A virtual power plant transaction method based on inter-energy fusion scenarios

Abstract

This invention discloses a virtual power plant trading method based on a transportation-energy integration scenario, belonging to the field of virtual power plant trading technology. It includes the following steps: (1) transportation-energy data fusion collection and preprocessing; (2) hierarchical resource aggregation and status assessment; (3) dynamic trading demand prediction and matching; (4) blockchain trusted execution and feedback optimization; and (5) collaborative management of the entire battery lifecycle. This invention enables efficient collaborative trading of transportation and energy resources, significantly improving resource collaboration efficiency. Simultaneously, the trading response speed meets the grid's needs, and a trusted and secure trading system is constructed, maximizing battery value utilization and yielding significant economic and environmental benefits.

Description

Technical Field

[0001] This invention belongs to the field of virtual power plant trading technology, specifically relating to a virtual power plant trading method based on a power exchange integration scenario. Background Technology

[0002] Currently, the integration of transportation and energy has become one of the core directions for the construction of new power systems. Virtual power plants (VPPs), as a key technology for distributed resource aggregation and management, are gradually being incorporated into flexible transportation resources such as electric vehicles and charging / swapping stations. The most representative existing technology is the "blockchain + V2G (vehicle-to-grid)" virtual power plant trading model. This technology uses blockchain to achieve the storage and sharing of distributed energy data, combined with smart contracts to automate the power trading process, and has been applied in pilot projects in Shanghai, Zhejiang, and other places. Specifically, the core architecture of the existing technology includes three parts: first, a resource aggregation layer, which accesses distributed resources such as electric vehicles and charging / swapping stations through a centralized platform to obtain their basic power information and trading intentions; second, a trading matching layer, which generates trading prices and selects trading counterparties based on the predicted power generation and electricity demand of the virtual power plant through smart contracts; and third, an execution feedback layer, which completes the power charging and discharging scheduling and transaction settlement, while synchronizing the results to the blockchain nodes. For example, a virtual power plant project in Zhejiang Province aggregated V2G charging piles and commercial air conditioning loads using this model, achieving a maximum daily load regulation capacity of 920,000 kilowatts and a cumulative response power of over 10 million kilowatt-hours. A pilot heavy-duty truck charging and battery swapping station in Shanghai aggregated the 200,000 kilowatt-hour battery capacity of 500 electric heavy-duty trucks, achieving an annual discharge of 260,000 kilowatt-hours, but still adopted a crude transaction management approach of fixed-time charging and discharging.

[0003] Although the above technologies have achieved a preliminary integration of transportation resources and energy networks, they still have four major shortcomings in practical applications, all of which are directly related to the key problems that this invention aims to solve:

[0004] 1. Insufficient Depth of Transportation-Energy Resource Integration: Existing technologies only treat electric vehicles as static energy storage units for power aggregation, without considering the dynamic characteristics of transportation scenarios. For example, when the real-time location of electric heavy trucks deviates from the plan due to adjustments in transportation tasks, the existing system cannot update its charging and discharging availability in a timely manner, resulting in approximately 30% of potential adjustable resources being idle, severely impacting the regulation capabilities of the virtual power plant. Taking the electric container truck transportation scenario at Shanghai Yangshan Port as an example, changes in vehicle scheduling caused by tidal operations result in a daily adjustable capacity loss of 52,000 kWh for the traditional system.

[0005] 2. Significant lag in transaction response: Existing technologies rely on day-ahead forecast data to formulate transaction plans, lacking a real-time dynamic adjustment mechanism. The existing system uses a standard LSTM model for load forecasting, and when there are intraday load fluctuations in the power grid, the forecast deviation rate can reach 8.7%, and the transaction response delay is generally 15-30 minutes, which is far from meeting the distribution network's demand for second-level and minute-level regulation, resulting in a serious disconnect between transaction plans and actual supply and demand.

[0006] 3. Vulnerabilities exist in the trusted transaction system: Although existing blockchain transaction models have achieved data storage, they have not resolved the contradiction between "data authenticity verification" and "privacy protection." Sensitive data such as real-time load data of charging and battery swapping stations and travel information of electric vehicle users are either unusable for transaction optimization due to encryption or pose a risk of privacy leakage due to public sharing. At the same time, there is a lack of effective identification mechanisms for abnormal transactions such as double payment and mismatch between electricity and amount.

[0007] 4. The full life-cycle value of batteries is not fully explored: Current technology does not differentiate between the characteristics of power batteries with different states of health (SOH), and uniformly applies the same charging and discharging strategies to participate in trading. For retired power batteries with an SOH below 80%, blindly charging and discharging will not only further shorten their lifespan by more than 40%, but also increase the risk of grid harmonic pollution, while wasting their potential as backup energy storage. A pilot project in a community in Jincheng showed that the secondary utilization cycle of unclassified retired batteries is less than 2 years, while it can be extended to more than 5 years after scientific classification. Summary of the Invention

[0008] To address the aforementioned deficiencies in existing technologies, this invention provides a virtual power plant trading method based on a transportation and energy integration scenario. Through technical means such as resource hierarchical aggregation, dynamic prediction and matching, reliable execution feedback, and full lifecycle management, it enables efficient collaborative trading of transportation and energy resources.

[0009] The technical solution adopted in this invention is as follows:

[0010] A virtual power plant trading method based on energy exchange integration scenarios includes the following steps:

[0011] (1) Data fusion acquisition and preprocessing of transportation and energy:

[0012] (1.1) Accurate collection of multi-source data: Collection of traffic data, energy data, and environmental data;

[0013] (1.2) Data preprocessing and fusion: The collected traffic data, energy data and environmental data are preprocessed, and the traffic, energy and environmental data are associated and matched by spatiotemporal alignment algorithm to generate a standardized dataset;

[0014] (2) Hierarchical resource aggregation and status assessment:

[0015] (2.1) Level 1 aggregation: Single resource layer aggregation, to assess the adjustable capacity and response capability of a single electric vehicle or a single charging and battery swapping station;

[0016] (2.2) Second-level aggregation: Scene resource layer aggregation, which is classified and collaborative capability is evaluated according to application scenarios;

[0017] The scenario classification criteria are divided into four categories: port container truck group, urban delivery group, community charging and swapping group, and trunk transportation group, with a preset weight coefficient for each category.

[0018] (2.3) Three-level aggregation: aggregation at the virtual power plant level to generate a global resource profile;

[0019] The aggregated capacity, response speed, and SOH distribution parameters of all scenario resource groups are summarized to calculate the total adjustable capacity, average response delay, and peak-valley regulation potential of the virtual power plant.

[0020] Construct a resource profile tagging system, including five dimensions: total capacity, response level, SOH structure, predicted availability period, and geographical distribution. Each tag corresponds to quantitative data, providing a basis for decision-making in transaction matching.

[0021] (3) Dynamic transaction demand forecasting and matching:

[0022] (3.1) GA-LSTM dual-timescale demand forecasting;

[0023] (3.2) Smart contract multi-target transaction matching;

[0024] (3.3) Transaction plan generation and confirmation, specifically:

[0025] (3.3.1) Generate a standardized transaction order containing transaction ID, pseudonym of the participant, time period, electricity, price, response requirements, and liability for breach of contract. The participant's identity is identified by a pseudonym generated using zero-knowledge proof, which is bound to the real identity and stored in the consortium blockchain.

