Multi-energy digital twinning cooperative scheduling method, device, equipment and medium

Abstract

The invention discloses a multi-energy digital twin collaborative scheduling method, device and equipment and a medium, relates to the technical field of computers, is applied to an industrial energy management system, and comprises the following steps: carrying out data preprocessing based on operation data of industrial equipment and an edge computing node; determining a calibrated twinborn model based on a current digital twinborn model in a digital twinborn layer in the system and the preprocessing result; the current digital twin model comprises an equipment-level model and a system-level model; multi-device and multi-energy coupled collaborative energy scheduling analysis is carried out through a cloud collaborative layer in the system and related data of the calibrated twin model, and a global energy scheduling strategy is determined; verifying the global energy scheduling strategy based on a block chain verification layer and a cross-chain verification protocol in the system; and when the verification is passed, the global energy scheduling strategy is issued to the edge computing node for execution, and cooperative scheduling is completed. According to the invention, multi-equipment and multi-energy collaborative optimization is effectively realized, and the coping ability to a complex working condition scene is improved.

Description

Technical Field

[0001] This invention relates to the field of computer technology, and in particular to a multi-energy digital twin collaborative scheduling method, apparatus, equipment, and medium. Background Technology

[0002] With the rapid development of the economy and society, current industrial energy management systems face problems such as data silos, insufficient coordination between energy and production, volatility of renewable energy, poor economic efficiency of energy storage, and poor system stability.

[0003] However, existing energy dispatch solutions often focus on a single technology (such as using only blockchain, primarily for data sharing and transaction records). This results in a lack of accurate equipment operating status as a basis for practical applications, which can easily lead to resource waste or supply-demand imbalance, affecting the overall energy efficiency management effect. Furthermore, it is difficult to handle the complex coupling relationships between multiple devices and energy sources, and lacks the ability to cope with complex operating scenarios.

[0004] Therefore, how to adjust the energy distribution strategy according to the real-time operating conditions of industrial equipment in complex working scenarios is an urgent problem to be solved in the field. Summary of the Invention

[0005] In view of this, the purpose of this invention is to provide a multi-energy digital twin collaborative scheduling method, device, equipment, and medium, which can effectively realize the collaborative optimization of multiple devices and multiple energy sources, improve the ability to cope with complex operating conditions, and solve the problems of data silos, insufficient energy and production coordination, renewable energy volatility, poor energy storage economics, and poor system stability in existing solutions, thereby improving energy utilization efficiency. The specific solution is as follows:

[0006] Firstly, this application provides a multi-energy digital twin collaborative scheduling method, applied to an industrial energy management system, including:

[0007] Data preprocessing is performed based on equipment operation data from industrial equipment and local edge computing nodes to determine the preprocessing results; the edge computing nodes include edge intelligent computing models.

[0008] Based on the current digital twin model in the digital twin layer of the industrial energy management system and the preprocessing results, model calibration is performed to determine the calibrated twin model; the current digital twin model includes device-level and system-level digital twin models;

[0009] By using the cloud collaboration layer in the industrial energy management system and the relevant data from the calibrated twin model, a collaborative energy scheduling analysis involving multiple devices and multiple energy sources is performed to determine the global energy scheduling strategy.

[0010] Based on the blockchain verification layer and cross-chain verification protocol in the industrial energy management system, the global energy scheduling strategy is verified to determine the strategy verification result; the blockchain verification layer adopts a dual-chain architecture that separates private chain and consortium chain.

[0011] When the strategy verification result shows that the verification is successful, the global energy scheduling strategy is sent to the corresponding edge computing node and the strategy execution operation is triggered to complete the collaborative scheduling operation and obtain the collaborative scheduling result.

[0012] Optionally, the data preprocessing based on the equipment operation data of the industrial equipment and the local edge computing node of the equipment is performed to determine the preprocessing result; the edge computing node includes an edge intelligent computing model, comprising:

[0013] Data is collected based on heterogeneous sensors deployed on industrial equipment to determine equipment operating data;

[0014] Based on the local edge computing node of the industrial equipment, Kalman filtering is performed on the equipment operation data to complete the data noise reduction operation and obtain the noise-reduced data.

[0015] Based on the device-level energy consumption prediction model in the edge intelligent computing model, and combined with the noise-reduced data, local energy consumption is lightly predicted to determine the local energy scheduling strategy corresponding to the industrial equipment; the device-level energy consumption prediction model is a model built based on a pruned neural network.

[0016] The preprocessing result is determined based on the noise-reduced data and the local energy scheduling strategy.

[0017] Optionally, the step of performing model calibration based on the current digital twin model in the digital twin layer of the industrial energy management system and the preprocessing results to determine the calibrated twin model includes:

[0018] The preprocessed results are received by the digital twin modeling layer in the industrial energy management system according to a preset communication protocol.

[0019] Energy consumption prediction is performed based on the current digital twin model in the digital twin modeling layer and the denoised data to determine the energy consumption prediction result; wherein, the current digital twin model includes device-level and system-level digital twin models; the device-level digital twin model is an energy consumption mapping function established by synchronizing the energy consumption status of the industrial equipment in real time; the system-level digital twin model is an energy consumption prediction model coupled with electrical energy, thermal energy and cold energy established by integrating data from multiple industrial equipment.

[0020] Determine whether the difference between the energy consumption prediction result and the energy consumption prediction value in the local energy dispatch strategy is greater than a preset dynamic threshold to determine the difference judgment result;

[0021] If the difference judgment result is yes, then the current digital twin model is updated using Bayesian method based on the denoised data to determine the calibrated twin model;

[0022] The model parameters of the calibrated twin model are sent to the corresponding edge computing nodes, and a model parameter overwrite operation is triggered.

[0023] Optionally, the step of performing collaborative energy scheduling analysis involving multiple devices and multiple energy sources using the cloud-based collaboration layer in the industrial energy management system and relevant data from the calibrated twin model to determine a global energy scheduling strategy includes:

[0024] The cloud collaboration layer in the industrial energy management system receives the equipment health index generated by the digital twin layer and assigns a weight value to each piece of industrial equipment based on the equipment health index.

[0025] The cloud collaboration layer, the calibrated twin model, the weight values, the distributed optimization algorithm, and the deep learning algorithm are used to perform collaborative energy scheduling analysis involving multiple devices and multiple energy sources in order to determine the global energy scheduling strategy.

[0026] Optionally, after acquiring data based on heterogeneous sensors deployed on industrial equipment to determine equipment operating data, the method further includes:

[0027] The hash value is determined based on a preset hash algorithm and the device's operating data.

[0028] The hash value and the device signature information corresponding to the industrial equipment are uploaded to the private blockchain;

[0029] Based on the private chain and the preset improved PBFT consensus mechanism, the received hash value and device signature information are verified, and when the verification is successful, the hash value and device signature information are written into the corresponding block.

[0030] Optionally, the verification of the global energy scheduling strategy based on the blockchain verification layer and cross-chain verification protocol in the industrial energy management system includes:

[0031] Based on the aforementioned private blockchain and the pre-defined improved PBFT consensus mechanism, and in conjunction with smart contracts, the temperature constraints and grid capacity constraints of the scheduling instructions in the global energy scheduling strategy are verified to determine the strategy verification results.

