, the loss function of the online transfer learning engine is ,in is the prediction error, is the constraint term of the physical constraint equation, λ=0.1-0.5.

[0009] Preferably, the elastic resource scheduler allocates FPGA acceleration resources according to a priority preemption strategy based on the real-time load of the container cluster. The priority calculation formula is P=0.6E+0.4R, where E is the task urgency and R is the resource demand coefficient.

[0010] Preferably, the polynomial degree of the RLWE homomorphic encryption is n=1024, and the modulus is , the private key update period T and the data sampling frequency f satisfy .

[0011] Preferably, the federated learning coordinator adopts a differential privacy mechanism with Gaussian noise injection, and the noise standard deviation is =0.1×data sensitivity, privacy budget =0.5, failure probability .

[0012] Preferably, the hardware configuration of the edge intelligent gateway includes NVIDIA Jetson AGX Xavier chip and Xilinx Zynq UltraScale+ RFSoC, supports 5G communication and TSN time-sensitive network, and has a transmission bandwidth ≥1Gbps.

[0013] Preferably, the physical constraint equations include thermodynamic equations , kinematic equations , or the electrical characteristic equation , where Q is heat, c is specific heat capacity, and m is the mass of the substance. is the temperature change, F is the net external force, is the mass of the object, a is the acceleration of the object, V is the voltage, I is the current, R is the resistance, and the constraint term is calculated as the weighted sum of squared errors.

[0014] Preferably, the elastic resource scheduler integrates the Prometheus monitoring system to collect the CPU / GPU / FPGA utilization of the container cluster in real time, sets the preemption threshold to CPU ≥ 80% or GPU ≥ 90%, and uses a Kubernetes custom scheduling plug-in to implement resource allocation.

[0015] Preferably, the blockchain traceability adopts the Hyperledger structure framework, the consensus mechanism is the Practical Byzantine Fault Tolerance algorithm, the block generation time is ≤ 2 seconds, the minimum number of nodes is 4, and the data tampering detection rate is 100%.

[0016] The present invention has the following beneficial effects:

[0017] 1. The edge intelligent gateway of the present invention achieves a data compression rate of ≥85% and a terminal processing delay of ≤5ms through a wavelet noise reduction algorithm and an improved Huffman encoder. Combined with 5G and TSN networks, it significantly improves the efficiency of multimodal data acquisition and transmission, solves the processing lag problem of traditional platforms, and meets industrial real-time requirements.

[0018] 2. The dynamic twin modeling layer of the present invention adopts the physical constraint loss function of the online transfer learning engine and the differential privacy mechanism of the federated learning coordinator to achieve real-time model updates and distributed collaborative privacy protection, overcoming the defects of static modeling and data leakage.

[0019] 3. The intelligent decision control layer of the present invention integrates the MPC-DRL hybrid algorithm to improve the control accuracy of complex systems. The elastic resource scheduler dynamically allocates resources based on the priority formula and Prometheus monitoring, solving the problems of low accuracy and resource waste of traditional algorithms.

[0020] 4. The security interaction layer of the present invention supports encrypted computing through RLWE homomorphic encryption, and blockchain traceability achieves block generation time ≤ 2 seconds and tamper detection rate 100%, effectively solving data security and non-traceability issues and improving system reliability. BRIEF DESCRIPTION OF THE DRAWINGS

[0021] Figure 1 This is a block diagram of the platform system of the present invention. DETAILED DESCRIPTION

[0022] The technical solutions in the embodiments of the present invention will be clearly and completely described below in conjunction with the drawings in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, rather than all the embodiments.

[0023] Reference Figure 1 , a digital twin control platform, including: a multi-protocol adaptive edge intelligent gateway 101, used to collect multimodal data of physical entities in real time; a dynamic twin modeling layer 200, including an online transfer learning engine 201 and a federated learning coordinator 204; an intelligent decision control layer 300, integrating an MPC-DRL hybrid algorithm 301 and an elastic resource scheduler 304; a secure interaction layer 400, supporting RLWE homomorphic encryption 401 and blockchain traceability 402.