[0026] (3.3.2) The order is synchronized to all blockchain nodes for pre-deposit. After the participants confirm through the edge terminal, the order officially takes effect and locks the adjustable capacity of the corresponding resources. The lock duration is the transaction period plus a 30-minute buffer period.

[0027] (4) Optimization of trusted execution and feedback in blockchain:

[0028] (4.1) Precise execution at edge nodes;

[0029] (4.2) Trusted storage and verification of blockchain data;

[0030] (4.3) Closed-loop optimization iteration;

[0031] (5) Collaborative management of the entire battery lifecycle:

[0032] (5.1) Construction of digital twin archives for batteries;

[0033] (5.2) Adaptation of phased trading strategies;

[0034] (5.3) Full life cycle revenue management.

[0035] Furthermore, in step (1.1), the collection of traffic data specifically involves:

[0036] The vehicle-mounted terminal collects real-time GPS location and speed information of electric vehicles at a frequency of 1Hz; the TMS transportation management system synchronizes and plans the trip, including departure time, destination, and estimated stay duration; and the battery management system collects battery status data every 30 seconds, including SOH data, SOC data, single-cell voltage equalization, real-time charging and discharging power, and temperature.

[0037] The specific steps for collecting energy-related data are as follows:

[0038] The edge intelligent controller collects real-time load, remaining battery quantity and status of the charging and swapping station every 15 seconds; obtains 15-minute output forecast data, time-of-use electricity price and ancillary service price of distributed photovoltaic / wind power from the power grid dispatch system; and obtains time-of-use metering of traded electricity through smart meters.

[0039] The specific steps for collecting environmental data are as follows:

[0040] Sensors deployed at charging and battery swapping stations and transportation trunk lines collect real-time road conditions, wind speed, and light intensity every 5 minutes to correct energy forecasting errors.

[0041] The specific interface standard is as follows:

[0042] Traffic data is transmitted using the J1939 protocol, and energy data is transmitted using the DL / T 645 protocol. All data is uploaded to the traffic and energy cloud platform via 4G / 5G network.

[0043] The specific steps (1.2) are as follows:

[0044] The Kalman filter algorithm is used to smooth the vehicle GPS data. The state equation is set as x(k)=A x(k-1)+w(k) and the observation equation is set as z(k)=H x(k)+v(k), where the process noise w(k) and the observation noise v(k) both follow a Gaussian distribution.

[0045] The Raida criterion is used to remove outliers from the load data of charging and swapping stations. The mean μ and standard deviation σ of historical load data are calculated. Data that exceeds the interval [μ-3σ, μ+3σ] are marked as outliers and are filled in using linear interpolation.

[0046] By using a spatiotemporal alignment algorithm, traffic, energy, and environmental data are correlated and matched to generate a standardized dataset with a time granularity of 15 minutes.

[0047] Further, step (2.1) specifically includes:

[0048] (2.1.1) Calculation of adjustable capacity for electric vehicles:

[0049] Adjustable capacity = Rated battery capacity × SOC × Adjustment coefficient × Traffic availability coefficient;

[0050] Among them: the adjustment coefficient is dynamically set according to the SOH; the traffic availability coefficient is calculated based on the trip plan;

[0051] (2.1.2) Calculation of adjustable capacity of charging and battery swapping stations:

[0052] Adjustable capacity = (maximum charging power - current charging power) - minimum load power;

[0053] Among them: the minimum load power is the minimum power required to maintain the basic operation of the battery swapping station; the number of dispatchable batteries is also calculated.

[0054] (2.1.3) Response capability marker: Record the response latency and continuous response duration parameters of individual resources as indicators for transaction matching.

[0055] Furthermore, step (2.2) specifically includes:

[0056] (2.2.1) Calculation of aggregation parameters:

[0057] Scene aggregation capacity = Σ (adjustable capacity of individual resources × scene weight coefficient);

[0058] Scene response speed = Σ (individual resource response latency × resource capacity percentage);

[0059] SOH Distribution in a Scenario = Statistics on the number and capacity percentage of batteries with different SOH levels in this scenario;

[0060] (2.2.2) Grouping constraints: The aggregated capacity of each scenario resource group must be ≥1MW, and the individual resources must be located at the same spot market clearing node to ensure the execution of scheduling instructions.

[0061] Furthermore, step (3.1) specifically includes:

[0062] (3.1.1) Model construction: The LSTM network optimized by the genetic algorithm is used. The genetic algorithm is used to optimize the number of hidden layer units, time step, dropout ratio and learning rate of the LSTM. The encoding method is a 10-bit integer chromosome. The fitness function adopts the validation set MSE.

[0063] (3.1.2) Day-ahead forecast: Input historical transaction data of the past 7 days, traffic plan for the next day, and weather forecast data, and output the basic transaction electricity demand at 96 time points within 24 hours;

[0064] (3.1.3) Intraday Rolling Forecast: By integrating real-time traffic conditions, grid load fluctuation data, and dynamic data on battery status changes, the day-ahead forecast results are corrected using the following formula:

[0065] Corrected demand = day-ahead forecast × (1 + grid load fluctuation coefficient × 0.3 + traffic delay coefficient × 0.7).

[0066] Where: Power grid load fluctuation coefficient = (current load - predicted load) / predicted load, traffic delay coefficient = (actual travel time - planned travel time) / planned travel time;

[0067] (3.1.4) Prediction result output: Generate a demand curve that includes time period, predicted power consumption, confidence interval, and adjustment trigger threshold. When the actual deviation exceeds the threshold, the prediction will be automatically restarted.

[0068] Furthermore, step (3.2) specifically includes:

[0069] (3.2.1) Contract preset rules: Based on local power trading rules, clarify the terms of trading types, price ranges, and settlement methods, and distinguish between load-type and generation-type virtual power plant identities;

[0070] (3.2.2) Matching objective function: With the fastest response speed, the lowest transaction cost, and the minimum battery loss as multiple objectives, the weighted summation method is used to transform it into a single objective optimization;

[0071] (3.2.3) Classification matching logic:

[0072] Electricity sales scenario: Prioritize matching resource groups with response speed ≤ 5 minutes and SOH 60%-80%. The screening criteria are transaction unit price ≤ grid peak-valley price difference × 0.9. The counterparty is determined by sorting the response speed in ascending order and the price in descending order.

[0073] Electricity purchase scenario: Prioritize matching charging and battery swapping stations in areas with abundant photovoltaic output and SOH ≥ 80%. The screening criteria are that the transaction unit price is ≥ the grid off-peak electricity price × 1.1, and the stations are sorted in ascending order of price and descending order of capacity.

[0074] Ancillary service sites: Only resource groups with a response speed of ≤2 minutes and SOH ≥85% are matched, and the charging and discharging power fluctuation range is required to be ≤10% / minute, sorted by adjustment accuracy;

[0075] (3.2.4) Conflict detection and arbitration: The smart contract automatically detects transaction conflicts, including double-spending conflicts and validity conflicts. It arbitrates conflicting transactions according to the principles of time priority and capacity priority, removes invalid transactions and records them in the log.

[0076] Further, step (4.1) specifically includes:

[0077] (4.1.1) Instruction issuance: The blockchain consortium chain parses the effective orders into charging and discharging instructions, and issues them to the resource group in real time through the edge smart controller. The instructions include start time, end time, power curve, and safety threshold parameters.