[0032] Optionally, after determining the strategy verification result, the method further includes:

[0033] If the strategy verification result shows that the verification is successful, then during the execution of the global energy scheduling strategy by the edge computing node, the collected device operation data and the executed scheduling instructions are verified and stored on the private chain based on the blockchain verification layer and the dual evidence storage strategy to determine the evidence storage information.

[0034] Based on the consortium blockchain, a pre-compiled contract is invoked to verify the evidence storage information, and when the verification is successful, the consortium blockchain evidence storage operation is triggered.

[0035] After completing the notarization operations corresponding to several blocks in the consortium blockchain, the final notarization is completed by calling the smart contract corresponding to the main chain in the blockchain verification layer.

[0036] Secondly, this application provides a multi-energy digital twin collaborative scheduling device for use in an industrial energy management system, comprising:

[0037] The data preprocessing module is used to preprocess data based on the equipment operation data of industrial equipment and the local edge computing nodes of the equipment to determine the preprocessing results; the edge computing nodes include edge intelligent computing models.

[0038] The model calibration module is used to calibrate the model based on the current digital twin model in the digital twin layer of the industrial energy management system and the preprocessing results, so as to determine the calibrated twin model; the current digital twin model includes device-level and system-level digital twin models;

[0039] The scheduling analysis module is used to perform collaborative energy scheduling analysis of multiple devices and multiple energy sources through the cloud collaboration layer in the industrial energy management system and the relevant data of the calibrated twin model, so as to determine the global energy scheduling strategy.

[0040] The strategy verification module is used to verify the global energy scheduling strategy based on the blockchain verification layer and cross-chain verification protocol in the industrial energy management system, so as to determine the strategy verification result; the blockchain verification layer adopts a dual-chain architecture that separates private chain and consortium chain.

[0041] The strategy execution module is used to send the global energy scheduling strategy to the corresponding edge computing node when the strategy verification result shows that the verification is successful, and to trigger the strategy execution operation to complete the collaborative scheduling operation and obtain the collaborative scheduling result.

[0042] Thirdly, this application provides an electronic device, comprising:

[0043] Memory, used to store computer programs;

[0044] A processor is used to execute the computer program to implement the steps of the aforementioned multi-energy digital twin collaborative scheduling method.

[0045] Fourthly, this application provides a computer-readable storage medium for storing a computer program, which, when executed by a processor, implements the steps of the aforementioned multi-energy digital twin collaborative scheduling method.

[0046] As can be seen, this application is applied to an industrial energy management system. Data preprocessing is performed based on equipment operation data of industrial equipment and local edge computing nodes to determine the preprocessing result. The edge computing node includes an edge intelligent computing model. Model calibration is performed based on the current digital twin model in the digital twin layer of the industrial energy management system and the preprocessing result to determine the calibrated twin model. The current digital twin model includes device-level and system-level digital twin models. Multi-device, multi-energy coupling collaborative energy scheduling analysis is performed through the cloud collaboration layer in the industrial energy management system and related data from the calibrated twin model to determine a global energy scheduling strategy. The global energy scheduling strategy is verified based on the blockchain verification layer and cross-chain verification protocol in the industrial energy management system to determine the strategy verification result. The blockchain verification layer adopts a dual-chain architecture with separate private and consortium chains. When the strategy verification result indicates successful verification, the global energy scheduling strategy is distributed to the corresponding edge computing node, triggering strategy execution to complete the collaborative scheduling operation and obtain the collaborative scheduling result. In other words, this application first collects and preprocesses data locally on industrial equipment, then calibrates it with the current digital twin model in the digital twin layer of the industrial energy management system. Through the cloud-based collaboration layer of the industrial energy management system and the relevant data from the calibrated twin model, it performs collaborative energy scheduling analysis involving multiple devices and multiple energy sources. Then, the global energy scheduling strategy is verified through a blockchain verification layer. Upon successful verification, the global energy scheduling strategy is distributed to the corresponding edge computing nodes, triggering strategy execution to complete the collaborative scheduling operation and obtain the collaborative scheduling result. This effectively achieves collaborative optimization of multiple devices and multiple energy sources, improves the ability to cope with complex operating scenarios, and solves the problems of data silos, insufficient energy and production coordination, renewable energy volatility, poor energy storage economics, and poor system stability in existing solutions, thereby improving energy utilization efficiency. Attached Figure Description

[0047] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on the provided drawings without creative effort.

[0048] Figure 1 A flowchart of a multi-energy digital twin collaborative scheduling method provided in this application;

[0049] Figure 2 A flowchart of a specific multi-energy digital twin collaborative scheduling method provided in this application;

[0050] Figure 3 This application provides a flowchart of a multi-energy coordinated scheduling process;

[0051] Figure 4 A three-layer closed-loop feedback flowchart is provided for this application;

[0052] Figure 5 A code diagram illustrating a private blockchain instruction compliance verification contract provided in this application;

[0053] Figure 6 A flowchart of a dual-chain blockchain operation is provided for this application;

[0054] Figure 7 A schematic diagram of a green electricity transaction settlement contract for a consortium blockchain provided in this application;

[0055] Figure 8 This application provides a code diagram illustrating the process of calling a main chain smart contract to complete the final notarization.

[0056] Figure 9 An emergency intervention response flowchart is provided for this application;

[0057] Figure 10 This application provides a code diagram illustrating the issuance of a blockchain circuit breaker instruction.

[0058] Figure 11 A schematic diagram of a multi-energy digital twin collaborative scheduling device provided in this application;

[0059] Figure 12 This application provides a structural diagram of an electronic device. Detailed Implementation

[0060] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0061] Existing energy dispatch solutions often focus on a single technology (such as using only blockchain, primarily for data sharing and transaction records). This results in a lack of accurate equipment operating status as a basis for practical applications, which can easily lead to resource waste or supply-demand imbalance, affecting the overall energy efficiency management effect. Furthermore, it is difficult to handle the complex coupling relationships between multiple devices and energy sources, and lacks the ability to cope with complex operating scenarios.

[0062] To this end, this application provides a multi-energy digital twin collaborative scheduling scheme, which effectively realizes the collaborative optimization of multiple devices and multiple energy sources, improves the ability to cope with complex working conditions, and solves the problems of data silos, insufficient energy and production coordination, renewable energy volatility, poor energy storage economics and poor system stability in existing schemes, thereby improving energy utilization efficiency.

[0063] See Figure 1 As shown, this invention discloses a multi-energy digital twin collaborative scheduling method, applied to an industrial energy management system, comprising:

[0064] Step S11: Perform data preprocessing based on the equipment operation data of the industrial equipment and the local edge computing nodes of the equipment to determine the preprocessing results; the edge computing nodes include edge intelligent computing models.

