Construction method of source network load storage regulation and control strategy simulation verification platform

Abstract

The invention discloses a construction method of a source network load storage regulation and control strategy simulation verification platform, and relates to the technical field of source network load storage regulation and control. The problems existing in the prior art are solved. The method specifically comprises the following steps: demand analysis and target setting; establishing a source network load storage dynamic model; building a heterogeneous computing platform; designing a layered regulation and control strategy; constructing a multi-scene test environment; verifying strategy validity through block chain consensus, and dynamically evaluating a toughness index in combination with digital twinning; and generating a carbon energy joint optimization report. The FPGA is adopted to accelerate electromagnetic transient simulation, the calculation efficiency is improved, block chain three-level verification (provincial-local-station) + digital twinning dynamic coupling is adopted, strategy dynamic evaluation and credible execution are achieved, second-level strategy verification and dynamic adjustment are guaranteed, layered regulation and control strategies are executed, the layered regulation and control strategies comprise second-level frequency modulation, minute-level AGC and hour-level economic dispatching, and the method is suitable for large-scale popularization and application. And finally, a carbon energy joint optimization report is generated, and the emission reduction benefit and the economic benefit of the strategy are quantified.

Description

Technical Field

[0001] This invention relates to the field of source-grid-load-storage regulation technology, and in particular to a method for constructing a simulation and verification platform for source-grid-load-storage regulation strategies. Background Technology

[0002] The intermittency and volatility of renewable energy pose three major challenges to the power grid: 1) Frequency stability problem: The high proportion of wind and solar power generation leads to a decrease in system inertia and a significant increase in frequency regulation pressure; 2) Carbon emission control problem: Traditional economic dispatch does not take into account carbon flow distribution, and regional carbon intensity deviation can reach up to 40%; 3) Extreme disaster risk: Typhoons can easily cause power grid tower collapses, resulting in load loss.

[0003] However, the current mainstream methods for verifying control strategies in existing technologies have obvious shortcomings: (1) Limitations of simulation platform: Traditional electromechanical-electromagnetic transient hybrid simulation (such as RTDS+PSASP) has the following problems: time scale is fragmented, and it is impossible to achieve collaborative verification of second-level frequency regulation and hour-level scheduling; data silo problem, a case of a regional power grid shows that the data delay between SCADA, EMS and meteorological system is 3-5 minutes; lack of realism verification, a virtual power plant strategy performed well in simulation, but the control failed due to communication delay after actual deployment.

[0004] (2) Insufficient application of blockchain: The existing blockchain verification scheme only realizes single-level verification: the provincial chain verification delay is >2 seconds, which cannot meet the AGC control requirements; and no device-level trusted execution environment has been established. A certain energy storage power station once caused a false SOC alarm due to the controller being tampered with.

[0005] Against this backdrop, the present invention proposes a method for constructing a simulation verification platform for source-grid-load-storage regulation strategy, addressing the core problems of existing technologies such as fragmented verification, data silos, and insufficient disaster response. Summary of the Invention

[0006] The purpose of this invention is to address the shortcomings of existing technologies by proposing a method for constructing a simulation and verification platform for source-grid-load-storage regulation strategies.

[0007] To achieve the above objectives, the present invention adopts the following technical solution: A method for constructing a simulation and verification platform for source-grid-load-storage regulation strategy includes the following steps: S1: Requirements Analysis and Goal Setting; S2: Establish a dynamic model of source-grid-load-storage; S3: Build a heterogeneous computing platform; S4: Design a tiered control strategy; S5: Build a multi-scenario testing environment; S6: Verify the effectiveness of the strategy through blockchain consensus and dynamically assess resilience indicators using digital twins; S7: Generate a joint carbon energy optimization report; The sources of the demand include historical operating data from the power grid dispatch center, policy documents, and user-side data; the target setting is specifically a quantitative target, including technical indicators, carbon emission indicators, and resilience indicators. The source-grid-load-storage dynamic model includes a thermal power / wind and solar power unit model, a transmission network model, a load response model, and an energy storage converter model; The heterogeneous computing platform comprises a data interface layer, a computing engine layer, and a policy verification layer. The data interface layer supports IEC 61850 and Modbus protocols; The computing engine layer uses FPGA to accelerate electromagnetic transient simulation; The strategy verification layer integrates a blockchain multi-level verification module and a digital twin inference module.

[0008] Preferably, the implementation of the blockchain multi-level verification module includes the following aspects: 1) Provincial-level blockchain verification system; The node hardware configuration includes: ① a server cluster equipped with Intel Xeon Scalable processors, with each node configured with SGX enclave memory ≥256MB; ② a secure communication protocol based on the national cryptographic SM9 algorithm with end-to-end encryption and a key update cycle of ≤15 minutes. Additionally, the verification logic includes: ① Calculating power volatility: Among them, P rated The rated power of the tie line is used to trigger a strategy review when the R_ fluctuation at three consecutive sampling points is greater than 5% / min. ② Consensus mechanism: Improved HotStuff algorithm, with view switching timeout set to 3 seconds; 2) Local chain verification system; Among them, the node security module includes: ① The distribution network RTU equipment integrates SM2-SM3 dual algorithm chips, with a signature speed of ≥1000 times / second; ② Lightweight zero-knowledge proof is performed before data is uploaded to the chain, with a proof time of ≤10ms; Additionally, voltage compliance rate assessment: ① Sampling standards: At least 3 monitoring points must be deployed in each area, namely the beginning, middle and end points, with a sampling interval of ≤1 second; ② Qualification criteria: The duration of voltage deviation within ±7% accounts for ≥99.7% of the total duration. 3) Station-level chain verification system; Among them, real-time performance is guaranteed: the energy storage PCS controller adopts a time-triggered architecture with a clock synchronization accuracy of ≤1μs; Another option is a trusted execution environment: ① The TEE chip must be certified by CC EAL5+ and support dynamic trust measurement; ② Critical instructions must be decrypted and executed within the enclave. 4) Cross-chain interaction protocol; The data format adopts the IEEE 2030.5 standard extended fields; In addition, the conflict resolution uses the Shapley value algorithm based on game theory to assign verification weights.