[0078] (4.1.2) Differentiated execution strategy:

[0079] Electric vehicles: An intermittent charging and discharging strategy is adopted without affecting the trip; if the trip changes, the vehicle terminal will automatically send a change request to the platform, triggering order adjustment.

[0080] Charging and battery swapping stations: adopt a tiered strategy of charging first and then discharging. During off-peak hours, priority is given to charging batteries with SOH ≥ 80%, and during peak hours, priority is given to discharging batteries with SOH 60%-80%. At the same time, 20% capacity of backup batteries is reserved to cope with sudden demand.

[0081] Execution monitoring: The edge node collects actual charging and discharging data every 5 seconds and compares it with the command curve. If the deviation exceeds 5%, local adjustment is initiated, and if it exceeds 10%, it is immediately reported to the platform.

[0082] Furthermore, step (4.2) specifically includes:

[0083] (4.2.1) Encrypted data upload: Every 5 minutes, the actual charging and discharging data, battery status data, and vehicle location data are double-encrypted and uploaded to the consortium blockchain after generating ciphertext data;

[0084] (4.2.2) Zero-knowledge proof verification: A polynomial equation is constructed using FO commitment to verify the authenticity of the uploaded data. If the verification passes, the data is stored; if it fails, the data is destroyed and a source tracing warning is triggered.

[0085] (4.2.3) Evidence storage content: including encrypted transaction data, verification results, participant signatures, timestamps, generating a unique QR code, and linking transaction location information to form a traceability information package to ensure that the data is tamper-proof and traceable;

[0086] (4.2.4) Key Management: Paillier's private key is stored in a hierarchical manner across three trusted institutions using the Shamir secret sharing mechanism. Decryption requires the collaboration of at least two of these institutions to prevent key leakage.

[0087] Furthermore, step (4.3) specifically includes:

[0088] (4.3.1) Daily optimization: After the daily transaction ends, extract the full amount of data stored in the blockchain, calculate the response deviation rate and battery loss rate of each resource group, update the training dataset of the GA-LSTM model, and re-optimize the model parameters.

[0089] (4.3.2) Weekly optimization: Analyze the collaborative efficiency of the scenario resource group every week and adjust the scenario weight coefficient; update the adjustment coefficient according to the battery SOH change to ensure the adaptability of the strategy;

[0090] (4.3.3) Optimization result synchronization: The updated parameters are synchronized to each functional module to form a closed-loop management of prediction, execution, verification and optimization.

[0091] Furthermore, step (5.1) specifically includes:

[0092] (5.1.1) Archive dimension: Assign a unique ID to each battery, and build a digital twin archive based on blockchain-stored data, which includes factory parameters, charging and discharging history, dynamic changes in SOH, fault records, and transaction participation records, and update the status data every 15 minutes;

[0093] (5.1.2) Accurate SOH calculation: The SOH value is calculated by integrating voltage monitoring method, internal resistance measurement method and capacity testing method and using a weighted algorithm;

[0094] Step (5.2) specifically includes:

[0095] (5.2.1) Health period with SOH≥80%: Priority participation in frequency regulation and backup auxiliary services, charge and discharge depth controlled between 20%-80%, and constant current-constant voltage segmented charging strategy to avoid overcharge damage;

[0096] (5.2.2) Decay period of 60%≤SOH<80%: For medium frequency electric energy trading that participates in peak shaving and valley filling, the depth of charge and discharge should be controlled at 30%-70%, and the discharge power should be limited to 70% of the rated power to avoid high power impact;

[0097] (5.2.3) Retirement period with SOH < 60%: It is converted into a backup energy storage resource, only participating in emergency power supply to the grid, with a charge and discharge depth ≤ 50%, and the charging temperature controlled at 15-35℃, extending the secondary utilization cycle to 3-5 years;

[0098] Step (5.3) specifically includes:

[0099] (5.3.1) Revenue sharing: The revenue sharing ratio is set according to the battery SOH level and the type of transaction, and is automatically settled to the battery owner's account through smart contract every month;

[0100] (5.3.2) Decommissioning warning: When the SOH drops to 55%, a decommissioning warning will be automatically triggered, and it is recommended to switch to community energy storage or emergency power supply scenarios.

[0101] In summary, due to the adoption of the above technical solution, the beneficial effects of the present invention are:

[0102] 1. Significantly Improved Resource Coordination Efficiency: Through the fusion and hierarchical aggregation of transportation and energy data, the resource utilization rate of the virtual power plant is increased by more than 45% compared to existing technologies. Taking the Shanghai heavy-duty truck charging and swapping pilot scenario as an example, after adopting this method, the adjustable capacity of a single station increased from 200,000 kWh to 360,000 kWh, and the annual discharge volume increased from 260,000 kWh to 416,000 kWh, equivalent to the regulation capacity of three new traditional energy storage stations.

[0103] 2. Transaction response speed meets power grid requirements: The GA-LSTM prediction model reduces the intraday prediction deviation rate to below 3.1%, and the combination of edge computing and blockchain shortens the transaction response latency to less than 2 minutes, a reduction of more than 90% compared to existing technologies. In simulation tests on the Zhejiang power grid, this method controls the response deviation rate to load fluctuations to within 5%, fully meeting the second-level adjustment requirements of the distribution network.

[0104] 3. Constructing a trusted and secure transaction system: The application of differential privacy encryption, double homomorphic encryption, and zero-knowledge proof technologies ensures the authenticity of transaction data while achieving end-to-end protection of user privacy. Testing shows that this method can resist over 99% of common data attacks, increases the difficulty of tampering with transaction information by three orders of magnitude compared to existing blockchain solutions, and reduces the risk of user data leakage to below 0.1%.

[0105] 4. Maximizing Battery Value: The differentiated trading strategy based on SOH extends the overall lifespan of power batteries by more than 30%, and the secondary utilization cycle of retired batteries can reach 3-5 years. In the pilot scenario in Jincheng community, after adopting this method, the total life-cycle revenue of batteries increased by 25%, and the cost per kilowatt-hour of a single battery decreased by 0.12 yuan.

[0106] 5. Significant economic and environmental benefits: The transaction revenue of virtual power plant operators is 30%-50% higher than that of existing technologies, and the battery maintenance costs of participating users are reduced by 20%; at the same time, a single charging and battery swapping station can reduce carbon emissions by 1,200 tons per year, providing technical support for achieving the "dual carbon" goal. Detailed Implementation

[0107] To make the objectives, technical solutions, and advantages of the embodiments of the present invention clearer, the technical solutions of the embodiments of the present invention will be clearly and completely described below in conjunction with the embodiments of the present invention. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. The components of the embodiments of the present invention described and shown herein can generally be arranged and designed in various different configurations.

[0108] Therefore, the following detailed description of the embodiments of the present invention is not intended to limit the scope of the claimed invention, but merely to illustrate selected embodiments of the invention. All other embodiments obtained by those skilled in the art based on the embodiments of the present invention without inventive effort are within the scope of protection of the present invention.

[0109] It should be noted that labels and letters indicate similar items, therefore, once an item is defined, it does not need to be further defined and explained subsequently.