[0065] In this embodiment, combined with Figure 2 As shown, this embodiment achieves a closed-loop energy management system through a four-tier architecture: edge computing layer, digital twin layer, cloud collaboration layer, and blockchain layer. The edge computing layer consists of local edge computing nodes on the device. Specifically, raw data—that is, data generated during device operation—is first collected using heterogeneous sensors (temperature, pressure, current sensors, etc., forming a sensor network) deployed on the industrial equipment. The data is preprocessed by local edge computing nodes (such as the NVIDIA Jetson Nano, a low-power AI computing module designed for edge computing). Specifically: data is collected from heterogeneous sensors deployed on the industrial equipment to determine its operating data; Kalman filtering is applied to the operating data using the local edge computing nodes to reduce noise and obtain denoised data; a lightweight local energy consumption prediction is performed based on the device-level energy consumption prediction model within the edge intelligent computing model, combined with the denoised data, to determine the corresponding local energy scheduling strategy for the industrial equipment; the device-level energy consumption prediction model is a model built based on a pruned neural network; and the preprocessing result is determined based on the denoised data and the local energy scheduling strategy.

[0066] Furthermore, during data preprocessing, Kalman filtering is performed first for noise reduction:

[0067] ;

[0068] In the formula, This is the state transition matrix; With Kalman gain, noise variance is reduced by 40-60%; The observation matrix; This represents the optimal state at time k, i.e., the data after noise reduction. for The optimal state at time (posterior estimate). The value is the measurement at time k.

[0069] Combination Figure 3 As shown, after filtering, lightweight prediction using TinyML (Tiny Machine Learning, a technology combining edge computing and machine learning) will be performed. The device-level energy consumption prediction model in this process employs a pruned neural network.

[0070] ;

[0071] In the formula, Here, x is the activation function (Sigmoid or ReLU6), and x is the input data of the model. The bias term after pruning is optimized in sync with the weight matrix. This is the pruned weight matrix, where || (double vertical bar symbol) represents the norm or modulus of the vector. This represents the output of the predictive model, which is a local optimization strategy (such as motor start / stop suggestions), with a computational latency controlled within 50ms. The urgency of the model's output is analyzed; if urgent, the equipment is shut down; otherwise, twin synchronization is performed, i.e., data is synchronized to the digital twin layer.

[0072] After the above preprocessing is completed, the data will be uploaded to the digital twin layer via the OPC UA / MQTT protocol (Open Platform Communications Unified Architecture / Message Queuing Telemetry Transport) so that the layer can build a dual-scale model at both the device and system levels.

[0073] Furthermore, it should be understood that after determining the equipment operation data, this embodiment will perform edge data preprocessing and private chain notarization: First, based on a preset hash algorithm and the equipment operation data, a hash value is determined; then, the hash value and the equipment signature information corresponding to the industrial equipment are uploaded to the private chain; based on the private chain and the preset improved PBFT (Practical Byzantine Fault) consensus mechanism, the received hash value and the equipment signature information are verified, and when the verification is successful, the hash value and the equipment signature information are written into the corresponding block.

[0074] The raw data collected by the device is used to generate data fingerprints through lightweight hash calculations (such as using the SHA-256 algorithm). :

[0075] ;

[0076] In the formula, This indicates the device's identification information; This indicates the timestamp when the raw data was collected. Then, the hash value... Together with device signature The data is submitted to the private chain in the blockchain and written into the block after being verified by the improved PBFT consensus, so that the private chain can synchronize the data with the consortium chain in the blockchain to realize the dual evidence storage process.

[0077] Step S12: Based on the current digital twin model in the digital twin layer of the industrial energy management system and the preprocessing results, perform model calibration to determine the calibrated twin model; the current digital twin model includes device-level and system-level digital twin models.

[0078] In this embodiment, combined with Figure 2 The twin model in the digital twin layer also makes predictions and outputs results. After obtaining the results, combine them with... Figure 4As shown, continuous self-optimization of energy dispatch is achieved through a three-level feedback loop of edge-digital twin-cloud. The digital twin layer predicts output based on the local model of the device. Real-time calibration of the edge-digital twin layer: The preprocessing results uploaded according to a preset communication protocol are received through the digital twin modeling layer in the industrial energy management system; energy consumption prediction is performed based on the current digital twin model in the digital twin modeling layer and the denoised data to determine the energy consumption prediction result; wherein, the current digital twin model includes device-level and system-level digital twin models; the device-level digital twin model is an energy consumption mapping function established by synchronizing the energy consumption status of the industrial equipment in real time; the system-level digital twin model is an energy consumption prediction model coupled with electrical energy, thermal energy, and cold energy established by integrating data from multiple industrial equipment; the difference between the energy consumption prediction result and the energy consumption prediction value in the local energy scheduling strategy is determined to be greater than a preset dynamic threshold to determine the difference judgment result; if the difference judgment result is yes, the current digital twin model is updated using Bayesian methods based on the denoised data to determine the calibrated twin model; the model parameters of the calibrated twin model are sent to the corresponding edge computing node, and a model parameter overwrite operation is triggered.

[0079] Specifically, regarding real-time calibration of the edge-digital twin layer, the trigger condition is: when the device's local model prediction output... With the output of the current digital twin model in the digital twin layer Deviation between Exceeding the dynamic threshold Time (initial value 5%, increasing linearly with equipment aging): And there exists an expression as shown below:

[0080] ;

[0081] In the formula, The denominator of the above formula, representing the rated power of the equipment, helps prevent false triggering under low load conditions. If the value does not exceed the threshold, normal execution is performed without triggering calibration.

[0082] If the obtained deviation value exceeds the threshold, the calibration procedure is executed:

[0083] 1) Edge computing nodes upload raw data and local model prediction parameters To the digital twin layer;

[0084] 2) Initiating online Bayesian updates for the digital twin model:

[0085] ;

[0086] In the formula, A B indicates that A is proportional to B; p() is the prior probability function, and p(A|B) represents the probability of observing the current data A given a set of parameters B. The hyperparameters of the digital twin model are adjusted through Markov Chain Monte Carlo (MCMC) sampling.

[0087] 3) Updated model parameters The data is distributed to edge computing nodes, overwriting the original parameter set.

[0088] It's important to understand that the device-level twin model within the digital twin layer is an energy consumption mapping function established by synchronizing the physical device states in real time. :

[0089] ;

[0090] In the formula, Gaussian noise is dynamically corrected through online learning. Let T represent the device state of the i-th device, and let T denote the transpose of the matrix. It is a mapping function.

[0091] See Figure 3 The LSTM prediction shown is a system-level twin model that integrates data from multiple devices to establish an electrical-thermal-cold coupled LSTM prediction network.

[0092] ;

[0093] In the formula, The hidden state is the memory core of the LSTM network. It is a vector that encapsulates information about all the input history sequences before time step t, and is used to predict the future. This is the memory state of the previous time step (t-1); To integrate multi-device data input into the LSTM model at time step t; It is the set of all trainable weights and bias parameters that the model learns from historical data during the training phase; The total system energy consumption for the next 24 hours is predicted in this embodiment with an error rate of <4%; For a hidden state of higher dimensions Mapped to a single predicted value The weight matrix.

[0094] Step S13: Through the cloud collaboration layer in the industrial energy management system and the relevant data of the calibrated twin model, perform collaborative energy scheduling analysis of multi-device and multi-energy coupling to determine the global energy scheduling strategy.

[0095] In this embodiment, the cloud-based collaborative layer in the industrial energy management system receives twin model data from the digital twin layer and solves for a strategy. Specifically, the cloud-based collaborative layer in the industrial energy management system receives the equipment health index generated by the digital twin layer and assigns weight values ​​to each industrial device based on the equipment health index. Through the cloud-based collaborative layer, the calibrated twin model, the weight values, the distributed optimization algorithm, and the deep learning algorithm, a collaborative energy scheduling analysis involving multiple devices and multiple energy sources is performed to determine the global energy scheduling strategy.

