Block chain and brain-like computing collaborative virtual power plant transaction guarantee system and method

Abstract

The invention provides a block chain and brain-like computing collaborative virtual power plant transaction guarantee system and method. According to the system, a four-layer progressive architecture is innovatively constructed; the physical layer comprises a plurality of distributed energy nodes, and each node is embedded into an edge agent unit and is used for collecting real-time operation state data, converting continuous analog quantity into discrete pulse signals, uploading the discrete pulse signals to the neural layer and receiving and executing an upper-layer scheduling instruction; the neural layer adopts an event-driven mechanism to perform inter-node communication, triggers signal transmission when state variable change exceeds a threshold value, and dynamically adjusts connection topology and weight according to node reputation and physical distance to realize edge intelligence and local collaboration; the decision-making layer receives the pulse feature convergence information, and executes hierarchical scheduling calculation at multiple levels to generate a verified scheduling instruction and issues the scheduling instruction step by step; and the protocol layer maintains a double-chain storage structure of an event chain and a value chain, verifies an access identity, a scheduling instruction and settlement data by using an intelligent contract, and protects a data transmission link in combination with anti-quantum encryption.

Description

Technical Field

[0001] This invention relates to the field of computers, and specifically to a virtual power plant transaction security system and method that combines blockchain and neuromorphic computing. Background Technology

[0002] In recent years, the global energy system has been undergoing a profound structural transformation. With the rapid development of renewable energy, distributed generation, and energy storage technologies, the operation of energy systems is characterized by high distribution, dynamism, and intelligence. Virtual power plants, as a new energy management and dispatching model, aggregate dispersed, heterogeneous, and multi-type distributed energy resources into a dispatchable and optimizable whole through advanced communication, control, and information processing technologies, achieving unified coordination and efficient operation. This model demonstrates enormous potential in improving renewable energy utilization, participating in market transactions, and providing ancillary services, and is gradually becoming an important component of the energy internet.

[0003] However, existing virtual power plant systems still face numerous systemic bottlenecks in the context of large-scale renewable energy integration and multi-party collaborative operation. Firstly, regarding data reliability, traditional virtual power plants typically rely on centralized data collection and reporting mechanisms, requiring distributed energy terminals to upload operational data to a central server for unified processing. In this model, data authenticity depends entirely on the credibility of the central node, lacking cross-party anti-tampering and verification mechanisms, making it prone to node false reporting, delayed reporting, or falsified operational status. Data distortion not only affects the accuracy of scheduling and control but may also trigger trust crises and transaction disputes in multi-party markets.

[0004] Secondly, regarding real-time performance, the volatility of renewable energy output requires scheduling systems to have response capabilities at the second or even millisecond level. However, centralized optimization models experience a sharp increase in computational complexity as the number of nodes grows, with central servers often requiring several minutes or even longer to generate scheduling schemes, which can no longer meet the dynamic response requirements of highly penetrating renewable energy scenarios. Simultaneously, the latency of ordinary blockchain consensus mechanisms is also difficult to match with real-time control, making it difficult to reduce the overall system scheduling latency to a usable range for engineering applications.

[0005] Furthermore, in terms of collaborative efficiency and scalability, traditional virtual power plant architectures typically operate in a "centralized aggregation—unified computing" model. When the number of nodes reaches tens of thousands, the global optimization computational load grows non-linearly, and the data transmission and network communication load increases dramatically, easily causing communication congestion and scheduling delays. Due to the lack of hierarchical optimization and edge autonomy mechanisms, it is difficult to achieve rapid local optimization and global coordination and unification among distributed resources of different geographical regions and types, significantly limiting the system's flexibility and response speed.

[0006] In terms of security, the wide and diverse geographical distribution of distributed energy nodes makes them highly vulnerable to cyberattacks. Once subjected to data injection, spoofing, or denial-of-service attacks, it can lead to local or even global control failures. Current security measures are primarily based on traditional public-key encryption systems, lacking mechanisms to address the security threats posed by future quantum computing. Furthermore, the system lacks layered proactive defense and collaborative immunity mechanisms; once a single point of failure is compromised, it can easily trigger a chain reaction, creating system-wide security vulnerabilities.

[0007] More importantly, existing research focuses primarily on improving algorithms or specific functional modules, such as economic dispatch, load forecasting, or aggregated trading models. However, a unified system architecture framework has not yet been established, lacking top-level design from a systems engineering perspective. Traditional systems often feature "advanced algorithms but outdated architectures." When facing complex environments and multi-party interactions in the energy internet scenario, they are unable to achieve cross-level collaborative optimization and end-to-end reliable operation, making it difficult to support future-oriented evolution and upgrades.

[0008] In conclusion, the development of virtual power plants is no longer limited to algorithmic or localized optimization problems, but urgently requires a solution that involves systemic innovation at the architectural level. A novel system architecture integrating blockchain's trust mechanism and neuromorphic computing's collaborative mechanism must be constructed to achieve trusted access to multiple energy sources, rapid response from edge intelligence, inter-regional collaborative optimization, and multi-layered security protection. This would fundamentally solve the systemic bottlenecks of existing virtual power plants in terms of trustworthiness, real-time performance, collaboration, and security, providing architectural-level support for the highly reliable and intelligent operation of the energy internet. Summary of the Invention

[0009] To address the problems faced by existing virtual power plants in large-scale distributed energy access scenarios, such as low data reliability, high scheduling response latency, insufficient coordination efficiency, and lack of resistance to quantum attacks and proactive security protection capabilities, this invention proposes a virtual power plant transaction security system and method that combines blockchain and neuromorphic computing.

[0010] Firstly, a virtual power plant transaction security system that integrates blockchain and neuromorphic computing is provided, including: The physical layer contains several distributed energy nodes, each of which embeds an edge agent unit. The physical layer is used to collect real-time operating status data of the device, convert the real-time operating status data into discrete pulse signals and upload them to the neural layer. It also serves as the system's bottom-level perception and execution end to receive and execute scheduling instructions from the upper layer. The neural layer, built on top of the physical layer through the edge proxy unit based on a spiking neural network, is used for inter-node communication using an event-driven mechanism. It triggers signal transmission only when the state variable changes exceed a preset threshold, and dynamically adjusts the connection topology and weights between nodes based on node reputation and physical distance. The decision layer, located above and connected to the neural layer, receives feature aggregation information uploaded by the neural layer, performs hierarchical scheduling calculations at multiple levels, generates verified scheduling instructions, and sends them down to the physical layer level by level through the neural layer. The hierarchical scheduling calculations include edge real-time optimization, regional collaborative optimization, and central global optimization. The feature aggregation information is generated by feature extraction and summarization processing of pulse signals through a spiking neural network. The protocol layer, located above the decision layer and connected to both the neural layer and the decision layer, is used to maintain a dual-chain storage structure containing an event chain and a value chain. It uses smart contracts to verify access identities, scheduling instructions, and settlement data, and combines quantum-resistant encryption technology to encrypt and protect the data transmission links between layers.

[0011] Secondly, a method for ensuring virtual power plant transactions through the collaboration of blockchain and neuromorphic computing is provided, including: Collect the digital identity information and hardware physical characteristics of the node applying for access, generate a unique access credential through hash binding, and construct a trusted physical node set based on the blockchain consortium chain to verify the unique access credential; The nodes in the trusted physical node set are networked and connected based on the spiking neural network algorithm, and the network topology is dynamically reconstructed according to the node reputation and spatiotemporal attributes. The communication between nodes adopts an event-driven mechanism, which triggers the generation and transmission of pulse signals only when the state variable changes exceed a preset threshold. Based on the characteristic information carried by the pulse signal, hierarchical scheduling calculation is performed to generate verified scheduling instructions and issue them for execution. The hierarchical scheduling calculation includes scheduling calculation from immediate response at the edge side and collaborative optimization at the regional side to global overall planning at the central side. The dual-chain architecture is used to classify and store event data and value data in the scheduling and computing process, execute multi-dimensional value settlement based on smart contracts, and combine quantum-resistant encryption technology to encrypt the data throughout the process.