[0110] In the description of this invention, it should be noted that the terms "center," "upper," "lower," "left," "right," "vertical," "horizontal," "inner," and "outer," indicating orientation or positional relationships, or the orientation or positional relationships in which the product of this invention is conventionally placed during use, are merely for the purpose of simplifying the description of this invention and do not indicate or imply that the device or element referred to must have a specific orientation, or be constructed and operated in a specific orientation. Therefore, they should not be construed as limitations on this invention. Furthermore, the terms "first," "second," and "third," etc., are only used to distinguish descriptions and should not be construed as indicating or implying relative importance.

[0111] Furthermore, terms such as "horizontal" and "vertical" do not imply that components must be absolutely horizontal or suspended, but rather that they can be slightly tilted. For example, "horizontal" simply means that its direction is more horizontal than "vertical," and does not mean that the structure must be completely horizontal, but can be slightly tilted.

[0112] In the description of this invention, it should also be noted that, unless otherwise explicitly specified and limited, the terms "set," "install," "connect," and "link" should be interpreted broadly. For example, they can refer to a fixed connection, a detachable connection, or an integral connection; they can refer to a mechanical connection or an electrical connection; they can refer to a direct connection or an indirect connection through an intermediate medium; and they can refer to the internal communication between two components. Those skilled in the art can understand the specific meaning of the above terms in this invention based on the specific circumstances.

[0113] A virtual power plant trading method based on energy exchange integration scenarios includes the following steps:

[0114] Step 1: Transportation-Energy Data Fusion Acquisition and Preprocessing

[0115] 1. Precise collection of multi-source data

[0116] —Traffic-related data: Real-time GPS location and speed information of electric vehicles are collected at a frequency of 1Hz through the on-board unit (OBU); the trip is planned synchronously from the TMS transportation management system (including departure time, destination, and estimated stay duration); battery status data is collected every 30 seconds through the BMS (battery management system), including SOH (calculated based on voltage-internal resistance joint detection method), SOC (state of charge), cell voltage equalization, real-time charge and discharge power, and temperature (accuracy ±0.5℃).

[0117] —Energy-related data: Real-time load, remaining battery quantity and status of charging and battery swapping stations are collected every 15 seconds through edge intelligent controllers; 15-minute output forecast data, time-of-use electricity prices (peak-valley price difference ≥ 0.7 yuan / kWh) and ancillary service prices (peak shaving and frequency regulation compensation standards) of distributed photovoltaic / wind power are obtained from the power grid dispatch system; Time-of-use metering of traded electricity is achieved through smart meters (accuracy 0.01kWh).

[0118] —Environmental data: Sensors deployed at charging and battery swapping stations and transportation trunk lines collect real-time traffic conditions (congestion level 1-5), wind speed (accuracy ±0.1m / s), and light intensity (accuracy ±10lux) every 5 minutes to correct energy forecasting biases.

[0119] — Interface standards: Vehicle data is transmitted using the J1939 protocol, and energy data is transmitted using the DL / T 645 protocol. All data is uploaded to the transportation energy cloud platform via 4G / 5G network, and the data collection integrity rate is no less than 99%.

[0120] 2. Data Preprocessing and Fusion

[0121] —Noise Reduction Processing: The Kalman filter algorithm is used to smooth the vehicle GPS data. The state equation is set as x(k)=A x(k-1)+w(k), and the observation equation is set as z(k)=H x(k)+v(k). The process noise w(k) and the observation noise v(k) both follow a Gaussian distribution. After processing, the position error is controlled within 5 meters.

[0122] —Outlier removal: Outliers in the load data are removed using the Raida criterion (3σ rule). The mean μ and standard deviation σ of the historical load data are calculated. Data that exceeds the interval [μ-3σ, μ+3σ] are marked as outliers and then filled in using linear interpolation.

[0123] —Data Fusion: By using a spatiotemporal alignment algorithm to associate and match traffic, energy, and environmental data, a standardized dataset is generated with a 15-minute time granularity, containing 18 feature fields including "vehicle ID, location, battery status, charging and swapping station load, electricity price, and weather".

[0124] Step 2: Hierarchical resource aggregation and status assessment

[0125] 1. Level 1 Aggregation (Single Resource Layer): Assess the adjustable capacity and response capabilities of a single electric vehicle or a single charging / battery swapping station.

[0126] (1) Calculation of adjustable capacity of electric vehicles:

[0127] Adjustable capacity = Rated battery capacity × SOC × Adjustment coefficient × Traffic availability coefficient

[0128] Among them: the adjustment coefficient is dynamically set according to SOH (0.8 when SOH≥80%, 0.5 when 60%≤SOH<80%, and 0.2 when SOH<60%); the traffic availability coefficient is calculated based on the trip plan, with 1.0 for the dwell time, 0.1 for the travel time, and 0.6 for the expected arrival time at the charging and battery swapping station within 1 hour.

[0129] Example: An electric heavy truck equipped with a 140kWh battery, with SOH=85% and SOC=70%, is currently docked at a port. The adjustable capacity is 140×0.7×0.8×1.0=78.4kWh.

[0130] (2) Calculation of adjustable capacity of charging and battery swapping stations:

[0131] Adjustable capacity = (Maximum charging power - Current charging power) - Minimum load power

[0132] Among them: the minimum load power is the minimum power to maintain the basic operation of the battery swapping station (including the monitoring system, battery swapping equipment standby, etc., usually taken as 15% of the total capacity); at the same time, the number of dispatchable batteries needs to be calculated, that is, the number of batteries that meet the current SOH requirements and have not been reserved × the capacity of a single battery.

[0133] (3) Response capability marker: Record parameters such as response delay (vehicles ≤ 5 minutes, charging and swapping stations ≤ 2 minutes) and continuous response duration (≥ 1 hour) of individual resources as core indicators for transaction matching.

[0134] 2. Secondary Aggregation (Scenario Resource Layer): Categorization and Collaboration Capability Assessment Based on Application Scenarios

[0135] Scenario classification criteria: divided into four categories: port truck group (fixed routes, concentrated stops), urban delivery group (flexible routes, dispersed stops), community charging and swapping group (fixed stations, random battery swapping by users), and trunk transportation group (long distance, battery swapping along the way). Each category of scenario has a preset weight coefficient (port truck group 1.2, trunk transportation group 1.0, urban delivery group 0.9, community charging and swapping group 0.8).

[0136] (1) Calculation of aggregation parameters:

[0137] Scene aggregation capacity = Σ (adjustable capacity of individual resources × scene weight coefficient)

[0138] Scene response speed = Σ (individual resource response latency × resource capacity percentage)

[0139] SOH Distribution in a Scenario = Statistics on the number and capacity percentage of batteries with different SOH levels in this scenario (e.g., for port truck groups with SOH ≥ 80%, the battery percentage must be ≥ 60%).

[0140] (2) Grouping constraints: The aggregated capacity of each scenario resource group must be ≥1MW, and the individual resources must be located at the same spot market clearing node (220 kV and above voltage level bus) to ensure the efficient execution of dispatch instructions.

[0141] 3. Three-level aggregation (virtual power plant layer): Generates a global resource profile.

[0142] (1) Summarize the aggregated capacity, response speed, SOH distribution and other parameters of all scenario resource groups, and calculate the core indicators such as total adjustable capacity of virtual power plants (≥5MW), average response delay (≤3 minutes), peak-valley regulation potential (peak discharge capacity / valley charging capacity ≥0.8).