[0096] Specifically, in combination Figure 3 As shown, the cloud-based collaboration layer uses distributed optimization algorithms (such as ADMM, Alternating Direction Method of Multipliers) to solve the global scheduling strategy:

[0097] ;

[0098] In the formula, N is the total number of devices / units; For local decision variables, i.e., the local scheduling scheme of the i-th device; This is the average of the local variables across all devices; The local cost function is the scheduling scheme executed by the i-th device. The price paid; For global cost function; These are the physical and security constraints that the i-th device itself must satisfy; , , For linear coupling constraints, describe how the decision of the i-th device interacts with the global variable z through a linear relationship. Iterative updates are performed using the alternating direction multiplier method:

[0099] ;

[0100] In the formula, Let be the local decision variable for the (k+1)th iteration; This represents the average value of all device local variables in the (k+1)th iteration; For penalty parameters; For Lagrange multipliers, information is contained about the degree of constraint violation between the local decision of the i-th device and the global formulation z; This represents the average value of all device local variables in the k-th iteration.

[0101] The global optimization strategy obtained through the above operations will be dynamically distributed to edge computing nodes by the Kubernetes (k8s) platform for execution.

[0102] In addition, combined Figure 4 As shown, in this embodiment, the edge-digital twin-cloud three-level feedback also includes strategy iteration at the digital twin-cloud layer:

[0103] (1) Adaptive optimization loop:

[0104] 1) The cloud receives the device health index generated by the twin layer every 15 minutes (adaptable to other intervals). ;

[0105] 2) The cloud-based scheduling algorithm dynamically adjusts device weights based on HI:

[0106] ;

[0107] In the formula, The adjusted device weight for the i-th device; Let i be the original device weight of the i-th device; Let be the device health index of the i-th device. This way, devices with poor health indices will have a lower weight in the optimization objective.

[0108] 3) Updated scheduling strategy It will be issued after being verified by the blockchain.

[0109] (2) Synchronization of elastic parameters:

[0110] 1) Normal operating conditions: The twin model synchronizes all parameters to the cloud daily. ;

[0111] 2) Abnormal operating conditions Trigger instant incremental synchronization The latency is less than 30 seconds.

[0112] Step S14: Based on the blockchain verification layer and cross-chain verification protocol in the industrial energy management system, the global energy scheduling strategy is verified to determine the strategy verification result; the blockchain verification layer adopts a dual-chain architecture that separates private chain and consortium chain.

[0113] Combination Figure 2As shown in this embodiment, after determining the global energy dispatch strategy, it needs to be verified based on the blockchain verification layer and cross-chain verification protocol to determine whether it can be issued for execution. Specifically, based on the private chain and the preset improved PBFT consensus mechanism, and combined with smart contracts, the temperature constraints and grid capacity constraints of the dispatch instructions in the global energy dispatch strategy are verified to determine the strategy verification result. The blockchain verification layer adopts a dual-chain architecture, a private chain-consortium chain separation architecture, and achieves collaborative management of sensitive instructions and public transactions through a cross-chain verification protocol. The private chain focuses on the secure verification of dispatch instructions, while the consortium chain manages the storage of energy transaction evidence. The two achieve trusted data interaction through a zero-knowledge proof cross-chain bridge.

[0114] Regarding private blockchains (scheduling instruction chains):

[0115] 1) The consensus mechanism is an improved PBFT protocol, used to introduce reputation-weighted voting. Node voting weights are determined by both reputation score and real-time network quality.

[0116] ;

[0117] In the formula, Calculate the reputation score for the j-th edge node; The real-time network quality of the j-th edge computing node; The voting rights of the j-th edge computing node; Calculate the reputation score for the k-th edge node; Let be the real-time network latency of the j-th node; Let be the real-time network latency of the k-th node; The fault tolerance threshold is satisfied. , This is a rounding operation, meaning the value is rounded down to the nearest integer. Consensus latency. (Measured value).

[0118] 2) Smart contract design, including instruction compliance verification contract, digital signature verification module, and data storage structure. The instruction compliance verification contract, for example... Figure 5 As shown, this is used to verify the temperature constraints and grid capacity constraints of the scheduling instructions in the global energy scheduling strategy. The digital signature verification module is used to verify the identity of edge computing nodes using the Elliptic Curve Digital Signature Algorithm (ECDSA).

[0119] ;

[0120] In the formula, k is the temporary key; Q is the public key; Verify() is the signature verification function; M is the original message to be verified; (r, s) is a pair of values ​​generated by the edge node after signing the message M with its private key; G is the base point of the elliptic curve; x represents the point x; n is a publicly known large prime number in the elliptic curve cryptosystem; mod is the modulo operator, which represents the remainder after dividing two numbers.

[0121] Furthermore, the data storage structure uses a Merkle Patricia Tree (MPT) to store instruction hashes and root hashes. It anchors to the consortium blockchain every two minutes.

[0122] About consortium blockchains (energy trading blockchains):

[0123] 1) Its consensus mechanism is a hybrid DPoS+BFT consensus, including super node election: every 24 hours based on the amount of tokens staked. and contribution value Ranking: The top 21 nodes are selected as supernodes, taking turns producing blocks every 3 seconds; Byzantine Fault Tolerance (BFT): A block must obtain... Each node signs to confirm. This is a rounding operation, which rounds the value up.

[0124] 2) Combining Figure 6 As shown, its privacy protection technologies include zk proofs of transaction data, also known as zk-SNARKs (Zero-Knowledge Succinct Non-Interactive Argument of Knowledge, a protocol that generates zero-knowledge proofs to verify the authenticity of information without exposing the underlying data), homomorphic encrypted storage, and smart contract functionality. The smart contract functionality includes, for example... Figure 7 The green electricity trading settlement contract shown; the circuit constructed by the zk proof satisfies:

[0125] ;

[0126] In the formula, n is the total number of constraints in this circuit; The weighting coefficients are publicly known. The public result is calculated based on the secret input and the public coefficient. For secret input, that is, transaction details data that need to be kept confidential; For use in restricting secret input The reasonable range of values. In this way, the prover generates the proof. Without leaking (Transaction details).

[0127] Furthermore, homomorphic encryption notarization includes the encrypted storage of energy transaction data:

[0128] ;

[0129] In the formula, m represents the original transaction data that needs to be encrypted; The ciphertext after m is encrypted; This is the public key parameter; It is a random number; It is a multiplicative group of integers modulo n; where n is the product of large prime numbers. Verifiable by regulators. Without decryption, among which , This is the original transaction data.

[0130] Regarding the cross-chain collaboration mechanism, it includes a cross-chain verification bridge and a dual notarization process. Specifically, in the cross-chain verification bridge, the private chain instruction hash is synchronized to the consortium chain via zero-knowledge proof.

[0131] 1) Proof of Private Chain Generation: ,in, ZK represents the instruction hash stored in the Merkel-Patricia tree (MPT) in the private chain;

[0132] 2) Consortium blockchain verification: ;

[0133] 3) After successful verification, a notarization event is generated on the consortium blockchain.