[0012] In another aspect, this application also provides an electronic device, comprising: at least one processor and a memory; the memory and the processor are connected via a bus; The memory is used to store one or more programs; When the one or more programs are executed by the at least one processor, a method for securing virtual power plant transactions through blockchain and neuromorphic computing collaboration, as described above, is implemented.

[0013] Furthermore, this application also provides a computer-readable storage medium having a computer program stored thereon, which, when executed, implements the aforementioned method for securing virtual power plant transactions through collaboration between blockchain and neuromorphic computing.

[0014] Compared with the prior art, the beneficial effects of the present invention are as follows: This invention provides a virtual power plant transaction security system and method that integrates blockchain and neuromorphic computing. The system includes a physical layer comprising several distributed energy nodes, each embedding an edge agent unit. This physical layer collects real-time operating status data of the devices, converts the data into discrete pulse signals, uploads them to the neural layer, and serves as the system's bottom-level perception and execution end to receive and execute scheduling instructions from the upper layer. The neural layer, built upon the physical layer using edge agent units based on a spiking neural network, collects event-driven mechanisms for inter-node communication, and ensures that state variables change when they exceed a preset threshold. The system triggers signal transmission and dynamically adjusts the connection topology and weights between nodes based on node reputation and physical distance. The decision layer, located above and connected to the neural layer, receives feature aggregation information uploaded by the neural layer, performs hierarchical scheduling calculations at multiple levels, generates verified scheduling instructions, and distributes them step-by-step to the physical layer through the neural layer. The protocol layer, located above and connected to both the neural and decision layers, maintains a dual-chain storage structure containing event chains and value chains. It uses smart contracts to verify access identities, scheduling instructions, and settlement data, and employs quantum-resistant encryption technology to encrypt and protect the data transmission links between layers.

[0015] Specifically, this invention achieves highly reliable, robust, and intelligent operation of a virtual power plant in a complex energy internet environment through the design of brain-like collaborative scheduling and multi-layered security protection mechanisms. The system, based on brain-like computing principles, organically combines event-driven neural dynamics, distributed optimization, and on-chain trust mechanisms to form a self-perceiving, self-learning, self-optimizing, and self-defending intelligent scheduling system, achieving significant improvements in performance, stability, and security.

[0016] In terms of scheduling performance, this invention adopts a hierarchical, progressive, brain-like collaborative optimization architecture. Edge nodes can complete local state prediction and control response in milliseconds, the regional coordination layer achieves rapid regional convergence and collaborative correction through the Distributed Augmented Lagrange Multiplication Model (ADMM), and the central layer completes global consistency optimization within seconds. Simulation results show that compared with the traditional centralized optimization framework, the overall system response speed is improved by about 70%, energy allocation deviation is reduced by about 40%, and near-linear scalability is maintained even when the node scale reaches tens of thousands.

[0017] In terms of communication efficiency, the system is based on a pulse event-driven model, triggering computation and communication only when state changes exceed a threshold, thus avoiding bandwidth waste caused by periodic polling. Test results show that this mechanism can reduce the system's communication load by 60%–80%, while reducing the average latency to 1 / 5 of the original system while ensuring data accuracy, achieving real-time and efficient transmission of scheduling and control signals.

[0018] In terms of security protection, the multi-layered protection system of this invention integrates three mechanisms: anomaly detection, reputation evaluation, and immune response. The edge layer uses a pulse autoencoder to achieve early identification of device anomalies; the region layer uses Bayesian fusion to achieve reliable judgment and local isolation of abnormal events; and the central layer uses an on-chain knowledge graph to achieve cross-domain threat correlation analysis. Experimental results show that, under the conditions of multi-point data tampering and denial-of-service (DoS) attacks, the average recovery time of the system is reduced by more than 50% compared with traditional systems, and it can recover to a safe steady state within seconds.

[0019] Regarding information trustworthiness, the system introduces quantum-resistant lattice-based cryptographic algorithms (Dilithium signature, KyberKEM encryption) and combines them with distributed key generation (DKG) and threshold signature mechanisms to achieve tamper-proof and verifiable scheduling instructions and data transmission throughout the entire process. All optimization results and security response processes are solidified by blockchain smart contracts, and any node can perform signature verification and timing checks, ensuring the authenticity and traceability of the system's operating results.

[0020] In terms of system robustness and adaptability, the dynamic plasticity of neuromorphic networks allows the connection weights between nodes to be adaptively adjusted according to the operating state. Faced with communication interruptions, node failures, or sudden load changes, the system can maintain functional continuity through topology reconstruction and local autonomy, exhibiting a "self-healing" characteristic similar to biological neural networks. Simulations show that when 10% of the nodes fail, the system can still maintain more than 90% of its operating performance indicators and can restore its overall coordination capability within tens of seconds.

[0021] In summary, compared with the traditional virtual power plant dispatching framework, the present invention has the following significant effects and advantages: 1. High real-time performance – Event-driven architecture reduces scheduling and protection response time to milliseconds to seconds; 2. High scalability – The distributed, optimized structure supports linear expansion of large-scale nodes; 3. Secure and reliable – A multi-layered protection system and on-chain encryption mechanism ensure that information cannot be forged and the process is traceable; 4. Intelligent Adaptation – Brain-like Plastic Networks Enable Continuous Learning and Self-Evolution; 5. Robust and stable – The system can maintain stable operation even under abnormal and attack scenarios.

[0022] Therefore, this invention not only innovatively integrates brain-like computing and blockchain trust mechanisms into the energy dispatch system in theory, but also significantly improves the operating efficiency, security protection capabilities and system reliability of virtual power plants in engineering practice, and has broad application prospects and promotion value. Attached Figure Description

[0023] Figure 1 This is a schematic diagram of the overall architecture of the virtual power plant transaction guarantee system that combines blockchain and neuromorphic computing according to the present invention. Figure 2 This is a schematic diagram of the overall brain-like collaborative scheduling structure of the virtual power plant transaction guarantee system that combines blockchain and brain-like computing according to the present invention. Figure 3 This is a schematic diagram of an implementation case of the virtual power plant transaction guarantee system that combines blockchain and neuromorphic computing according to the present invention. Figure 4 The flowchart shows the virtual power plant transaction security method of blockchain and neuromorphic computing collaboration of the present invention. Figure 5 This is a schematic diagram of an electronic device structure according to the present invention. Detailed Implementation

[0024] This invention aims to fundamentally address the systemic problems of existing virtual power plant systems, such as insufficient trustworthiness, limited optimization efficiency, weak security protection, and inadequate collaboration capabilities. It constructs a novel virtual power plant trust assurance system that integrates blockchain and neuromorphic computing, oriented towards multi-source heterogeneous energy resources. This system does not primarily focus on algorithm optimization; instead, it achieves functional collaboration, architectural self-adaptation, and end-to-end trustworthy operation through architectural innovation, thereby providing a unified system platform for the intelligent and secure operation of the future energy internet.