[0024] In this embodiment, the edge intelligent gateway 101 is used to collect multimodal data of physical entities in real time, supports multi-protocol adaptation, and performs edge preprocessing and transmission. The edge intelligent gateway 101 improves data collection efficiency and compatibility, reduces transmission delays, ensures the real-time nature of multi-source data, and enhances the system's adaptability to industrial scenarios; the dynamic twin modeling layer 200 is used to dynamically update the digital twin model through the online transfer learning engine 201, and the federated learning coordinator 204 coordinates distributed learning to build a real-time virtual mapping of physical entities. The dynamic twin modeling layer 200 achieves high adaptability and distributed collaboration of the model, the online transfer learning engine 201 improves modeling accuracy, and the federated learning coordinator 204 protects data privacy and optimizes the dynamic performance of the system; intelligent decision-making The control layer 300 is used to integrate the MPC-DRL hybrid algorithm 301 to optimize real-time control decisions, the elastic resource scheduler 304 dynamically allocates computing resources to support system operation, the intelligent decision-making control layer 300 improves control accuracy and resource utilization, the MPC-DRL hybrid algorithm 301 adapts to complex scenarios, and the elastic resource scheduler 304 ensures efficient operation and improves the intelligence level of the system; the secure interaction layer 400 is used to protect data privacy through RLWE homomorphic encryption 401, and the blockchain traceability 402 records operation history to ensure data interaction security and traceability. The secure interaction layer 400 enhances system security, RLWE homomorphic encryption 401 supports encrypted computing, and the blockchain traceability 402 realizes tampering detection to improve data credibility and compliance.

[0025] In the present invention, the edge intelligent gateway 101 has a built-in wavelet noise reduction algorithm and an improved Huffman encoder. The data compression rate of the edge intelligent gateway 101 is ≥85%, the end-side processing delay is ≤5ms, and it supports OPC UA, MQTT and Modbus protocol conversion.

[0026] In this embodiment, the edge intelligent gateway 101 has a built-in wavelet noise reduction algorithm and an improved Huffman encoder, which collects multimodal data of physical entities in real time and performs noise reduction and compression. It supports OPC UA, MQTT and Modbus protocol conversion to realize edge data processing and transmission. The edge intelligent gateway 101 improves data quality through the wavelet noise reduction algorithm and the improved Huffman encoder to achieve a data compression rate of ≥85%, and an end-side processing delay of ≤5ms, ensuring efficient real-time processing. Multi-protocol support enhances compatibility and optimizes data collection and transmission efficiency in industrial scenarios.

[0027] In the present invention, the loss function of the online transfer learning engine 201 is ,in is the prediction error, is the constraint term of the thermodynamic equation or kinematic equation, =0.1-0.5.

[0028] In this embodiment, the online transfer learning engine 201 uses the loss function Dynamically update the digital twin model, where is the prediction error, is the constraint term of the physical constraint equation, =0.1-0.5 adjusts the physical constraint weight to achieve real-time modeling. The online transfer learning engine 201 combines data-driven and physical laws to improve model accuracy and adaptability. Ensure compliance with actual physical properties, Flexible adjustment optimizes modeling balance, enabling the digital twin control platform to quickly respond to changes in physical entities.

[0029] In the present invention, the elastic resource scheduler 304 allocates FPGA acceleration resources according to a priority preemption strategy based on the real-time load of the container cluster. The priority calculation formula is P=0.6E+0.4R, where E is the task urgency and R is the resource demand coefficient.

[0030] In this embodiment, the elastic resource scheduler 304 allocates FPGA acceleration resources based on the real-time load of the container cluster through a priority preemption strategy. The priority calculation formula P=0.6E+0.4R comprehensively considers the task urgency E and the resource demand coefficient R, and dynamically optimizes the allocation of computing resources. The elastic resource scheduler 304 improves resource utilization efficiency. The priority formula P=0.6E+0.4R ensures that urgent tasks are executed first. FPGA acceleration supports high-performance computing and enhances the responsiveness of the digital twin control platform to dynamic loads and system stability.

[0031] In the present invention, the polynomial degree n of RLWE homomorphic encryption 401 is 1024, and the modulus , the private key update period T and the data sampling frequency f satisfy .