[0143] (2) Construct a resource profile tag system, including five dimensions: "total capacity - response level - SOH structure - predicted available time period - geographical distribution". Each tag corresponds to quantitative data (such as response level divided into "second-level response", "minute-level response" and "hour-level response"), to provide decision-making basis for transaction matching.

[0144] Step 3: Dynamic Transaction Demand Forecasting and Matching

[0145] 1. GA-LSTM Dual-Time-Scale Demand Forecasting

[0146] (1) Model construction: The LSTM network optimized by the genetic algorithm (GA) is used. GA is used to optimize the number of hidden layer units (50-200), time step (12-72), dropout ratio (0.2-0.5) and learning rate (0.001-0.01) of the LSTM. The encoding method is a 10-bit integer chromosome (the first 6 bits represent the time window and the last 4 bits represent the number of units). The fitness function is the validation set MSE (mean squared error).

[0147] (2) Forecast before 08:30 on T-1 day: Input historical transaction data of the past 7 days, traffic plan for the next day (such as port container truck operation schedule), and weather forecast data, and output the basic transaction electricity demand at 96 time points within 24 hours. The forecast accuracy should be above 95% (MAPE≤3.1%).

[0148] (3) Intraday rolling forecast (updated every 15 minutes): Integrating real-time traffic conditions (congestion coefficient), power grid load fluctuation data (real-time deviation rate), battery status changes, and other dynamic data, the day-ahead forecast results are corrected using the following formula:

[0149] Corrected demand = Day-ahead forecast × (1 + Grid load fluctuation coefficient × 0.3 + Traffic delay coefficient × 0.7)

[0150] Among them: power grid load fluctuation coefficient = (current load - predicted load) / predicted load, traffic delay coefficient = (actual travel time - planned travel time) / planned travel time.

[0151] (4) Prediction result output: Generate a demand curve containing "time period - predicted power - confidence interval - adjustment trigger threshold". When the actual deviation exceeds the threshold (±10%), the prediction will be automatically restarted.

[0152] 2. Smart Contract Multi-Target Transaction Matching

[0153] (1) Contract preset rules: Based on the Guangdong Power Trading Rules, the core terms such as the trading varieties (electric energy, demand response, ancillary services), price range (0-1500 yuan / MWh), and settlement method are clearly defined, and the identities of load-type and generation-type virtual power plants are distinguished (load-type only buys electricity, while generation-type can buy and sell).

[0154] (2) Matching objective function: With "fastest response speed, lowest transaction cost, and minimum battery loss" as multiple objectives, the weighted summation method is used to transform it into a single objective optimization, with the weights allocated as response speed (0.4), transaction cost (0.3), and battery loss (0.3).

[0155] (3) Classification matching logic:

[0156] — Electricity sales scenario (participating in peak shaving): Prioritize matching resource groups with response speed ≤ 5 minutes and SOH 60%-80%. The screening criteria are transaction unit price ≤ grid peak-valley price difference × 0.9. The counterparty is determined by sorting the response speed in ascending order and the price in descending order.

[0157] —Electricity purchase scenario (valley filling): Prioritize matching charging and battery swapping stations in areas with abundant photovoltaic output (sunlight intensity ≥ 800 lux) and SOH ≥ 80%. The screening criteria are that the transaction unit price is ≥ the grid valley electricity price × 1.1, and the stations are sorted in ascending order of price and descending order of capacity.

[0158] —Auxiliary service scenario (frequency adjustment): Only resource groups with response speed ≤ 2 minutes and SOH ≥ 85% are matched, and the charging and discharging power fluctuation range is required to be ≤ 10% / minute, sorted by adjustment accuracy.

[0159] (4) Conflict detection and arbitration: The smart contract automatically detects transaction conflicts, including double payment conflicts (two transactions matching the same resource at the same time) and validity conflicts (electricity and amount do not match). The conflicting transactions are arbitrated using the "first-come, first-served + capacity priority" principle, invalid transactions are eliminated and the log is recorded.

[0160] 3. Transaction plan generation and confirmation

[0161] (1) Generate a standardized transaction order containing “transaction ID-participant pseudonym-time period-electricity-price-response requirements-liability for breach of contract”, wherein the participant identity is identified by a pseudonym generated by zero-knowledge proof and stored in the consortium blockchain in conjunction with the real identity.

[0162] (2) The order is synchronized to all blockchain nodes for pre-deposit. After the participants confirm through the edge terminal, the order officially takes effect and locks the adjustable capacity of the corresponding resources. The lock duration is the transaction period + 30-minute buffer period.

[0163] Step 4: Optimization of Trusted Execution and Feedback in Blockchain

[0164] 1. Precise execution at edge nodes

[0165] (1) Instruction issuance: The blockchain consortium chain parses the effective order into a charging and discharging instruction, and issues it to the resource group in real time through the edge smart controller. The instruction includes parameters such as "start time - end time - power curve - safety threshold (e.g., battery temperature ≤ 45℃)" and the issuance delay is ≤ 200ms.

[0166] (2) Differentiated execution strategy:

[0167] —Electric vehicles: Without affecting the trip, “intermittent charging and discharging” is adopted (such as port trucks discharging during loading and unloading intervals, each lasting 15-20 minutes); if the trip changes, the vehicle terminal automatically sends a change request to the platform, triggering order adjustment.

[0168] —Charging and battery swapping stations: adopt a "charge first, discharge later" tiered strategy, prioritizing charging batteries with SOH≥80% during off-peak periods and prioritizing discharging batteries with SOH 60%-80% during peak periods, while reserving 20% ​​capacity of backup batteries to cope with sudden demand.

[0169] —Execution monitoring: The edge node collects actual charging and discharging data every 5 seconds and compares it with the command curve. If the deviation exceeds 5%, local adjustment is initiated, and if it exceeds 10%, it is immediately reported to the platform.

[0170] 2. Blockchain-based trusted evidence storage and verification

[0171] (1) Data encryption upload: Every 5 minutes, the actual charging and discharging data, battery status data, and vehicle location data are encrypted using double encryption (first layer Paillier semi-homomorphic encryption, second layer BGV fully homomorphic encryption) and then uploaded to the consortium blockchain to generate ciphertext data.

[0172] (2) Zero-knowledge proof verification: The FO commitment is used to construct a polynomial equation to verify the authenticity of the uploaded data (e.g., the power data must meet the condition "charging and discharging power × duration = cumulative power"). If the verification is successful, the evidence is stored; if it fails, the data is destroyed and a traceability warning is triggered.

[0173] (3) Evidence content: including encrypted transaction data, verification results, participant signatures, timestamps, generating a unique QR code, and linking transaction location information to form a traceability information package to ensure that the data is tamper-proof and traceable.

[0174] (4) Key management: Paillier private keys are stored in layers among three trusted institutions (power grid company, operator and regulatory agency) through the Shamir secret sharing mechanism. At least two institutions need to cooperate to decrypt the keys to prevent key leakage.

[0175] 3. Closed-loop optimization iteration

[0176] (1) Daily optimization: After the daily transaction ends (T+1 day), extract the full amount of data stored in the blockchain, calculate the response deviation rate, battery loss rate and other indicators of each resource group, update the training dataset of the GA-LSTM model, and re-optimize the model parameters to improve the prediction accuracy of the next day by 2%-3%.