[0025] In terms of overall design goals, this invention aims to establish a multi-layered architecture system with blockchain as the foundation of trust and neuromorphic computing as the core of collaboration. The system is characterized by trustworthiness, efficiency, security, and self-evolution, achieving trusted access to distributed energy resources, real-time information interaction, edge self-optimization, global collaborative control, and multi-layered proactive protection. Through blockchain's distributed identity authentication, data traceability, and consensus mechanisms, it ensures that data from various energy nodes is verifiable, behavior is traceable, and records are tamper-proof. Through the event-driven model and pulse information transmission mechanism of neuromorphic computing, it achieves rapid response, adaptive connection, and hierarchical collaboration between nodes, thereby completing dynamic optimization and decision execution within millisecond to second timescales.

[0026] At the system level, this invention proposes a four-layer progressive architecture, including a physical layer, a neural layer, a decision layer, and a protocol layer. The physical layer is responsible for real-time data acquisition and action execution from energy devices and sensors, forming the foundation for the system's perception and execution. The neural layer constructs a distributed adaptive network through an event-driven impulse computation model, achieving edge intelligence and local collaboration. The decision layer undertakes global coordination and hierarchical optimization tasks, realizing multi-scale energy scheduling and strategy reasoning through a regional-central two-level collaborative mechanism. The protocol layer, with blockchain as its core, implements identity verification, data on-chaining, reputation management, and smart contract execution, ensuring the system's reliable operation in cross-participant, multi-task scenarios. This four-layer architecture forms a closed-loop system of perception, cognition, decision-making, and trust from bottom to top, enabling the system to possess a complete operational chain from data acquisition to global control.

[0027] In terms of security, this invention ensures the long-term reliability of the system in future computing environments by introducing multi-layered security mechanisms and quantum-resistant algorithms. The system employs post-quantum cryptography algorithms at the communication link and data signature levels, combined with distributed key management and threshold signature mechanisms, effectively preventing data leakage and instruction forgery. At the operational level, an anomaly detection and immune response module is embedded in the neuromorphic network, enabling adaptive identification and local isolation of malicious data injection, spoofing interference, and sensor anomalies, forming a self-healing security defense system.

[0028] Through the aforementioned system design and collaborative mechanism, this invention achieves breakthroughs in three aspects of virtual power plants: structural innovation, functional integration, and security and reliability. The system not only overcomes the bottlenecks of traditional centralized control platforms in terms of scalability and real-time performance, but also overcomes the weaknesses of existing virtual power plants in data trust and security protection. With "structural self-adaptation, autonomous decision-making, and reliable collaboration" as its core characteristics, this system constructs a verifiable, scalable, and evolvable new energy internet operating framework, providing a practical system foundation for aggregated scheduling and safe operation in environments with a high proportion of distributed energy access.

[0029] To better understand the present invention, the following description, in conjunction with the accompanying drawings and embodiments, will further illustrate the content of the present invention.

[0030] Example 1: The virtual power plant trusted assurance system proposed in this invention is a novel architecture. It is not a simple superposition of existing virtual power plant scheduling models with technologies such as blockchain and neuromorphic computing. Instead, it starts from the essential needs of systems engineering and the energy internet, constructing a highly reliable, highly collaborative, and adaptive multi-layered system architecture for multi-source distributed energy. From the underlying physical energy to the high-level trust protocols, this system achieves the organic integration of information flow, energy flow, and control flow, forming a trusted operational closed loop throughout the entire lifecycle. Its core objective is to achieve highly reliable aggregation and control of massive heterogeneous resources in complex and dynamic energy environments with minimal latency and computational resource consumption, while ensuring the verifiability, security, and autonomy of each layer of the system.

[0031] Specifically, the system adopts a four-layer progressive architecture, including the physical layer, neural layer, decision layer, and protocol layer, such as... Figure 1 As shown, the four levels are logically progressive from bottom to top, mutually coordinated, and form a closed-loop mechanism of two-way interaction between the upper and lower levels during operation.

[0032] In this embodiment, the physical layer consists of various distributed energy terminals and intelligent sensing devices, including photovoltaic power generation units, wind power generation devices, energy storage equipment, controllable loads, and electric vehicle charging and discharging infrastructure. Each physical node is equipped with an agent unit with edge computing capabilities. This unit can be an embedded computing module or a small industrial computer with a neuromorphic chip, responsible for state data acquisition, encoding, and local preprocessing. It also possesses a certain degree of autonomous control capability, enabling efficient information uplink and rapid command downlink. The physical layer is not only the entry point for interaction between the system and energy entities but also the foundation for the brain-like network to acquire environmental information. Therefore, its design imposes strict requirements on data accuracy, time synchronization, and sensor calibration to ensure that the upper-layer optimization modules receive real-time, reliable, and verifiable data support.

[0033] In this embodiment, the neural layer, located above the physical layer, draws on the dynamic characteristics and event-driven principles of biological neural networks to construct an adaptive communication and collaboration mechanism between nodes using a brain-like information processing model. The core idea of ​​this layer is to achieve information interaction and collaborative optimization between distributed energy nodes by simulating the dynamic connections, impulse transmission, and plasticity learning of neurons. Instead of using a traditional fixed topology, the neural layer allows the connection relationships and communication frequencies between nodes to be dynamically adjusted based on reputation status, task relevance, and physical proximity, forming a real-time reconfigurable adaptive topology network. This structure enables the network to automatically reduce the connection weights of nodes or isolate risky nodes when their states are abnormal, thereby significantly improving the robustness and security of the system.

[0034] Furthermore, each node in the neural layer runs a lightweight brain-like model (such as a spiking neural network (SNN) or an event-driven sparse neural network), which performs feature extraction and preliminary decision-making by receiving data pulses from the physical layer, and exchanges information with neighboring nodes in the form of event triggers, achieving sparse and efficient distributed collaborative computing. This mechanism effectively reduces the network communication burden, enabling the system to complete optimization and response on a millisecond timescale.

[0035] In this embodiment, the decision layer above the neural layer undertakes the task of global coordination and optimization decision-making for the system. Structurally, this layer is divided into three levels: edge real-time optimization, regional collaborative optimization, and central global optimization, to achieve dynamic collaborative control at different spatial scales.

[0036] In one specific embodiment, edge real-time optimization leverages the event-driven capabilities of neural layer nodes to process local state and prediction data in real time, enabling rapid power regulation and adaptive load adjustment. Regional collaborative optimization is achieved jointly by clusters of geographically or electrically proximate nodes, exchanging key feature information through a pulse event network and running a graph neural network-based regional optimization model to achieve resource coordination within a local area. Central global optimization integrates regional optimization results with grid operation information, runs a hybrid deep prediction and mathematical programming solver, and generates a global scheduling and control scheme. All decision commands are solidified and verified via a blockchain network to ensure the credibility of command sources, consistency of execution, and traceability of results.

[0037] In this embodiment, the top layer of the system is the protocol layer, which serves as the foundation for trust and contract execution throughout the entire system. This layer uses blockchain as its core, implementing functions such as identity authentication, data on-chaining, reputation management, and smart contract execution. To meet real-time and security requirements, the protocol layer adopts a dual-chain structure: an event chain for high-frequency recording of scheduling events, resource availability, and state changes; and a value chain for low-frequency but highly secure trusted operations and recording. The two chains are interconnected through encrypted indexes and event identifiers, forming a verifiable data mapping relationship. The smart contracts in this layer incorporate a dynamic reputation scoring mechanism, which automatically updates node reputation based on execution timeliness and response accuracy, ensuring consistency between long-term behavioral incentives for participants and the system's security boundaries. The protocol layer also deploys post-quantum-safe algorithms and a hot-swap mechanism in the critical signature stage to guarantee the long-term security of the system under future computational threats.