[0134] Furthermore, the dual-certification process utilizes a three-tier architecture—private chain, consortium chain, and main chain—to achieve end-to-end trusted data certification. The specific steps are as follows:

[0135] ① First, perform edge data preprocessing and private chain notarization as described in step S11. The private chain constructs a Merkel-Patricia Tree (MPT) to store all instruction hashes, generates a root hash, and synchronizes it to the consortium chain periodically (every 2 minutes).

[0136] ② Zero-knowledge proof cross-chain verification. Private blockchains generate existence proofs. Prove a specific hash It belongs to the current state tree. The expression generated by the proof is shown below:

[0137] ;

[0138] In the formula, path is Path proof in MPT; This is a Merkel-Patricia tree; SNARK.Prove() is the proof generator function. Subsequent cross-chain submissions will... The verification contract is sent to the consortium blockchain. The consortium blockchain calls the pre-compiled contract to perform proof verification to complete on-chain verification. Upon successful verification, a consortium blockchain notarization event is triggered.

[0139] ③ Multi-dimensional evidence storage on the consortium blockchain. The consortium blockchain performs the following parallel operations: energy transaction record and instruction hash anchoring, and compound hash generation. Among these, the energy transaction record stores encrypted transaction data. It is implemented using Paillier homomorphic encryption:

[0140] .

[0141] Furthermore, the instruction hash is anchored as follows: embedded in the transaction metadata. Establish a connection with the private blockchain; generate compound hashes by calculating transaction batch hashes. ,in, This is a collection of transactions that currently need to be packaged and stored.

[0142] ④ Main chain final confirmation. The consortium blockchain will update the latest state every 10 blocks (approximately 30 seconds). Submitting to the Ethereum main chain via a relay, and then calling the main chain smart contract to complete the final notarization, such as... Figure 8 As shown, the main chain's notarized transaction hash As the ultimate trusted credential.

[0143] ⑤ Closed-loop verification mechanism. Any participant can verify data integrity through the following path: Reverse tracing: Lightweight node verification: Edge devices only need to obtain the Merkle path for verification. Validity:

[0144] ;

[0145] In the formula, The Merkle root hash of the current block in the private chain; A set used for proof A sequence of hash values ​​existing in a Merkle tree.

[0146] (4) Performance optimization design.

[0147] ① Layered and sharded architecture. Private blockchains are partitioned by factory (e.g., area A, area B), while consortium blockchains are sharded by energy type (photovoltaic, wind power, etc.). Cross-shard communication uses an atomic locking protocol.

[0148] ;

[0149] In the formula, A cryptographic commitment used to lock the transaction state; Commit() is a commit action: once confirmed... Once successful locking has been achieved on all relevant shards, the coordinator issues a command to formally commit the transaction on both the left and right shards; Release() is a release action, and after the transaction is committed, the previously generated shards are used. As proof, release or clear the temporary state set during the locked phase.

[0150] ② Lightweight node verification. Edge devices verify the validity of commands via Merkle Proof:

[0151] .

[0152] Step S15: When the strategy verification result shows that the verification is successful, the global energy scheduling strategy is sent to the corresponding edge computing node, and the strategy execution operation is triggered to complete the collaborative scheduling operation and obtain the collaborative scheduling result.

[0153] In this embodiment, combined with Figure 3 As shown, when the policy verification result indicates successful verification, Kubernetes (K8s) distributes the global energy scheduling policy to the edge computing nodes for execution. And as... Figure 4 As shown, the system will also learn from the feedback provided during execution, specifically through a three-level feedback loop between the edge, digital twin, and cloud layers. This enables continuous self-optimization of energy dispatch and coordinated operation across all levels. In addition to the real-time calibration of the edge-digital twin layer and the strategy iteration of the digital twin-cloud layer described above, the three-level feedback loop also includes emergency intervention at the cloud-edge layer.

[0154] Regarding emergency intervention at the cloud-edge layer, combined with Figure 9 As shown, the specific process is as follows:

[0155] (1) Establish a three-level dynamic response system and implement differentiated emergency strategies according to the severity of the fault.

[0156] 1) Level 1 Alarm (Device-level Emergency Response):

[0157] When the temperature of a single device is detected to exceed 90% of the safety threshold (i.e. When the temperature exceeds a certain threshold, the edge computing layer immediately initiates an autonomous shutdown protocol. This response is executed entirely locally, implemented through pre-configured hard interrupt logic in the smart gateway, with response latency strictly controlled within 100 milliseconds. Simultaneously, it triggers blockchain light node notarization, recording the device ID (IdentityDocument), the over-temperature value, and the action timestamp. .

[0158] 2) Level 2 Alarm (Regional Collaborative Intervention):

[0159] When the total power in a certain area exceeds the rated capacity by 15% (i.e.) When this happens, the system initiates a dual-track response: the cloud-based rescheduling engine re-solves the optimization problem within 5 seconds.

[0160] ;

[0161] In the formula, The original scheduling scheme; To optimize the algorithm, the output of the solution is required; for A single element in a vector; For the safety capacity of the power grid; The goal is to find a new scheduling scheme that is as close as possible to the original optimal plan, based on the L2 norm, which is the square root of the sum of the squares of each element in the matrix. Furthermore, blockchain circuit breaker instructions are automatically issued via smart contracts, such as... Figure 10 As shown.

[0162] 3) Level 3 Alarm (System-level Disaster Response):

[0163] When a cascading failure occurs in multiple areas (such as a grid frequency deviation > 1.5 Hz), the system executes a defense-in-depth strategy:

[0164] ① Instantaneous backup power access: Automatically call upon backup power through pre-deployed smart contracts:

[0165] ;

[0166] In the formula, This represents the actual output power of the backup power supply. The maximum output power that the energy storage system can safely provide at the current moment; It is the difference between the system's current total power demand and the safe capacity that the power grid can provide.

[0167] ② Emergency rollback of digital twin model: Restore the twin parameters to the latest stable version. (Version history based on blockchain evidence), among which, For parameters of the digital twin model; This refers to the parameter version of the digital twin model 720 minutes prior to the current time t.

[0168] ③ Global Strategy Reset: The cloud optimization engine switches to conservative mode, and the objective function weights are adjusted to... ,in, For stability weights, Economic weighting.

[0169] The entire process described above is completed within 1 minute, prioritizing system survivability.

[0170] Furthermore, regarding the implementation of the feedback loop for the aforementioned emergency intervention, in this embodiment:

[0171] 1) Edge actuators will consume actual energy. The status code (STATUS) is fed back to the cloud.

[0172] 2) Optimize engine to calculate target deviation :

[0173] ;

[0174] In the formula, For planned energy consumption.

[0175] 3) According to Dynamically adjust the penalty coefficient of the ADMM algorithm The updated penalty coefficient is obtained. :

[0176] .

[0177] This embodiment employs a three-layer closed-loop feedback mechanism—"cloud-edge-device"—simplifying operation and control processes. Users only need to set basic parameters, and the system automatically generates the optimal scheduling strategy and distributes it to edge nodes for execution, requiring no manual intervention. Furthermore, the Kubernetes platform dynamically manages edge node resources, flexibly adjusting computing tasks according to actual needs, further enhancing system usability. Compared to traditional manual adjustment methods, this system significantly reduces the workload of maintenance personnel, lowers operational difficulty, and reduces the probability of errors.