[0032] In this embodiment, RLWE homomorphic encryption 401 uses a polynomial of degree n=1024 and a modulus The RLWE algorithm is used to homomorphically encrypt the data, and the private key update period T is set according to the data sampling frequency f. , to ensure encryption computing and data security, RLWE homomorphic encryption 401 supports data processing in encrypted state, n=1024 and Ensure high security strength, Dynamically updating private keys improves anti-attack capabilities and enhances data privacy protection and security of the digital twin control platform.

[0033] In the present invention, the federated learning coordinator 204 adopts a differential privacy mechanism of Gaussian noise injection, and the noise standard deviation =0.1×data sensitivity, privacy budget =0.5, failure probability .

[0034] In this embodiment, the federated learning coordinator 204 uses a differential privacy mechanism with Gaussian noise injection to coordinate distributed learning. The noise standard deviation is =0.1×data sensitivity, privacy budget =0.5, failure probability ,To protect data privacy in multi-party collaboration, the federated learning coordinator 204 =0.1×sensitivity to precisely control noise, =0.5 and Ensure strong privacy protection and low failure rate, improve the security and reliability of distributed modeling in digital twin control platforms, while maintaining model accuracy.

[0035] In the present invention, the hardware configuration of the edge intelligent gateway 101 includes the NVIDIA Jetson AGX Xavier chip and the Xilinx Zynq UltraScale+ RFSoC, supporting 5G communication and TSN time-sensitive network, with a transmission bandwidth ≥1Gbps.

[0036] In this embodiment, the hardware configuration of the edge intelligent gateway 101 includes the NVIDIA Jetson AGX Xavier chip and the Xilinx Zynq UltraScale+ RFSoC, which supports 5G communication and TSN time-sensitive network, with a transmission bandwidth of ≥1Gbps, and is used to efficiently collect and process multimodal data of physical entities. The edge intelligent gateway 101 improves computing power through high-performance chips, and the 5G and TSN networks achieve low-latency transmission with a bandwidth of ≥1Gbps, enhancing the real-time performance and reliability of the data of the digital twin control platform and meeting the efficient communication requirements of industrial scenarios.

[0037] In the present invention, the physical constraint equations include the thermodynamic equations , kinematic equations , or the electrical characteristic equation , where Q is heat, c is specific heat capacity, and m is the mass of the substance. is the temperature change, F is the net external force, is the mass of the object, a is the acceleration of the object, V is the voltage, I is the current, R is the resistance, and the constraint term is calculated as the weighted sum of squared errors.

[0038] In this embodiment, the online transfer learning engine 201 introduces physical constraint equations, and uses the constraint terms Incorporating loss functions to constrain violation penalties and ensure that the model complies with physical laws, the online transfer learning engine 201 uses physical constraint equations to improve the scientific nature of the model. The weighted sum of squared errors quantifies the degree of violation, enhances the accuracy and reliability of digital twin control platform modeling, and adapts to complex physical scenarios.

[0039] In the present invention, the elastic resource scheduler 304 integrates the Prometheus monitoring system to collect the CPU / GPU / FPGA utilization of the container cluster in real time, sets the preemption threshold to CPU ≥ 80% or GPU ≥ 90%, and uses a Kubernetes custom scheduling plug-in to implement resource allocation.

[0040] In this embodiment, the elastic resource scheduler 304 integrates the Prometheus monitoring system to collect the CPU / GPU / FPGA utilization of the container cluster in real time, sets the preemption threshold, and dynamically allocates resources through the Kubernetes custom scheduling plug-in. The elastic resource scheduler 304 uses Prometheus to achieve precise load monitoring, the preemption threshold ensures efficient resource utilization, and the Kubernetes plug-in optimizes allocation flexibility, thereby improving the computing performance and system stability of the digital twin control platform.

[0041] In the present invention, blockchain traceability 402 adopts the Hyperledger Fabric framework, i.e., the hyperledger structure framework, the consensus mechanism is the Practical Byzantine Fault Tolerance, i.e., the practical Byzantine fault tolerance algorithm, the block generation time is ≤ 2 seconds, the minimum number of nodes is 4, and the data tampering detection rate is 100%.