[0038] From an overall structural perspective, the system forms a two-way closed loop, both bottom-up and top-down: Bottom-up, energy and state data from the physical layer undergo event-driven processing and regional aggregation at the neural layer, are abstracted level by level into optimizable features, and uploaded to the decision layer to form a global optimization result; top-down, global instructions are verified by the protocol layer and constrained by contracts, then decomposed layer by layer before finally being transmitted to specific execution nodes. The entire process achieves dynamic response and real-time closed-loop control through an event-triggered mechanism.

[0039] The system's design fully embodies the deep integration of blockchain trust mechanisms and brain-like computing collaboration mechanisms. The former provides a unified trust foundation and data immutability among multiple stakeholders, while the latter greatly enhances the system's distributed computing and collaborative decision-making capabilities. The combination of these two elements forms a verifiable and self-evolving intelligent energy system architecture, significantly improving the virtual power plant's real-time performance, security, robustness, and scalability. This system not only provides architectural support for aggregated dispatching under conditions of high renewable energy penetration but also lays a sustainable technological foundation for the intelligent operation of the future energy internet.

[0040] Example 2: A virtual power plant transaction security system that integrates blockchain and neuromorphic computing, such as Figure 1 As shown, it includes: The physical layer contains several distributed energy nodes, each of which embeds an edge agent unit. The physical layer is used to collect real-time operating status data of the device, convert the real-time operating status data into discrete pulse signals and upload them to the neural layer. It also serves as the system's bottom-level perception and execution end to receive and execute scheduling instructions from the upper layer. The neural layer, built on top of the physical layer through the edge proxy unit based on a spiking neural network, is used for inter-node communication using an event-driven mechanism. It triggers signal transmission only when the state variable changes exceed a preset threshold, and dynamically adjusts the connection topology and weights between nodes based on node reputation and physical distance. The decision layer, located above and connected to the neural layer, is used to receive feature aggregation information uploaded by the neural layer, perform hierarchical scheduling calculations at multiple levels, generate verified scheduling instructions, and send them down to the physical layer level by level through the neural layer. The protocol layer, located above the decision layer and connected to both the neural layer and the decision layer, is used to maintain a dual-chain storage structure containing an event chain and a value chain. It uses smart contracts to verify access identities, scheduling instructions, and settlement data, and combines quantum-resistant encryption technology to encrypt and protect the data transmission links between layers.

[0041] The hierarchical scheduling calculation includes edge real-time optimization, regional collaborative optimization, and central global optimization. The feature convergence information is generated by feature extraction and summarization processing of pulse signals through a spiking neural network.

[0042] The virtual power plant trusted assurance system of this invention is designed with the core principles of continuity and functional synergy across the entire system chain. It forms a verifiable, traceable, low-latency, and highly robust closed-loop system from physical data acquisition to optimization decision execution. The system's functional modules, based on logical connections and hierarchical responsibilities, constitute a system architecture composed of core units such as trusted access and identity authentication, dynamic reputation management, brain-like collaborative scheduling, security protection, and quantum-resistant encryption. The modules achieve efficient coupling and hierarchical autonomy through event triggering and on-chain verification.

[0043] In this embodiment, the system first implements trusted access and identity authentication functions, which form the foundation of the system's security boundary. All nodes must complete digital identity registration and verification through the consortium blockchain's certificate authority before accessing the network. The consortium blockchain generates a digital certificate based on the node's registration information and combines it with the node's hardware characteristics to form a neural fingerprint. These two fingerprints are hashed and bound together, then uploaded to the blockchain as the node's unique identifier, serving as its basic performance parameter vector, which specifically includes: The node access process in the physical layer includes: The node initiates a registration request to the consortium blockchain certificate authority at the protocol layer and receives a digital certificate generated by the certificate authority, which is generated from the node's registration information. A unique access credential is generated by binding the digital certificate with the hardware neural fingerprint hash formed by combining the hardware characteristics of the node. Based on the unique access credential, an access request containing a quantum-resistant signature is initiated to the protocol layer, and the node access is completed after verification by the protocol layer consortium blockchain verification node. Based on the digital certificate and the hardware neural fingerprint, a unique access credential is generated through hash binding, and an access request containing a quantum-resistant signature is sent to the protocol layer to achieve anti-forgery access for physical nodes.

[0044] In one specific embodiment, the access credential hash value is used. It can be represented as:

[0045] in, This indicates a serial operation. Indicates a unique node number. Represents a vector of basic performance parameters of a node. This represents a hardware neural fingerprint; it is only available when a node initiates an access request to the system with a signature attached. And verified by consortium blockchain validator nodes Access is only permitted at certain times. This design binds the unforgeable fingerprint of a physical node to its digital identity, preventing cloning, forgery, or unauthorized access. This mechanism ensures the uniqueness and authenticity of node access. A node can only enter the system network after carrying a signature and being confirmed by the consortium blockchain verification node, preventing forgery, cloning, or unauthorized access, thus establishing a trusted starting point for the system during the access phase.

[0046] After identity authentication is completed, the node enters the brain-like dynamic connection network, where its connection weight and communication frequency will be continuously adjusted according to the reputation management mechanism. The system's reputation model uses the node's data reliability, latency performance, and execution accuracy in historical operation as scoring criteria to define a node reputation score. In time The value is updated iteratively as new information arrives:

[0047] in, As a smoothing factor, For the updated node reputation score, This is the instantaneous quality score calculated for the current period, and the calculation method is as follows:

[0048] in, Indicates a score for data accuracy. This indicates the score for the accuracy of instruction execution. This represents the average response delay score.

[0049] Weight satisfy These correspond to data accuracy score, instruction execution accuracy score, and average response latency score, respectively. Connection weight. In credit score and network physical distance Dynamic adjustment under the combined effect:

[0050] in, This is the distance attenuation coefficient, ensuring that the network topology balances high reputation priority and proximity priority. For node i, a dynamic reputation score is given. A dynamic reputation score is assigned to node j; this dynamic connectivity plasticity allows the brain-like network topology to be continuously optimized as the network operates, reducing overall network communication latency.

[0051] At the data management level, this invention designs a hybrid on-chain / off-chain distributed storage mechanism. High-frequency but low-value real-time data is stored locally on nodes to form a neuromorphic cache network, while low-frequency but high-value metadata is stored on-chain. Assume the system generates the following dataset: Each data Importance label With access frequency The storage strategy optimization problem can be expressed as:

[0052] in Indicates on-chain storage. For on-chain storage costs, For off-chain storage costs, This is an importance threshold. To ensure the verifiability of off-chain data, all data stores its hash index. On-chain, signature verification and hash comparison ensure that the data has not been tampered with during any access.

[0053] The core computing module of the system is a brain-like collaborative scheduling engine. This engine achieves multi-level collaborative control from edge nodes to central nodes through a hierarchical optimization structure, specifically including: The edge real-time optimization unit, embedded in the edge agent unit, is used to process locally sensed pulse signals and prediction data based on the event-driven capabilities of neural layer nodes, in order to generate control actions with the goal of minimizing economic costs and physical deviations. The regional collaborative optimization unit is used to aggregate the impulse features of nodes within a region using a graph neural network, exchange key feature information through an impulse event network, and perform collaborative computation between adjacent node clusters using a distributed augmented Lagrange algorithm to output the coordinated optimization decision. The central global optimization unit integrates regional optimization results with power grid operation information, combines a hybrid deep prediction and mathematical programming solver for the whole network constrained operation, solves the global optimization model, and decomposes the optimization results into instructions for issuance to correct regional and edge strategies.