[0177] (2) Weekly optimization: Analyze the collaborative efficiency of the scenario resource group every week and adjust the scenario weight coefficient (e.g., if the response compliance rate of the city delivery group is ≥90% for 3 consecutive days, the weight coefficient is increased from 0.9 to 0.95); update the adjustment coefficient according to the battery SOH change to ensure the adaptability of the strategy.

[0178] (3) Optimization result synchronization: The updated parameters are synchronized to each functional module to form a closed-loop management of "prediction-execution-verification-optimization", which increases the overall response compliance rate of the virtual power plant by 1%-2% per month.

[0179] Step 5: Collaborative Management of the Entire Battery Lifecycle

[0180] 1. Construction of Battery Digital Twin Profile

[0181] (1) Archive dimension: Each battery is assigned a unique ID, and a digital twin archive containing "factory parameters (rated capacity, cycle life) - charge and discharge history (cumulative number of times, depth, temperature) - SOH dynamic change - fault record - transaction participation record" is constructed based on blockchain evidence data, and the status data is updated every 15 minutes.

[0182] (2) Accurate SOH calculation: The SOH value is calculated by combining the voltage monitoring method (measuring the voltage deviation of individual cells), the internal resistance measurement method (tracking the growth trend of internal resistance), and the capacity test method (a full charge and discharge test once a month). The error is controlled within ±2%.

[0183] 2. Adaptation to phased trading strategies

[0184] (1) SOH≥80% (healthy period): Priority participation in high-frequency auxiliary services such as frequency modulation and standby, charge and discharge depth controlled at 20%-80%, and charging adopts constant current-constant voltage segmented strategy (0-80% constant current fast charging, 80%-100% constant voltage slow charging) to avoid overcharging damage.

[0185] (2) 60%≤SOH<80% (decay period): Participate in medium frequency electric energy trading such as peak shaving and valley filling, control the depth of charge and discharge at 30%-70%, and limit the discharge power to 70% of the rated power to avoid high power impact.

[0186] (3) SOH<60% (retirement period): It is converted into a backup energy storage resource and only participates in emergency power supply of the power grid (such as during fault repair period). The depth of charge and discharge is ≤50%, the charging temperature is controlled at 15-35℃, and the secondary utilization cycle is extended to 3-5 years.

[0187] 3. Full lifecycle revenue management

[0188] (1) Revenue sharing: The revenue sharing ratio is set according to the battery SOH level and the type of transaction (70% for frequency regulation service during the healthy period and 40% for standby service during the retirement period), and is automatically settled to the battery owner's account through smart contract every month.

[0189] (2) Decommissioning warning: When SOH drops to 55%, a decommissioning warning will be automatically triggered, recommending conversion to community energy storage, emergency power supply and other scenarios, to realize the full-chain value mining of "vehicle power station - backup energy storage - recycling and dismantling".

[0190] Example

[0191] A port electric truck charging and swapping trunk line project covers high-frequency battery swapping scenarios such as steel transportation and port trucks. Five intelligent bidirectional battery swapping stations have been put into operation, with a total equipment capacity of nearly 9,000 kilowatts. The project connects 500 new energy electric heavy trucks equipped with CTB-400 vehicle-storage shared batteries (each battery has a capacity of 400 kWh and an initial state of equilibrium (SOH) of 100%), and is equipped with a 10 MW distributed photovoltaic power station. The project participates in the Shanghai electricity spot market and demand response transactions.

[0192] Implementation process:

[0193] 1. Data Acquisition and Preprocessing (T-1 day 00:00-08:00)

[0194] (1) Collect GPS location data of 500 trucks (320 trucks staying in the port and 180 trucks for trunk transportation), TMS plan (the peak operation period is from 8:00 to 18:00 the next day, each truck stops 4 times for loading and unloading, each time for 30 minutes), and BMS data (average SOC 65%, SOH ≥ 80% account for 85%).

[0195] (2) Collect real-time load data (average 2200kW) and photovoltaic output forecast data (peak 8MW the next day, occurring between 12:00 and 14:00) of 5 battery swapping stations, as well as time-of-use electricity prices in Shanghai (0.38 yuan / kWh during off-peak hours, 1.08 yuan / kWh during peak hours, and 0.5 yuan / kWh for peak shaving compensation).

[0196] (3) After processing with Kalman filtering and the Laida criterion, a standardized dataset with 96 time points was generated, and the proportion of outlier data decreased from 3.2% to 0.5%.

[0197] 2. Layered resource aggregation (08:00-09:00 on T-1 day)

[0198] (1) Level 1 aggregation: Adjustable capacity of a single truck = 400 × 0.65 × 0.8 × 1.0 = 208 kWh (stationary state) or 400 × 0.65 × 0.8 × 0.1 = 20.8 kWh (driving state), total adjustable capacity of 5 battery swapping stations = (9000 - 1350) - 1350 = 6300 kW (1350 kW is the minimum load);

[0199] (2) Secondary aggregation: classified into port truck group (320 stationary trucks, aggregation capacity of 66,560 kWh, response speed of 3 minutes) and trunk transportation group (180 traveling trucks, aggregation capacity of 3,744 kWh, response speed of 5 minutes).

[0200] (3) Three-level aggregation: The total adjustable capacity of the virtual power plant is 70.3MWh, the average response delay is 3.5 minutes, and the capacity with SOH≥80% accounts for 82%.

[0201] 3. Dynamic transaction matching (09:00-17:00 on T-1 day)

[0202] (1) GA-LSTM forecast: The previous forecast for the next day's peak electricity demand (9:00-11:00, 14:00-17:00) is 35MWh, and the off-peak electricity demand (0:00-6:00) is 40MWh; the intraday rolling forecast at 10:00 has been revised to 38MWh due to the earlier port operations;

[0203] (2) Smart contract matching: During peak periods, priority is given to matching port container truck groups, with a unit price of RMB 0.97 / kWh for matching 35MWh of grid peak shaving transactions; during peak photovoltaic output periods (12:00-14:00), 12MWh of photovoltaic surplus electricity is matched at a unit price of RMB 0.42 / kWh.

[0204] (3) Conflict detection: Two duplicate matching transaction orders were removed, and 12 valid orders were generated in the end, involving a transaction electricity volume of 47MWh.

[0205] 4. Reliable Execution and Feedback (T-day 00:00-24:00)

[0206] (1) Off-peak period 0:00-6:00: The battery swapping station executes the power purchase order and prioritizes charging 300 batteries with SOH≥80%. The charging power is controlled at 800kW / station. The actual charging amount is 39.8MWh, which is 0.5% different from the prediction.

[0207] (2) Peak period 9:00-11:00: 320 container trucks discharged during loading and unloading intervals, each truck discharged 30kWh per time, with a cumulative discharge of 38.4MWh. Edge nodes were monitored every 5 seconds, and the deviation rate was <3%.

[0208] (3) Data storage: Data is encrypted and uploaded to the consortium blockchain every 5 minutes. The zero-knowledge proof verification is 100% successful, generating 144 traceability QR codes. The keys are jointly managed by Shanghai Power Grid, Qiyuan Core Power and Shanghai Energy Regulatory Bureau.

[0209] 5. Battery lifecycle management (continuous execution)

[0210] (1) Digital twin archives: Real-time updates of charge and discharge data for each battery. Monthly SOH testing shows that the internal resistance growth rate of batteries using the differentiated strategy is reduced by 28% compared to the traditional method;

[0211] (2) Strategy adjustment: 12 batteries with SOH reduced to 79% will be automatically converted to peak shaving, and the depth of charge and discharge will be reduced from 80% to 70%, with the expected secondary utilization cycle extended to 4 years.