[0178] Furthermore, in this embodiment, if the verification result shows that the verification is successful, then during the execution of the global energy scheduling strategy by the edge computing node, based on the blockchain verification layer and the dual notarization strategy, the collected device operation data and the executed scheduling instructions are verified and notarized on the private chain to determine the notarization information; based on the consortium chain, the pre-compiled contract is called to verify the notarization information, and when the verification is successful, the consortium chain notarization operation is triggered; after completing the consortium chain notarization operation corresponding to several blocks in the consortium chain, the final notarization is completed by calling the smart contract corresponding to the main chain in the blockchain verification layer.

[0179] In summary, the blockchain-based multi-energy digital twin collaborative scheduling scheme proposed in this embodiment demonstrates significant advantages over existing technologies in several aspects. These improvements not only enhance the overall system performance but also bring considerable economic and social benefits in practical applications.

[0180] (1) Improve energy efficiency. By integrating digital twin modeling with lightweight artificial intelligence models (Kalman filtering, TinyML), this system can achieve accurate energy consumption prediction and optimization at both the device and system levels. Specifically, edge computing nodes collect and process industrial equipment operation data in real time, and combine them with localized artificial intelligence models to make preliminary optimization decisions and generate efficient scheduling strategies. In experimental testing, for a real-world case of a steel plant, the system reduced overall energy consumption by approximately 20%. In contrast, traditional energy management systems often fail to achieve such high energy-saving effects due to the lack of real-time monitoring and dynamic adjustment mechanisms. Furthermore, the cloud platform uses distributed optimization algorithms (such as ADMM) to coordinate energy allocation among multiple devices and combines deep learning models (such as LSTM) to predict long-term energy consumption trends, further optimizing the global scheduling strategy and significantly improving the energy utilization rate of the entire industrial park.

[0181] (2) Enhancing Data Credibility. Blockchain technology is used to store the version hashes of the digital twin model and energy transaction records, ensuring data immutability. Through the Hyperledger Fabric consortium blockchain and DPoS consensus mechanism, the parameter hashes of each model update are uploaded to the blockchain, forming an immutable version traceability chain. Smart contracts automatically execute energy trading rules (carbon quota allocation, demand response incentives) and manage enterprise access permissions through an ABAC strategy, protecting sensitive data privacy. Experimental results show that the accuracy rate of data verified using blockchain reaches 99.9%, far exceeding that of traditional systems that do not use blockchain (typically around 95%). This not only improves data credibility but also enhances trust between entities, promoting data sharing and collaboration.

[0182] (3) Reduce system latency. Edge computing nodes deploy low-power embedded hardware (such as NVIDIA Jetson Nano) to run lightweight artificial intelligence models for local energy consumption prediction and optimization, greatly reducing reliance on the cloud and lowering network latency. Experiments show that the response time of localized decision-making is less than 200ms, while traditional cloud-based processing methods require several seconds or even longer to complete the same task. This rapid response capability is particularly important for applications with high real-time requirements (such as power load regulation and emergency fault handling), effectively avoiding resource waste or security risks caused by latency.

[0183] (4) Support for cross-entity collaboration. The combination of blockchain and digital twin technologies enables secure data sharing and collaborative optimization among different enterprises and devices. In the green electricity trading case between photovoltaic power plants and microgrids, smart contracts automatically settle transactions, reducing manual review costs and improving transaction transparency and efficiency. Compared to traditional manual review methods, the automated trading process shortens the settlement cycle by approximately 30% and reduces the risk of human error. Simultaneously, multi-energy coupling modeling (such as combined heat and power systems and energy storage scheduling models) supports collaborative optimization in complex scenarios, enabling more efficient mutual complementarity between different energy forms and improving the overall flexibility and stability of the energy system.

[0184] (5) Enhanced System Robustness. By verifying discrepancies between virtual and real data through smart contracts and automatically triggering anomaly handling mechanisms, the system's robustness is enhanced. For example, in the optimization case of a central air conditioning system, the digital twin model predicts load changes in real time, optimizes fan speed and cooling water flow, achieving energy savings of 15%. More importantly, when a significant deviation between the predicted and actual values ​​is detected, the system automatically adjusts the model parameters to reduce error accumulation and ensure long-term stable operation. Experimental data shows that after multiple iterations of optimization, the model's prediction accuracy improved by approximately 50%, and the system's stability was significantly enhanced.

[0185] (6) Simplify operation and control. A three-layer closed-loop feedback mechanism of "cloud-edge-device" was designed to simplify the operation and control process.

[0186] In summary, this embodiment, by introducing multiple advanced technologies such as blockchain, digital twins, edge computing, and cloud-based collaborative optimization, not only significantly improves energy utilization efficiency, enhances data credibility, and reduces system latency, but also supports cross-entity collaboration, improves system robustness, and simplifies operation and control processes. These improvements have achieved remarkable results in practical applications, providing a brand-new solution for industrial energy management.

[0187] Therefore, this application first collects and preprocesses data locally on industrial equipment, then calibrates it with the current digital twin model in the digital twin layer of the industrial energy management system. Through the cloud-based collaboration layer of the industrial energy management system and the relevant data from the calibrated twin model, it performs collaborative energy scheduling analysis involving multiple devices and multiple energy sources. Then, the global energy scheduling strategy is verified through a blockchain verification layer. Upon successful verification, the global energy scheduling strategy is distributed to the corresponding edge computing nodes, triggering strategy execution to complete the collaborative scheduling operation and obtain the collaborative scheduling result. This effectively achieves collaborative optimization of multiple devices and multiple energy sources, improves the ability to cope with complex operating scenarios, and solves the problems of data silos, insufficient energy and production coordination, renewable energy volatility, poor energy storage economics, and poor system stability in existing solutions, thereby improving energy utilization efficiency.

[0188] See Figure 11 As shown in the illustration, this application also discloses a multi-energy digital twin collaborative scheduling device, applied to an industrial energy management system, comprising:

[0189] Data preprocessing module 11 is used to perform data preprocessing based on the equipment operation data of industrial equipment and the local edge computing nodes of the equipment to determine the preprocessing results; the edge computing nodes include edge intelligent computing models.

[0190] The model calibration module 12 is used to perform model calibration based on the current digital twin model in the digital twin layer of the industrial energy management system and the preprocessing results, so as to determine the calibrated twin model; the current digital twin model includes device-level and system-level digital twin models;

[0191] The scheduling analysis module 13 is used to perform collaborative energy scheduling analysis of multiple devices and multiple energy sources through the cloud collaboration layer in the industrial energy management system and the relevant data of the calibrated twin model, so as to determine the global energy scheduling strategy.

[0192] The strategy verification module 14 is used to verify the global energy scheduling strategy based on the blockchain verification layer and cross-chain verification protocol in the industrial energy management system to determine the strategy verification result; the blockchain verification layer adopts a dual-chain architecture that separates private chain and consortium chain.

[0193] The strategy execution module 15 is used to send the global energy scheduling strategy to the corresponding edge computing node when the strategy verification result shows that the verification is successful, and to trigger the strategy execution operation to complete the collaborative scheduling operation and obtain the collaborative scheduling result.