[0054] In one specific implementation, the underlying edge real-time optimization is primarily responsible for local rapid response and execution strategy adjustments, with the goal of minimizing the sum of economic costs and physical deviations:

[0055] in It is a node The adjustment decision vector, It is the first Decision variables of each node The corresponding economic cost function, It is the first The actual output of each node It is the first one issued by the upper management. The settings for each node, The penalty coefficient is... For the first The objective function for each node. Regional collaborative optimization then runs a graph neural network model within clusters of adjacent nodes, based on the adjacency matrix between nodes. With node feature matrix Output the optimized decision after coordination:

[0056] in, , yes The degree matrix, It is an activation function. For the first l The weight matrix of the layer, For optimized decision-making after coordination, For the first l Layer features, This is the augmented adjacency matrix of the original adjacency matrix A. The central global optimization phase integrates regional results with real-time market data to solve the constrained optimization problem:

[0057] in To meet the needs of the entire network, For the first The set of physical constraints for each node. For the set of global decision variables, The total system cost, For decision variables At that time, the first The cost of a distributed energy unit.

[0058] For system security, anomaly detection and immune response mechanisms are crucial, and these specifically include: The decision-making level also includes: The edge layer anomaly detection unit is used to calculate the error through a pulse autoencoder and determine the node anomaly when the error meets the preset anomaly conditions. The regional layer immune response unit is used to combine evidence based on multi-node anomaly reports collected by the regional coordinator through Bayesian inference, calculate the posterior probability after the anomaly occurs, and execute local immune response actions after the calculated posterior probability exceeds a preset probability threshold. The central layer situational awareness unit is used to build a security knowledge graph based on on-chain records and reputation indexes, and to identify potential attack paths and covert data injection behaviors.

[0059] In one specific embodiment, the present invention implements anomaly pattern recognition based on a pulse autoencoder at edge nodes, mapping time-series data to a pulse event domain and training the reconstruction error. ,when Time-based anomaly detection:

[0060] in, It is the original feature vector. It is the vector reconstructed by the autoencoder. It is an empirical threshold. For error. After an anomaly is detected, the regional coordinator collects anomaly reports from multiple nodes, merges the evidence through Bayesian inference, and calculates the posterior probability of the anomaly occurring:

[0061] in, For posterior probability, For the evidence set, For likelihood, For prior probability, This represents the marginal probability of the evidence. If the posterior probability exceeds a set threshold, a local immune response action is executed, such as isolating the node or switching to a safe mode, and the event and its handling measures are then recorded on the blockchain.

[0062] In this embodiment, the transaction and value settlement mechanism is built on a dual-chain architecture of event chain and value chain. The event chain stores execution proofs, and the value chain triggers settlement based on the proofs from the event chain. The dual-chain storage structure of this protocol layer specifically includes: The event chain is used to store high-frequency status pulse records, hashes of scheduling instructions generated by the decision layer, and abnormal alarm events; The value chain is used to store low-frequency fund settlement records and credit score changes; The event chain and value chain are linked through a cryptographic index and combined with smart contracts to implement a settlement mechanism.

[0063] In one specific embodiment, the settlement amount calculation formula in the settlement mechanism is as follows:

[0064] in, For the first The settlement value of each node. For the first Effective response count per node As the benchmark price, For the first Individual node reputation The weighting coefficients, For the first Node response time The attenuation coefficient, For the first Execution accuracy of each node The reward coefficient, For the first Each node has unique reward and penalty clauses. The calculation rules and formula hashes for all coefficients are recorded on the blockchain to ensure verifiability.

[0065] In the demand-side resource bidding module, this invention employs the Relatively Robust Conditional Value at Risk (RRCVaR) model. First, it defines a given confidence level... and probability distribution CVaR under:

[0066] in, Represents conditional risk value. This represents the mathematical expectation of a random variable y under a given probability distribution U. , For the plan In state The loss function is RRCVaR, which is defined as the relative risk relative to the optimal baseline solution.

[0067] in, Represents relative conditional risk value. In distribution The optimal decision is made across all possible distributions. Minimize the maximum relative risk:

[0068] in, Representing the feasible region, this model ensures that the selected combination of demand response participants is highly robust under various external uncertainties, thereby improving execution quality and economy.

[0069] Finally, to address the threats posed by future quantum computing, the security and quantum-resistant design module introduces quantum-resistant public-key cryptographic algorithms into the critical encryption and signature stages of the system, such as lattice-based Dilithium signatures and KyberKEM encryption, specifically including: The protocol layer includes a quantum-safe module for: In the physical layer access and decision-layer instruction issuance stages, a lattice-based digital signature algorithm is used for digital signatures. In the data transmission link establishment stage between each layer, a lattice-based key encapsulation mechanism is used to establish an encrypted channel, and distributed key generation technology is used to distribute and manage the system's private keys.

[0070] In one specific embodiment, let the complexities of signature generation and verification be respectively... and The algorithm selected in this invention satisfies the following at security level 5:

[0071] On current hardware platforms, signing and verification can be completed in milliseconds, ensuring that the system can achieve quantum-resistant security without sacrificing real-time performance. Furthermore, mechanisms such as distributed key generation (DKG) and threshold signatures free key management from single-point dependencies, further enhancing the system's long-term security resilience.

[0072] In summary, this invention constructs a fully trustworthy, adaptive, and highly secure virtual power plant system functional architecture for distributed energy scenarios by organically combining core modules such as trusted access, dynamic collaboration, hierarchical optimization, and multi-layered protection. The interaction and collaborative operation of each functional module enable the system to possess the capabilities of data trustworthiness, efficient optimization, robust security, and self-evolving structure, providing fundamental support for the intelligent and trustworthy operation of the energy internet.

[0073] Example 3: The brain-like collaborative scheduling and multi-layered security protection mechanism proposed in this invention is the core and soul of the entire virtual power plant trusted assurance system. Its design concept originates from the hierarchical organizational structure and energy adaptive regulation mode of biological nervous systems. By introducing the event-driven, dynamic plasticity, and distributed collaborative principles of neural networks into the energy scheduling system, the virtual power plant possesses intelligent characteristics similar to the biological central nervous system: namely, rapid perception, autonomous decision-making, efficient collaboration, and self-healing security. Architecturally, this mechanism achieves deep coupling between scheduling and protection, enabling millisecond-level monitoring and response to the energy system's operating status, and automatically forming an immune response in the event of anomalies or attacks, thereby constructing an energy scheduling and security collaborative system with "intelligent evolution" capabilities.

[0074] The brain-like collaborative scheduling system of this invention adopts a hierarchical and progressive design approach, comprising three sub-layers: an edge-based real-time optimization layer, a regional collaborative optimization layer, and a central global optimization layer. These layers interact bidirectionally through event trigger signals, pulse-coded information streams, and consensus decision results, forming a bottom-up data perception channel and a top-down control command channel. Physically, the layers are not linearly subordinate but are coupled through a reconfigurable event bus and dynamic connection topology, ensuring stable operation even under conditions of topology changes, node failures, or communication delays. Figure 2 A schematic diagram of the overall system architecture of the brain-like collaborative scheduling mechanism is presented, showing a multi-level interconnected architecture from the physical device layer and the brain-like network layer to the protocol trust layer. The various layers of the system are bound together through distributed consensus and trust indexes, ensuring that information transmission paths, decision logic, and execution feedback are traceable and verifiable in any scheduling event.