[0212] Implementation results:

[0213] 1. Resource utilization rate: The adjustable resource utilization rate of trucks increased from 72% to 98%, the adjustable capacity of battery swapping stations increased from 4.5MW to 6.3MW, and the photovoltaic absorption rate increased from 85% to 99%;

[0214] 2. Response speed: Transaction response delay was reduced from 22 minutes to 1.8 minutes, and the power grid load fluctuation response deviation rate was controlled at 4.2%;

[0215] 3. Security performance: No data leakage occurred throughout the process, and three attempts to tamper with the transaction were all intercepted by the blockchain traceability system;

[0216] 4. Economic benefits: Monthly transaction revenue reached 286,000 yuan, an increase of 102,000 yuan compared to the traditional model; battery maintenance costs decreased by 18.5%;

[0217] 5. Environmental benefits: Reduces carbon emissions by 980 tons per month, equivalent to planting 54,000 trees.

[0218] The above description constitutes an embodiment of the present invention. The foregoing descriptions are preferred embodiments of the present invention. Unless there is a clear contradiction or a prerequisite for a particular preferred embodiment, the preferred embodiments can be arbitrarily combined and used. The embodiments and specific parameters described are merely for clearly illustrating the verification process of the invention and are not intended to limit the scope of patent protection of the present invention. The scope of patent protection of the present invention is still determined by its claims. Similarly, any equivalent structural changes made based on the content of the present invention's specification should also be included within the scope of protection of the present invention.

Claims

1. A virtual power plant trading method based on an energy exchange integration scenario, characterized in that, Includes the following steps: (1) Data fusion acquisition and preprocessing of transportation and energy: (1.1) Accurate collection of multi-source data: Collection of traffic data, energy data, and environmental data; The specific steps for collecting traffic-related data are as follows: The vehicle-mounted terminal collects real-time GPS location and speed information of electric vehicles at a frequency of 1Hz; the trip is planned synchronously from the TMS transportation management system, including departure time, destination, and estimated stay duration. The battery management system collects battery status data every 30 seconds, including SOH data, SOC data, single-cell voltage equalization, real-time charge / discharge power, and temperature. The specific steps for collecting energy-related data are as follows: The edge intelligent controller collects real-time load, remaining battery quantity and status of the charging and swapping station every 15 seconds; obtains 15-minute output forecast data, time-of-use electricity price and ancillary service price of distributed photovoltaic / wind power from the power grid dispatch system; and obtains time-of-use metering of traded electricity through smart meters. The collection of environmental data specifically involves: Sensors deployed at charging and battery swapping stations and transportation trunk lines collect real-time road conditions, wind speed, and light intensity every 5 minutes to correct energy forecasting errors. (1.2) Data preprocessing and fusion: The collected traffic data, energy data and environmental data are preprocessed, and the traffic, energy and environmental data are associated and matched by spatiotemporal alignment algorithm to generate a standardized dataset; (2) Hierarchical resource aggregation and status assessment: (2.1) Level 1 aggregation: Single resource layer aggregation, to assess the adjustable capacity and response capability of a single electric vehicle or a single charging and battery swapping station; (2.2) Second-level aggregation: Scene resource layer aggregation, which is classified and collaborative capability assessed according to application scenarios; The scenario classification criteria are divided into four categories: port container truck group, urban delivery group, community charging and swapping group, and trunk transportation group, with preset weight coefficients for each category. Specifically, it includes: (2.2.1) Calculation of aggregation parameters: Scene aggregation capacity = Σ (adjustable capacity of individual resources × scene weight coefficient); Scene response speed = Σ (individual resource response latency × resource capacity percentage); SOH Distribution in a Scenario = Statistics on the number and capacity percentage of batteries with different SOH levels in this scenario; (2.2.2) Grouping constraints: The aggregated capacity of each scenario resource group must be ≥1MW, and the individual resources must be located at the same spot market clearing node to ensure the execution of scheduling instructions; (2.3) Three-level aggregation: aggregation at the virtual power plant level to generate a global resource profile; Summarize the aggregated capacity, response speed, and SOH distribution parameters of all scenario resource groups, and calculate the total adjustable capacity, average response delay, and peak-valley regulation potential indicators of the virtual power plant; Construct a resource profile tagging system, including five dimensions: total capacity, response level, SOH structure, predicted availability period, and geographical distribution. Each tag corresponds to quantitative data, providing a basis for decision-making in transaction matching. (3) Dynamic transaction demand forecasting and matching: (3.1) GA-LSTM dual-timescale demand forecasting; (3.2) Smart contract multi-target transaction matching; Specifically, it includes: (3.2.1) Contract preset rules: Based on local power trading rules, clarify the terms of trading types, price ranges, and settlement methods, and distinguish between load-type and generation-type virtual power plant identities; (3.2.2) Matching objective function: With the fastest response speed, the lowest transaction cost, and the minimum battery loss as multiple objectives, the weighted summation method is used to transform it into a single objective optimization; (3.2.3) Classification matching logic: Electricity sales scenario: Prioritize matching resource groups with response speed ≤ 5 minutes and SOH 60%-80%. The screening criteria are transaction unit price ≤ grid peak-valley price difference × 0.

9. The counterparty is determined by sorting the response speed in ascending order and the price in descending order. Electricity purchase scenario: Prioritize matching charging and battery swapping stations in areas with abundant photovoltaic output and SOH ≥ 80%. The screening criteria are that the transaction unit price is ≥ the grid off-peak electricity price × 1.1, and the stations are sorted in ascending order of price and descending order of capacity. Ancillary service sites: Only resource groups with a response speed of ≤2 minutes and SOH ≥85% are matched, and the charging and discharging power fluctuation range is required to be ≤10% / minute, sorted by adjustment accuracy; (3.2.4) Conflict detection and arbitration: The smart contract automatically detects transaction conflicts, including double-spending conflicts and validity conflicts. It arbitrates conflicting transactions according to the principles of time priority and capacity priority, removes invalid transactions and records them in the log. (3.3) Transaction plan generation and confirmation, specifically: (3.3.1) Generate a standardized transaction order containing transaction ID, pseudonym of the participant, time period, electricity, price, response requirements, and liability for breach of contract. The participant's identity is identified by a pseudonym generated using zero-knowledge proof, which is bound to the real identity and stored in the consortium blockchain. (3.3.2) The order is synchronized to all blockchain nodes for pre-deposit. After the participants confirm through the edge terminal, the order officially takes effect and locks the adjustable capacity of the corresponding resources. The lock duration is the transaction period plus a 30-minute buffer period. (4) Optimization of trusted execution and feedback in blockchain: (4.1) Precise execution at edge nodes; Specifically, it includes: (4.1.1) Instruction issuance: The blockchain consortium chain parses the effective orders into charging and discharging instructions, and issues them to the resource group in real time through the edge smart controller. The instructions include start time, end time, power curve, and safety threshold parameters. (4.1.2) Differentiated execution strategy: Electric vehicles: An intermittent charging and discharging strategy is adopted without affecting the trip; if the trip changes, the vehicle terminal will automatically send a change request to the platform, triggering order adjustment. Charging and battery swapping stations: adopt a tiered strategy of charging first and then discharging. During off-peak hours, priority is given to charging batteries with SOH ≥ 80%, and during peak hours, priority is given to discharging batteries with SOH 60%-80%. At the same time, 20% capacity of backup batteries is reserved to cope with sudden demand. Execution monitoring: The edge node collects actual charging and discharging data every 5 seconds and compares it with the command curve. If the deviation exceeds 5%, local adjustment is initiated. If it exceeds 10%, the data is immediately reported to the platform. (4.2) Trusted storage and verification of blockchain data; (4.3) Closed-loop optimization iteration; (5) Collaborative management of the entire battery lifecycle: (5.1) Construction of digital twin archives for batteries; (5.2) Adaptation of phased trading strategies; Specifically, it includes: (5.2.1) Health period with SOH≥80%: Priority participation in frequency regulation and backup auxiliary services, charge and discharge depth controlled between 20%-80%, and constant current-constant voltage segmented charging strategy to avoid overcharge damage; (5.2.2) Decay period of 60%≤SOH<80%: For medium frequency electric energy trading that participates in peak shaving and valley filling, the depth of charge and discharge should be controlled at 30%-70%, and the discharge power should be limited to 70% of the rated power to avoid high power impact; (5.2.3) Retirement period with SOH < 60%: It is converted into a backup energy storage resource, only participating in emergency power supply to the grid, with a charge and discharge depth ≤ 50%, and the charging temperature controlled at 15-35℃, extending the secondary utilization cycle to 3-5 years; (5.3) Full life cycle revenue management.