[0194] In some specific embodiments, the data preprocessing module 11 can be used to: collect data based on heterogeneous sensors deployed on industrial equipment to determine equipment operating data; perform Kalman filtering on the equipment operating data based on the local edge computing nodes of the industrial equipment to complete data denoising and obtain denoised data; perform lightweight local energy consumption prediction based on the device-level energy consumption prediction model in the edge intelligent computing model and in combination with the denoised data to determine the local energy scheduling strategy corresponding to the industrial equipment; the device-level energy consumption prediction model is a model built based on a pruned neural network; and determine the preprocessing result based on the denoised data and the local energy scheduling strategy.

[0195] In some specific embodiments, the model calibration module 12 can be used to: receive the preprocessing results uploaded according to a preset communication protocol through the digital twin modeling layer in the industrial energy management system; perform energy consumption prediction based on the current digital twin model in the digital twin modeling layer and the denoised data to determine the energy consumption prediction result; wherein, the current digital twin model includes device-level and system-level digital twin models; the device-level digital twin model is an energy consumption mapping function established by synchronizing the energy consumption status of the industrial equipment in real time; the system-level digital twin model is an energy consumption prediction model coupled with electrical energy, thermal energy and cold energy established by integrating the data of multiple industrial equipment; determine whether the difference between the energy consumption prediction result and the energy consumption prediction value in the local energy scheduling strategy is greater than a preset dynamic threshold to determine the difference judgment result; if the difference judgment result is yes, perform a Bayesian update on the current digital twin model based on the denoised data to determine the calibrated twin model; and send the model parameters of the calibrated twin model to the corresponding edge computing node and trigger a model parameter overwrite operation.

[0196] In some specific embodiments, the scheduling analysis module 13 can be used to: receive the equipment health index generated by the digital twin layer through the cloud collaboration layer in the industrial energy management system, and assign weight values ​​to each of the industrial equipment based on the equipment health index; and perform collaborative energy scheduling analysis of multiple devices and multiple energy sources through the cloud collaboration layer, the calibrated twin model, the weight values, the distributed optimization algorithm, and the deep learning algorithm to determine the global energy scheduling strategy.

[0197] In some specific embodiments, the multi-energy digital twin collaborative scheduling device can also be used to: determine a hash value based on a preset hash algorithm and the equipment operating data; upload the hash value and the equipment signature information corresponding to the industrial equipment to the private chain; verify the received hash value and the equipment signature information based on the private chain and a preset improved PBFT consensus mechanism, and write the hash value and the equipment signature information into the corresponding block when the verification is successful.

[0198] In some specific embodiments, the strategy verification module 14 can be used to: verify the temperature constraints and grid capacity constraints of the scheduling instructions in the global energy scheduling strategy based on the private chain and the preset improved PBFT consensus mechanism, and in combination with smart contracts, so as to determine the strategy verification result.

[0199] In some specific embodiments, the multi-energy digital twin collaborative scheduling device can also be used to: if the strategy verification result shows that the verification is successful, then during the execution of the global energy scheduling strategy by the edge computing node, verify and store the collected device operation data and the executed scheduling instructions on the private chain based on the blockchain verification layer and the dual storage strategy to determine the storage information; based on the consortium chain, call the pre-compiled contract to verify the storage information, and when the verification is successful, trigger the consortium chain storage operation; after completing the consortium chain storage operation corresponding to several blocks in the consortium chain, complete the final storage by calling the smart contract corresponding to the main chain in the blockchain verification layer.

[0200] Furthermore, embodiments of this application also disclose an electronic device, Figure 12 This is a structural diagram of an electronic device 20 according to an exemplary embodiment. The content of the diagram should not be construed as limiting the scope of this application.

[0201] Figure 12 This is a schematic diagram of the structure of an electronic device 20 provided in an embodiment of this application. Specifically, the electronic device 20 may include: at least one processor 21, at least one memory 22, a power supply 23, a communication interface 24, an input / output interface 25, and a communication bus 26. The memory 22 stores a computer program, which is loaded and executed by the processor 21 to implement the relevant steps in the multi-energy digital twin collaborative scheduling method disclosed in any of the foregoing embodiments. Furthermore, the electronic device 20 in this embodiment may specifically be an electronic computer.

[0202] In this embodiment, the power supply 23 is used to provide operating voltage for each hardware device on the electronic device 20; the communication interface 24 can create a data transmission channel between the electronic device 20 and external devices, and the communication protocol it follows can be any communication protocol applicable to the technical solution of this application, and is not specifically limited here; the input / output interface 25 is used to acquire external input data or output data to the outside world, and its specific interface type can be selected according to specific application needs, and is not specifically limited here.

[0203] In addition, the memory 22, as a carrier for resource storage, can be a read-only memory, random access memory, disk or optical disk, etc. The resources stored thereon can include operating system 221, computer program 222, etc., and the storage method can be temporary storage or permanent storage.

[0204] The operating system 221 is used to manage and control the various hardware devices on the electronic device 20 and the computer program 222, which may be Windows Server, Netware, Unix, Linux, etc. In addition to including a computer program capable of performing the multi-energy digital twin collaborative scheduling method executed by the electronic device 20 as disclosed in any of the foregoing embodiments, the computer program 222 may further include a computer program capable of performing other specific tasks.

[0205] Furthermore, this application also discloses a computer-readable storage medium for storing a computer program; wherein, when the computer program is executed by a processor, it implements the aforementioned multi-energy digital twin cooperative scheduling method. Specific steps of this method can be found in the corresponding content disclosed in the foregoing embodiments, and will not be repeated here.

[0206] The various embodiments in this specification are described in a progressive manner, with each embodiment focusing on its differences from other embodiments. Similar or identical parts between embodiments can be referred to interchangeably. For the apparatus disclosed in the embodiments, since it corresponds to the method disclosed in the embodiments, the description is relatively simple; relevant parts can be referred to in the method section.

[0207] Those skilled in the art will further recognize that the units and algorithm steps of the various examples described in conjunction with the embodiments disclosed herein can be implemented in electronic hardware, computer software, or a combination of both. To clearly illustrate the interchangeability of hardware and software, the components and steps of the various examples have been generally described in terms of functionality in the foregoing description. 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 application.

[0208] The steps of the methods or algorithms described in conjunction with the embodiments disclosed herein can be implemented directly by hardware, a software module executed by a processor, or a combination of both. The software module can be located in random access memory (RAM), main memory, read-only memory (ROM), electrically programmable ROM, electrically erasable programmable ROM, registers, hard disk, removable disk, CD-ROM, or any other form of storage medium known in the art.

[0209] Finally, it should be noted that in this document, relational terms such as "first" and "second" are used only to distinguish one entity or operation from another, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Furthermore, the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such a process, method, article, or apparatus. Without further limitations, an element defined by the phrase "comprising one..." does not exclude the presence of other identical elements in the process, method, article, or apparatus that includes said element.

[0210] The technical solutions provided in this application have been described in detail above. Specific examples have been used to illustrate the principles and implementation methods of this application. The descriptions of the above embodiments are only for the purpose of helping to understand the methods and core ideas of this application. At the same time, for those skilled in the art, there will be changes in the specific implementation methods and application scope based on the ideas of this application. Therefore, the content of this specification should not be construed as a limitation of this application.