[0075] In terms of operation, the edge layer acts as the "nerve endings" of the system, responsible for the lowest-level state acquisition and local control. Each distributed energy node, such as a photovoltaic inverter, energy storage device, charging / discharging interface, or controllable load unit, embeds a lightweight, brain-like intelligent agent. This agent operates based on an event-driven mechanism, triggering computation and communication only when the node state undergoes a significant change (e.g., sudden output changes, frequency shifts, or communication delays exceeding a threshold), thus avoiding the resource waste caused by periodic sampling in traditional scheduling. Edge nodes not only possess sensing and execution capabilities during operation but also perform local prediction and optimization. When an abnormal state or power deviation is detected, the node recalculates the optimal response value within milliseconds based on its built-in local control function and transmits key summary information to the upper-level regional nodes via pulses. Because event-driven communication occurs only when necessary, the overall system communication load can be reduced by approximately 60%–80% compared to traditional models, while the response speed is increased by tens of times.

[0076] Within the edge layer, the control logic of each node is implemented by a neuromorphic spiking neural network (SNN). Its neuron model follows the leaky integration discharge (LIF) mechanism, where the membrane potential gradually accumulates under the influence of input current, generating a pulse signal and immediately resetting when a threshold is reached. This mechanism simulates the spiking activity characteristics of real neurons, allowing control decisions to be expressed as events rather than continuous signal streams, thus significantly reducing information bandwidth requirements. The synaptic weights of nodes are dynamically adjusted according to the temporal dependent plasticity (STDP) rule. When the system is in a certain operating mode for a long time, the connection strength between related nodes automatically strengthens, forming stable local response paths; while when the system environment changes, these connections gradually weaken or reconstruct, exhibiting a "learning" and "forgetting" ability similar to biological neural networks. This adaptive plasticity enables the virtual power plant to achieve self-regulation and performance optimization under changing external conditions, maintaining stable operation within a local range without relying on central commands.

[0077] The regional collaborative optimization layer is analogous to a "nerve cluster" in a brain-like system, responsible for integrating local information from several edge nodes to achieve coordinated optimization and anomaly protection within the region. Regional nodes are connected via pulse events, forming a sparse yet efficient information exchange topology. Each regional coordinator continuously receives state event streams uploaded by its subordinate nodes, performing feature extraction, cluster analysis, and similarity assessment to determine the overall operational status of the region. When a sub-region's state deviation exceeds a safety threshold, the coordinator initiates a distributed collaborative optimization algorithm to quickly generate an adjustment scheme by solving a simplified local constraint model, and reports the results to the central layer as encrypted events. This regional mechanism ensures the system possesses "local self-healing" capabilities: even if the central layer temporarily loses connection, the regional layers can maintain stable operation based on historical patterns and real-time information, and can implement islanded control when necessary.

[0078] To achieve efficient collaboration between regions, this invention introduces a Graph Neural Network (GNN) structure into the region optimization process. The energy flow relationship between node states and edges is modeled as a weighted graph, and the region coordinator achieves state estimation and optimization decision-making through convolutional aggregation and graph feature propagation. Compared with traditional linear prediction models, brain-like graph networks have significant nonlinear modeling capabilities and can maintain a fast convergence speed even with a very large number of nodes. By introducing the Distributed Augmented Lagrange Multiplication Model (ADMM), the system can achieve globally consistent optimization without fully sharing data, thus ensuring node data privacy and improving the parallelism and stability of optimization computation.

[0079] The regional layer not only undertakes collaborative optimization tasks but is also deeply coupled with the security protection system. Each regional coordinator has a built-in anomaly detection and immune response module. When a node experiences data or communication anomalies, the regional coordinator immediately downgrades its reputation status and notifies neighboring nodes to reduce their connection weights with it, preventing further spread of the anomaly. If multiple nodes report similar anomaly patterns, the coordinator uses a Bayesian fusion mechanism to calculate the posterior probability of the anomaly event. If the probability exceeds a set threshold, the system immediately implements a local isolation strategy, temporarily severing the communication link between the suspicious node and the backbone. This mechanism achieves simultaneous operation of security and optimization: while maintaining scheduling efficiency, the system can spontaneously identify potential security threats and achieve proactive defense.

[0080] At the central global optimization layer, the system undertakes global scheduling and strategy coordination tasks. The central optimization engine integrates aggregated information, predicted data, and network-wide operational constraints uploaded from the regional layers to form a comprehensive optimization model. This model considers factors such as power balance, energy storage scheduling, equipment capacity limitations, and line power flow constraints, aiming to minimize the overall system operating cost while ensuring coordination and consistency among regions. To improve real-time performance, the central optimization process employs a hierarchical solution and model pruning strategy, with pre-optimization performed by the regional layers followed by joint correction by the central layer. Verification has shown that even with tens of thousands of nodes, the system can still complete a global optimization iteration within seconds. Optimization results are solidified and signed via blockchain smart contracts, ensuring the verifiability and tamper-proof nature of scheduling instructions. Instructions are accompanied by digital signatures and encrypted indexes during issuance, ensuring that all data received by executing nodes undergoes consistency verification, preventing erroneous control behavior due to malicious tampering or communication errors.

[0081] In addition to scheduling functions, the brain-like collaborative mechanism of this invention also includes a multi-layered security protection system to resist complex threats such as network attacks, data forgery, and equipment failures. The security protection system is designed to exist in parallel with and support the scheduling mechanism, forming a closed-loop operation chain of "protection-response-immunity-recovery." The system protection is divided into three layers: an edge defense layer, a regional immunity layer, and a central situational awareness layer. At the edge layer, each node's neural agent embeds a pulse autoencoder model. By learning the time-series characteristics under normal operating conditions, it reconstructs the input data. When the reconstruction error exceeds an empirical threshold, the node is determined to be in an abnormal state and an alarm is triggered. The node then enters a safe mode, limiting its output power or temporarily cutting off communication to prevent the propagation of erroneous data. At the regional layer, the coordinator aggregates anomaly reports from multiple nodes and performs correlation analysis on events based on probabilistic reasoning methods to determine whether the anomalies are concentrated or correlated. If an anomaly pattern is identified as a cross-node attack, the system immediately triggers a regional immunity response mechanism, isolating the relevant nodes and adjusting the connection weight matrix, so that the abnormal nodes are gradually eliminated spontaneously by the network structure. The central layer monitors the system's security status over the long term through a global situational awareness module. This module constructs a security knowledge graph based on on-chain records and reputation indexes, enabling it to identify potential attack paths and covert data injection behaviors, and, when necessary, issue security policy updates to the regional layers, achieving cross-layer defense coordination.

[0082] At the information security level, this invention introduces a quantum-resistant encryption and signature system to address the threat posed by future quantum computing technology to traditional public-key cryptosystems. System communication and critical signature processes are all implemented using lattice-based cryptography, such as Dilithium signatures and the KyberKEM encryption algorithm. These algorithms can guarantee signature and verification within milliseconds at security level 5, achieving long-term cryptographic strength without sacrificing real-time performance. Furthermore, the system employs a distributed key generation and threshold signature mechanism, preventing any single node from independently controlling the system's private key, fundamentally eliminating the security vulnerabilities associated with centralized key storage. The key management process is bound to a blockchain smart contract, allowing for traceability and auditability of any key update and verification process, thus achieving full lifecycle cryptographic security management.

[0083] The brain-like collaborative scheduling and multi-layered security protection mechanism of this invention forms a complete closed-loop feedback chain during operation. Bottom-up, sensory data from the physical layer enters the feature extraction and optimization network of the neural layer in the form of pulse event streams; top-down, scheduling instructions generated by the central layer are verified by smart contracts and then distributed layer by layer until they are executed by edge nodes; horizontally, the regional layers form a dynamic balance through pulse event interaction and a reputation network, enabling the entire system to have adaptive coordination capabilities in both time and space dimensions. When a part of the system is attacked or malfunctions, the protection mechanism intervenes immediately, quickly restoring system stability through local isolation and topology adjustment. This system, combining intelligent scheduling and self-healing protection, enables the system to maintain continuous and efficient operation in the extremely complex energy internet environment.