2. The virtual power plant trading method based on an energy exchange integration scenario as described in claim 1, characterized in that, In step (1.1), the interface standard is specifically as follows: Traffic data is transmitted using the J1939 protocol, and energy data is transmitted using the DL / T 645 protocol. All data is uploaded to the traffic and energy cloud platform via 4G / 5G network. The specific steps (1.2) are as follows: The Kalman filter algorithm is used to smooth the vehicle GPS data. The state equation is set as x(k)=A x(k-1)+w(k) and the observation equation is set as z(k)=H x(k)+v(k), where the process noise w(k) and the observation noise v(k) both follow a Gaussian distribution. The Raida criterion is used to remove outliers from the load data of charging and swapping stations. The mean μ and standard deviation σ of historical load data are calculated. Data that exceeds the interval [μ-3σ, μ+3σ] are marked as outliers and are filled in using linear interpolation. By using a spatiotemporal alignment algorithm, traffic, energy, and environmental data are correlated and matched to generate a standardized dataset with a time granularity of 15 minutes.

3. A virtual power plant trading method based on an energy exchange integration scenario as described in claim 1, characterized in that, Step (2.1) specifically includes: (2.1.1) Calculation of adjustable capacity for electric vehicles: Adjustable capacity = Rated battery capacity × SOC × Adjustment coefficient × Traffic availability coefficient; Among them: the adjustment coefficient is dynamically set according to the SOH; the traffic availability coefficient is calculated based on the trip plan; (2.1.2) Calculation of adjustable capacity of charging and battery swapping stations: Adjustable capacity = (maximum charging power - current charging power) - minimum load power; Among them: the minimum load power is the minimum power required to maintain the basic operation of the battery swapping station; the number of dispatchable batteries is also calculated. (2.1.3) Response capability marker: Record the response latency and continuous response duration parameters of individual resources as indicators for transaction matching.

4. A virtual power plant trading method based on an energy exchange integration scenario as described in claim 1, characterized in that, Step (3.1) specifically includes: (3.1.1) Model construction: The LSTM network optimized by the genetic algorithm is used. The genetic algorithm is used to optimize the number of hidden layer units, time step, dropout ratio and learning rate of the LSTM. The encoding method is a 10-bit integer chromosome. The fitness function adopts the validation set MSE. (3.1.2) Day-ahead forecast: Input historical transaction data of the past 7 days, traffic plan for the next day, and weather forecast data, and output the basic transaction electricity demand at 96 time points within 24 hours; (3.1.3) Intraday Rolling Forecast: By integrating real-time traffic conditions, grid load fluctuation data, and dynamic data on battery status changes, the day-ahead forecast results are corrected using the following formula: Corrected demand = day-ahead forecast × (1 + grid load fluctuation coefficient × 0.3 + traffic delay coefficient × 0.7). Where: Power grid load fluctuation coefficient = (current load - predicted load) / predicted load, traffic delay coefficient = (actual travel time - planned travel time) / planned travel time; (3.1.4) Prediction result output: Generate a demand curve that includes time period, predicted power consumption, confidence interval, and adjustment trigger threshold. When the actual deviation exceeds the threshold, the prediction will be automatically restarted.

5. A virtual power plant trading method based on an energy exchange integration scenario as described in claim 1, characterized in that, Step (4.2) specifically includes: (4.2.1) Encrypted data upload: Every 5 minutes, the actual charging and discharging data, battery status data, and vehicle location data are double-encrypted and uploaded to the consortium blockchain after generating ciphertext data; (4.2.2) Zero-knowledge proof verification: A polynomial equation is constructed using FO commitment to verify the authenticity of the uploaded data. If the verification passes, the data is stored; if it fails, the data is destroyed and a source tracing warning is triggered. (4.2.3) Evidence storage content: including encrypted transaction data, verification results, participant signatures, timestamps, generating a unique QR code, and linking transaction location information to form a traceability information package to ensure that the data is tamper-proof and traceable; (4.2.4) Key management: Paillier's private key is stored in layers across three trusted institutions through the Shamir secret sharing mechanism. At least two institutions must work together to decrypt it to prevent key leakage.

6. A virtual power plant trading method based on an energy exchange integration scenario as described in claim 1, characterized in that, Step (4.3) specifically includes: (4.3.1) Daily optimization: After the daily transaction ends, extract the full amount of data stored in the blockchain, calculate the response deviation rate and battery loss rate of each resource group, update the training dataset of the GA-LSTM model, and re-optimize the model parameters. (4.3.2) Weekly optimization: Analyze the collaborative efficiency of the scenario resource group every week and adjust the scenario weight coefficient; update the adjustment coefficient according to the battery SOH change to ensure the adaptability of the strategy; (4.3.3) Optimization result synchronization: The updated parameters are synchronized to each functional module to form a closed-loop management of prediction, execution, verification and optimization.

7. A virtual power plant trading method based on an energy exchange integration scenario as described in claim 1, characterized in that, Step (5.1) specifically includes: (5.1.1) Archive dimension: Assign a unique ID to each battery, and build a digital twin archive based on blockchain-stored data, which includes factory parameters, charging and discharging history, dynamic changes in SOH, fault records, and transaction participation records, and update the status data every 15 minutes; (5.1.2) Accurate SOH calculation: The SOH value is calculated by integrating voltage monitoring method, internal resistance measurement method and capacity testing method and using a weighted algorithm; Step (5.3) specifically includes: (5.3.1) Revenue sharing: The revenue sharing ratio is set according to the battery SOH level and the type of transaction, and is automatically settled to the battery owner's account through smart contract every month; (5.3.2) Decommissioning warning: When the SOH drops to 55%, a decommissioning warning will be automatically triggered, and it is recommended to switch to community energy storage or emergency power supply scenarios.