Claims

1. A multi-energy digital twin collaborative scheduling method, characterized in that, Applications in industrial energy management systems include: Data preprocessing is performed based on equipment operation data from industrial equipment and local edge computing nodes to determine the preprocessing results; the edge computing nodes include edge intelligent computing models. Based on the current digital twin model in the digital twin layer of the industrial energy management system and the preprocessing results, model calibration is performed to determine the calibrated twin model; the current digital twin model includes device-level and system-level digital twin models; By using the cloud collaboration layer in the industrial energy management system and the relevant data from the calibrated twin model, a collaborative energy scheduling analysis involving multiple devices and multiple energy sources is performed to determine the global energy scheduling strategy. The global energy scheduling strategy is verified based on the blockchain verification layer and cross-chain verification protocol in the industrial energy management system to determine the strategy verification result; the blockchain verification layer adopts a dual-chain architecture that separates private chain and consortium chain. When the strategy verification result shows that the verification is successful, the global energy scheduling strategy is sent to the corresponding edge computing node and the strategy execution operation is triggered to complete the collaborative scheduling operation and obtain the collaborative scheduling result.

2. The multi-energy digital twin collaborative scheduling method according to claim 1, characterized in that, The data preprocessing is performed based on the equipment operation data of the industrial equipment and the local edge computing nodes of the equipment to determine the preprocessing results; The edge computing node includes an edge intelligent computing model, comprising: Data is collected based on heterogeneous sensors deployed on industrial equipment to determine equipment operating data; Based on the local edge computing node of the industrial equipment, Kalman filtering is performed on the equipment operation data to complete the data noise reduction operation and obtain the noise-reduced data. Based on the device-level energy consumption prediction model in the edge intelligent computing model, and combined with the noise-reduced data, local energy consumption is lightly predicted to determine the local energy scheduling strategy corresponding to the industrial equipment; the device-level energy consumption prediction model is a model built based on a pruned neural network. The preprocessing result is determined based on the noise-reduced data and the local energy scheduling strategy.

3. The multi-energy digital twin collaborative scheduling method according to claim 2, characterized in that, The step of calibrating the model based on the current digital twin model in the digital twin layer of the industrial energy management system and the preprocessing results to determine the calibrated twin model includes: The preprocessed results are received by the digital twin modeling layer in the industrial energy management system according to a preset communication protocol. Energy consumption prediction is performed based on the current digital twin model in the digital twin modeling layer and the denoised data to determine the energy consumption prediction result; wherein, the current digital twin model includes device-level and system-level digital twin models; the device-level digital twin model is an energy consumption mapping function established by synchronizing the energy consumption status of the industrial equipment in real time; the system-level digital twin model is an energy consumption prediction model coupled with electrical energy, thermal energy and cold energy established by integrating data from multiple industrial equipment. Determine whether the difference between the energy consumption prediction result and the energy consumption prediction value in the local energy dispatch strategy is greater than a preset dynamic threshold to determine the difference judgment result; If the difference judgment result is yes, then the current digital twin model is updated using Bayesian method based on the denoised data to determine the calibrated twin model; The model parameters of the calibrated twin model are sent to the corresponding edge computing nodes, and a model parameter overwrite operation is triggered.

4. The multi-energy digital twin collaborative scheduling method according to claim 1, characterized in that, The process involves using the cloud-based collaborative layer in the industrial energy management system and relevant data from the calibrated twin model to perform collaborative energy scheduling analysis involving multiple devices and energy sources, in order to determine a global energy scheduling strategy. This includes: The cloud collaboration layer in the industrial energy management system receives the equipment health index generated by the digital twin layer and assigns a weight value to each piece of industrial equipment based on the equipment health index. The cloud collaboration layer, the calibrated twin model, the weight values, the distributed optimization algorithm, and the deep learning algorithm are used to perform collaborative energy scheduling analysis involving multiple devices and multiple energy sources in order to determine the global energy scheduling strategy.

5. The multi-energy digital twin collaborative scheduling method according to claim 2 or 3, characterized in that, After acquiring data based on heterogeneous sensors deployed on industrial equipment to determine equipment operating data, the method further includes: The hash value is determined based on a preset hash algorithm and the device's operating data. The hash value and the device signature information corresponding to the industrial equipment are uploaded to the private blockchain; Based on the private chain and the preset improved PBFT consensus mechanism, the received hash value and device signature information are verified, and when the verification is successful, the hash value and device signature information are written into the corresponding block.

6. The multi-energy digital twin collaborative scheduling method according to claim 5, characterized in that, The verification of the global energy scheduling strategy based on the blockchain verification layer and cross-chain verification protocol in the industrial energy management system includes: Based on the aforementioned private blockchain and the pre-defined improved PBFT consensus mechanism, and in conjunction with smart contracts, the temperature constraints and grid capacity constraints of the scheduling instructions in the global energy scheduling strategy are verified to determine the strategy verification results.

7. The multi-energy digital twin collaborative scheduling method according to claim 6, characterized in that, After determining the strategy verification result, the following is also included: If the strategy verification result shows that the verification is successful, then during the execution of the global energy scheduling strategy by the edge computing node, the collected device operation data and the executed scheduling instructions are verified and stored on the private chain based on the blockchain verification layer and the dual evidence storage strategy to determine the evidence storage information. Based on the consortium blockchain, a pre-compiled contract is invoked to verify the evidence storage information, and when the verification is successful, the consortium blockchain evidence storage operation is triggered. After completing the notarization operations for several blocks in the consortium blockchain, the final notarization is completed by calling the smart contract corresponding to the main chain in the blockchain verification layer.

8. A multi-energy digital twin collaborative scheduling device, characterized in that, Applications in industrial energy management systems include: The data preprocessing module is used to preprocess data based on the equipment operation data of industrial equipment and the local edge computing nodes of the equipment to determine the preprocessing results; the edge computing nodes include edge intelligent computing models. The model calibration module is used to calibrate the model based on the current digital twin model in the digital twin layer of the industrial energy management system and the preprocessing results, so as to determine the calibrated twin model; the current digital twin model includes device-level and system-level digital twin models; The scheduling analysis module is used to perform collaborative energy scheduling analysis of multiple devices and multiple energy sources through the cloud collaboration layer in the industrial energy management system and the relevant data of the calibrated twin model, so as to determine the global energy scheduling strategy. The strategy verification module is used to verify the global energy scheduling strategy based on the blockchain verification layer and cross-chain verification protocol in the industrial energy management system, so as to determine the strategy verification result; the blockchain verification layer adopts a dual-chain architecture that separates private chain and consortium chain. The strategy execution module is used to send the global energy scheduling strategy to the corresponding edge computing node when the strategy verification result shows that the verification is successful, and to trigger the strategy execution operation to complete the collaborative scheduling operation and obtain the collaborative scheduling result.

9. An electronic device, characterized in that, include: Memory, used to store computer programs; A processor for executing the computer program to implement the multi-energy digital twin collaborative scheduling method as described in any one of claims 1 to 7.

10. A computer-readable storage medium, characterized in that, Used to store a computer program, which, when executed by a processor, implements the multi-energy digital twin collaborative scheduling method as described in any one of claims 1 to 7.

Images