[0084] Compared to traditional centralized scheduling systems, the greatest advantage of this invention lies in its biomimetic architecture and distributed intelligence. It no longer relies on a single central node for command issuance, but instead empowers each node with "thinking" capabilities, distributing the decision-making process across the network. When the system scales to tens of thousands of nodes, the computational burden of this architecture does not increase exponentially, but maintains near-linear scalability through local autonomy and regional collaboration. Simultaneously, the event-driven communication mechanism significantly reduces communication frequency, latency, and energy consumption without sacrificing accuracy; while the introduction of multi-layered protection mechanisms ensures the system remains robust and reliable even when facing complex security threats. Experimental results show that in typical virtual power plant simulation scenarios, the system response time of this invention is reduced by approximately 70% compared to traditional scheduling frameworks, energy allocation deviation is reduced by approximately 40%, and system recovery time under attack is reduced by more than half.

[0085] In summary, this invention deeply integrates the collaborative optimization mechanism of neuromorphic computing with the trust guarantee mechanism of blockchain to construct a virtual power plant system with autonomous learning, self-optimization, and self-defense capabilities. Structurally, the system achieves layered progression, information closed-loop, and secure coupling; functionally, it realizes trusted scheduling, dynamic collaboration, and proactive protection; and operationally, it achieves an efficient, reliable, and verifiable intelligent energy management model. This system breaks through the architectural limitations of traditional virtual power plants, providing an evolvable, verifiable, and scalable system foundation for the intelligent collaboration and trusted operation of the future energy internet.

[0086] This invention has been applied and verified in a regional energy management platform. The regional resource pool includes controllable industrial loads, commercial air conditioning systems, residential energy storage devices, and electric vehicle charging / discharging nodes. Figure 3As shown, during system operation, the system first collects user behavior data and historical execution records from all potential participants in the resource pool. The raw dataset is then preprocessed, including outlier removal, missing sample imputation, and timestamp unification, to ensure the integrity and consistency of the input data. Subsequently, the cleaned user behavior characteristics, historical response volume, historical bids, and execution reliability indicators are input into the proposed prediction model to predict each user's adjustable capacity, expected bid, and response accuracy in the current demand response cycle. Based on the prediction results, a relatively robust conditional value-at-risk model is used to quantitatively assess the risk performance of each user under various possible external scenarios. Combining the worst-case conditions of each probability distribution, the relative risk coefficient relative to the baseline optimal solution is calculated, and the set of participating users that minimizes overall risk is determined accordingly. The user list obtained through model screening represents the optimal demand response solution for this cycle. This solution demonstrates a high response rate and economy in actual scheduling and execution, significantly outperforming the traditional price ranking selection method under unfavorable conditions.

[0087] Example 4: Based on the same inventive concept, this invention also provides a method for ensuring virtual power plant transactions through the collaboration of blockchain and neuromorphic computing, such as... Figure 4 ,include: Step 1: Collect the digital identity information and hardware physical characteristics of the node applying for access, generate a unique access credential through hash binding, and construct a trusted physical node set based on the blockchain consortium chain to verify the unique access credential; Step 2: Connect the nodes in the trusted physical node set into a network based on the spiking neural network algorithm, and dynamically reconstruct the network topology according to the node reputation and spatiotemporal attributes. The communication between nodes adopts an event-driven mechanism, which triggers the generation and transmission of pulse signals only when the state variable changes beyond a preset threshold. Step 3: Based on the characteristic information carried by the pulse signal, perform hierarchical scheduling calculation, generate verified scheduling instructions and issue them for execution. The hierarchical scheduling calculation includes scheduling calculation from immediate response at the edge side and collaborative optimization at the regional side to global overall planning at the central side. Step 4: Use a dual-chain architecture to classify and store event data and value data in the scheduling and computation process, execute multi-dimensional value settlement based on smart contracts, and encrypt the entire process data using quantum-resistant encryption technology.

[0088] Example 5 like Figure 5As shown, the present invention also provides an electronic device, which may be a computer device, a microcontroller device, a smart mobile device, etc. The electronic device in this embodiment may include a processor, a memory, a transceiver component, etc. The memory, processor, and transceiver component are connected via a bus; the memory can be used to store executable programs, and an exemplary executable program may include instructions; the processor is used to execute the instructions stored in the memory. The memory can also be used to store data, which can be accessed and / or modified when instructions are executed.

[0089] The processor may be a Central Processing Unit (CPU), or it may be other general-purpose processors, digital signal processors (DSPs), application-specific integrated circuits (ASICs), field-programmable gate arrays (FPGAs), or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, etc. It is the computing and control core of the terminal, and it is suitable for implementing one or more instructions. Specifically, it is suitable for loading and executing one or more instructions in the storage medium to realize the corresponding method flow or corresponding function, so as to realize the steps of the virtual power plant transaction guarantee method of blockchain and brain-like computing collaboration in the above embodiments.

[0090] Example 6 Based on the same inventive concept, this invention also provides a readable storage medium, specifically an electronic device readable storage medium (Memory). This readable storage medium is a memory device within an electronic device used to store programs and data. It is understood that the storage medium here can include both built-in storage media within the electronic device and extended storage media supported by the electronic device. The storage medium provides storage space, which stores the terminal's operating system. Furthermore, this storage space also stores one or more instructions suitable for loading and execution by a processor. These instructions can be one or more executable programs (including program code). It should be noted that the storage medium here can be high-speed RAM or non-volatile memory, such as at least one disk storage device. Loading and executing one or more instructions stored in the storage medium by the processor can implement the steps of the blockchain and neuromorphic computing collaborative virtual power plant transaction security method described in the above embodiments.

[0091] Those skilled in the art will understand that embodiments of the present invention can be provided as methods, systems, or computer program products. Therefore, the present invention can take the form of a completely hardware embodiment, a completely software embodiment, or an embodiment combining software and hardware aspects. Furthermore, the present invention can take the form of a computer program product embodied on one or more computer-usable storage media (including, but not limited to, disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code.

[0092] This invention is described with reference to flowchart illustrations and / or block diagrams of methods, apparatus (systems), and computer program products according to embodiments of the invention. It will be understood that each block of the flowchart illustrations and / or block diagrams, and combinations of blocks in the flowchart illustrations and / or block diagrams, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, special-purpose computer, embedded processor, or other programmable data processing apparatus to produce a machine, such that the instructions, which execute via the processor of the computer or other programmable data processing apparatus, generate instructions for implementing the flowchart illustrations and / or block diagrams. Figure 1 One or more processes and / or boxes Figure 1 A device that provides the functions specified in one or more boxes.

[0093] These computer program instructions may also be stored in a computer-readable storage medium that can direct a computer or other programmable data processing device to function in a particular manner, such that the instructions stored in the computer-readable storage medium produce an article of manufacture including instruction means, which are implemented in a process Figure 1 One or more processes and / or boxes Figure 1 The function specified in one or more boxes.

[0094] These computer program instructions may also be loaded onto a computer or other programmable data processing equipment to cause a series of operational steps to be performed on the computer or other programmable equipment to produce a computer-implemented process, thereby providing instructions that execute on the computer or other programmable equipment for implementing the process. Figure 1 One or more processes and / or boxes Figure 1 The steps of the function specified in one or more boxes.

[0095] The above are merely embodiments of the present invention and are not intended to limit the present invention. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention are included within the scope of the claims of the present invention pending approval.