[0042] In this embodiment, blockchain traceability 402 adopts the Hyperledger Fabric framework and PBFT consensus mechanism, with a block generation time of ≤2 seconds and a minimum number of nodes of 4. It records operation history and detects data tampering to ensure traceability credibility. Blockchain traceability 402 achieves efficient consensus through Hyperledger Fabric and PBFT, with a minimum of 4 nodes to ensure distributed reliability and a tamper detection rate of 100%, improving the data credibility and security of the digital twin control platform.

[0043] The above description is only a preferred specific embodiment of the present invention, but the scope of protection of the present invention is not limited thereto. Any technician familiar with the technical field, within the technical scope disclosed by the present invention, who makes equivalent replacements or changes based on the technical solution and inventive concept of the present invention, should be covered by the scope of protection of the present invention.

Claims

1. A digital twin control platform, characterized in that: include: A multi-protocol-adaptive edge intelligent gateway (101) for collecting multimodal data of physical entities in real time; A dynamic twin modeling layer (200) includes an online transfer learning engine (201) and a federated learning coordinator (204), wherein the online transfer learning engine (201) dynamically updates the digital twin model, and the federated learning coordinator (204) coordinates distributed learning to construct a real-time virtual mapping of the physical entity; The edge intelligent gateway (101) has a built-in wavelet noise reduction algorithm and an improved Huffman encoder, the data compression rate of the edge intelligent gateway (101) is ≥85%, the terminal side processing delay is ≤5ms, and it supports OPC UA, MQTT and Modbus protocol conversion; The loss function of the online transfer learning engine (201) is ,in is the prediction error, is the constraint term of the physical constraint equation, =0.1-0.5; The federated learning coordinator (204) adopts a differential privacy mechanism of Gaussian noise injection, and the noise standard deviation =0.1×data sensitivity, privacy budget =0.5, failure probability ; An intelligent decision control layer (300) integrating an MPC-DRL hybrid algorithm (301) and a flexible resource scheduler (304) for optimizing real-time control decisions and dynamically allocating computing resources; The elastic resource scheduler (304) allocates FPGA acceleration resources according to a priority preemption strategy based on the real-time load of the container cluster. The priority calculation formula is P=0.6E+0.4R, where P is the priority, E is the task urgency, and R is the resource demand coefficient; The secure interaction layer (400) supports RLWE homomorphic encryption (401) and blockchain traceability (402), wherein the RLWE homomorphic encryption (401) is used to protect data privacy, and the blockchain traceability (402) records the operation history for data interaction security and traceability.

2. A digital twin control platform according to claim 1, characterized in that: The polynomial degree n of the RLWE homomorphic encryption (401) is 1024, and the modulus , the private key update period T and the data sampling frequency f satisfy .

3. The digital twin control platform according to claim 1, characterized in that: The hardware configuration of the edge intelligent gateway (101) includes NVIDIA Jetson AGX Xavier chip and Xilinx Zynq UltraScale+RFSoC, supports 5G communication and TSN time-sensitive network, and has a transmission bandwidth of ≥1Gbps.

4. The digital twin control platform according to claim 1, characterized in that: The physical constraint equations include thermodynamic equations , kinematic equations , or the electrical characteristic equation , where Q is heat, c is specific heat capacity, and m is the mass of the substance. is the temperature change, F is the net external force, is the mass of the object, a is the acceleration of the object, V is the voltage, I is the current, R is the resistance, and the constraint term is calculated as the weighted sum of squared errors.

5. The digital twin control platform according to claim 1, characterized in that: The elastic resource scheduler (304) is integrated with the Prometheus monitoring system to collect the CPU / GPU / FPGA utilization of the container cluster in real time, the preemption threshold is set to CPU ≥ 80% or GPU ≥ 90%, and a Kubernetes custom scheduling plug-in is used to implement resource allocation.

6. The digital twin control platform according to claim 1, characterized in that: The blockchain traceability (402) adopts a hyperledger structure framework, the consensus mechanism is a practical Byzantine fault-tolerant algorithm, the block generation time is ≤2 seconds, the minimum number of nodes is 4, and the data tampering detection rate is 100%.

Images