Claims

1. A virtual power plant transaction security system that integrates blockchain and neuromorphic computing, characterized in that, include: The physical layer contains several distributed energy nodes, each of which embeds an edge agent unit. The physical layer is used to collect real-time operating status data of the device, convert the real-time operating status data into discrete pulse signals and upload them to the neural layer. It also serves as the system's bottom-level perception and execution end to receive and execute scheduling instructions from the upper layer. The neural layer, built on top of the physical layer through the edge proxy unit based on a spiking neural network, is used for inter-node communication using an event-driven mechanism. It triggers signal transmission only when the state variable changes exceed a preset threshold, and dynamically adjusts the connection topology and weights between nodes based on node reputation and physical distance. The decision layer, located above and connected to the neural layer, receives feature aggregation information uploaded by the neural layer, performs hierarchical scheduling calculations at multiple levels, generates verified scheduling instructions, and sends them down to the physical layer level by level through the neural layer. The hierarchical scheduling calculations include edge real-time optimization, regional collaborative optimization, and central global optimization. The feature aggregation information is generated by feature extraction and summarization processing of pulse signals through a spiking neural network. The protocol layer, located above the decision layer and connected to both the neural layer and the decision layer, is used to maintain a dual-chain storage structure containing an event chain and a value chain. It uses smart contracts to verify access identities, scheduling instructions, and settlement data, and combines quantum-resistant encryption technology to encrypt and protect the data transmission links between layers.

2. The system according to claim 1, characterized in that, The node access process in the physical layer includes: The node initiates a registration request to the consortium blockchain certificate authority at the protocol layer and receives a digital certificate generated by the certificate authority, which is generated from the node's registration information. A unique access credential is generated by binding the digital certificate with the hardware neural fingerprint hash formed by combining the hardware characteristics of the node. Based on the unique access credential, an access request containing a quantum-resistant signature is initiated to the protocol layer, and the node access is completed after verification by the protocol layer consortium blockchain verification node. Based on the digital certificate and the hardware neural fingerprint, a unique access credential is generated through hash binding, and an access request containing a quantum-resistant signature is sent to the protocol layer to achieve anti-forgery access for physical nodes.

3. The system according to claim 1, characterized in that, The neural layer includes a topology dynamic reconstruction module, which updates the connection weights between different nodes according to the following formula: in, For node i, a dynamic reputation score is given. For the dynamic reputation score of node j, The distance between nodes. The attenuation coefficient is... Let be the dynamic connection weight between node i and node j.

4. The system according to claim 1, characterized in that, The decision-making level includes: The edge real-time optimization unit, embedded in the edge agent unit, is used to process locally sensed pulse signals and prediction data based on the event-driven capabilities of neural layer nodes, in order to generate control actions with the goal of minimizing economic costs and physical deviations. The regional collaborative optimization unit is used to aggregate the impulse features of nodes within a region using a graph neural network, exchange key feature information through an impulse event network, and perform collaborative computation between adjacent node clusters using a distributed augmented Lagrange algorithm to output the coordinated optimization decision. The central global optimization unit integrates regional optimization results with power grid operation information, combines a hybrid deep prediction and mathematical programming solver for the whole network constrained operation, solves the global optimization model, and decomposes the optimization results into instructions for issuance to correct regional and edge strategies.

5. The system according to claim 4, characterized in that, The calculation formula in the edge instant optimization unit, which aims to minimize economic cost and physical deviation, is as follows: in, It is a node The adjustment decision vector, It is the first Decision variables of each node The corresponding economic cost function, It is the first The actual output of each node It is the first one issued by the upper management. The settings for each node, The penalty coefficient is... For the first The objective function for each node; The calculation formula for graph neural network aggregation in the regional collaborative optimization unit is as follows: in, yes The degree matrix, It is an activation function. Let be the weight matrix of the l-th layer. For optimized decision-making after coordination, For the first l Layer features, Let A be the augmented adjacency matrix of the original adjacency matrix A; The central global optimization unit solves the global optimization model using the following formula: in, To meet the needs of the entire network, For the first The set of physical constraints for each node. For the set of global decision variables, The total system cost, For decision variables At that time, the first The cost of a distributed energy unit This is the feasible region.

6. The system according to claim 1, characterized in that, The dual-chain storage structure of the protocol layer includes: The event chain is used to store high-frequency status pulse records, hashes of scheduling instructions generated by the decision layer, and abnormal alarm events; The value chain is used to store low-frequency fund settlement records and credit score changes; The event chain and the value chain are linked through a cryptographic index and a settlement mechanism is implemented using smart contracts.

7. The system according to claim 6, characterized in that, The settlement mechanism is implemented through the following calculation formula: in, For the first The settlement value of each node. For the first Effective response count per node As the benchmark price, For the first Individual node reputation The weighting coefficients, For the first Node response time The attenuation coefficient, For the first Execution accuracy of each node The reward coefficient, For the first Special reward and punishment clauses for each node.

8. The system according to claim 1, characterized in that, The protocol layer includes a quantum-safe module for: In the physical layer access and decision-layer instruction issuance stages, a lattice-based digital signature algorithm is used for digital signatures. In the data transmission link establishment stage between each layer, a lattice-based key encapsulation mechanism is used to establish an encrypted channel, and distributed key generation technology is used to distribute and manage the system's private keys.

9. The system according to claim 1, characterized in that, The decision-making level also includes: The edge layer anomaly detection unit is used to calculate the error through a pulse autoencoder and determine the node anomaly when the error meets the preset anomaly conditions. The regional layer immune response unit is used to combine evidence based on multi-node anomaly reports collected by the regional coordinator through Bayesian inference, calculate the posterior probability after the anomaly occurs, and execute local immune response actions after the calculated posterior probability exceeds a preset probability threshold. The central layer situational awareness unit is used to build a security knowledge graph based on on-chain records and reputation indexes, and to identify potential attack paths and covert data injection behaviors.

10. The system according to claim 9, characterized in that, The error is calculated using the following formula: in, It is the original feature vector. It is the vector reconstructed by the autoencoder. For error; The posterior probability is calculated using the following formula: in, For posterior probability, For the evidence set, For likelihood, For prior probability, This represents the marginal probability of the evidence.

11. A method for securing virtual power plant transactions through the collaboration of blockchain and neuromorphic computing, characterized in that, include: Collect the digital identity information and hardware physical characteristics of the node applying for access, generate a unique access credential through hash binding, and construct a trusted physical node set based on the blockchain consortium chain to verify the unique access credential; The nodes in the trusted physical node set are networked and connected based on the spiking neural network algorithm, and the network topology is dynamically reconstructed according to the node reputation and spatiotemporal attributes. The communication between nodes adopts an event-driven mechanism, which triggers the generation and transmission of pulse signals only when the state variable changes exceed a preset threshold. Based on the characteristic information carried by the pulse signal, hierarchical scheduling calculation is performed to generate verified scheduling instructions and issue them for execution. The hierarchical scheduling calculation includes scheduling calculation from immediate response at the edge side and collaborative optimization at the regional side to global overall planning at the central side. The dual-chain architecture is used to classify and store event data and value data in the scheduling and computing process, execute multi-dimensional value settlement based on smart contracts, and combine quantum-resistant encryption technology to encrypt the data throughout the process.

12. An electronic device, characterized in that, include: At least one processor and memory; The memory and processor are connected via a bus; The memory is used to store one or more programs; When the one or more programs are executed by the at least one processor, the virtual power plant transaction security method of blockchain and neuromorphic computing collaboration as described in claim 11 is implemented.

13. A readable storage medium, characterized in that, It contains an execution program, which, when executed, implements the virtual power plant transaction security method that combines blockchain and neuromorphic computing as described in claim 11.

Images