[0009] Preferably, the digital twin inference module includes a data input layer, a physical field modeling layer, a strategy inference layer, and a visualization and interaction layer; The data input layer is used to access meteorological data and monitor equipment status. The meteorological data includes typhoon data and ice storm data. Specifically: ① Real-time access to ECMWF typhoon forecast data, with a spatial resolution of 0.1°×0.1° and a temporal resolution of 1 hour, including central pressure, movement path, and radius of the 7-level wind circle; ② Access to freezing rain forecast data from the Central Meteorological Observatory, with a spatial resolution of 5km×5km, including ambient temperature, precipitation intensity, and duration. In addition, the equipment status is specifically obtained through the PHM system to obtain the remaining life of the transformer, and the monitoring parameters include dissolved gas in oil, winding hot spot temperature, and partial discharge.

[0010] Preferably, the physical field modeling layer includes typhoon scene modeling and ice storm scene modeling; wherein, the typhoon scene modeling includes: ① Wind field reconstruction: using CFD fluid dynamics simulation, with a computational domain size of 20km×20km×1km and a mesh size ≤50m; ② Tower failure criteria: ; In addition, the content of ice disaster scenario modeling includes: ① Icing growth model: ② Insulator flashover criterion: Flashover warning is triggered when the ice thickness is ≥15mm and the humidity is >90%.

[0011] Preferably, the strategy deduction layer is a dynamic contingency plan generation layer, which includes the generation of typhoon scenarios and ice storm scenarios. Specifically, ① when the typhoon scenario is generated: when P fail Automatic startup when >30%: Contingency Plan A: Reduce the output of wind farms upstream of the typhoon's path to 50% of their rated capacity; Contingency Plan B: Pre-set tree trimming instructions for the power transmission corridor, with tree trimming height ≥ conductor-to-ground distance + 3m; ② When generating an ice disaster scenario: Automatically trigger when the predicted ice thickness is >20mm: Contingency Plan C: Activate the DC de-icing device with a current density ≥3A / mm²; Contingency Plan D: Switch to the backup line.

[0012] Preferably, the visualization interaction layer includes a 3D geographic information system display and a risk level output. The content displayed by the 3D geographic information system includes the typhoon path prediction error zone and the tower failure heat map; the comprehensive risk index in the risk level output is R=0.6×Pfail+0.4×Flashover_prob.

[0013] Preferably, the design of the hierarchical control strategy includes second-level frequency regulation, minute-level AGC, and hour-level economic scheduling; The second-level frequency adjustment adopts an event-triggered distributed consensus algorithm, and the response time of the second-level frequency adjustment is ≤200ms from the frequency exceeding the limit to the command being issued. The minute-level AGC uses an improved distributed ADMM algorithm: ,in, Let P be the power generation cost function of the i-th generating unit; i The active power output of the i-th unit is MW; λ i Let z be the Lagrange multiplier vector; z be the global consistency variable; ρ be the penalty factor; the convergence condition of the minute-level AGC is that the difference in power allocation between two adjacent iterations is <1MW; The hourly economic scheduling model considers carbon constraints and is a stochastic optimization model. ; ; in, Let γ be the power generation cost of unit g during time period t, in yuan; γ is the carbon price weighting coefficient, in yuan / kgCO2; The bus carbon intensity during time period t is expressed in kgCO2 / kWh. Let g be the output of unit g during time period t, in MW; Let g be the output of unit g during time period t, in MW; Reserve capacity for time period t; The hourly economic scheduling scenario reduction uses K-means clustering to reduce the number of wind and solar power output scenarios from 1000 to 10 representative scenarios.

[0014] Preferably, the test environment covers normal weather, extreme weather, and network attack scenarios; The extreme weather scenarios include typhoon scenarios and ice storm scenarios. In the typhoon scenario, the wind field model adopts the Jensen wake model, and the turbulence intensity is set to IEC Class A. Tower failure: when the wind speed is >42m / s, the failure probability is calculated according to the Weibull distribution. In the ice storm scenario, the icing growth adopts the Makkonen model, the ambient temperature is -5℃~0℃, and the liquid water content is 0.1~0.3g / m³. The network attack scenario includes a False Data Injection attack and defense verification. The False Data Injection attack targets the SCADA system state estimation module; its attack magnitude is defined as voltage amplitude tampering ≤10% and phase angle ≤5°. The defense verification employs a χ² detector with the following detection threshold settings: ; Where α = 0.01, m is the number of measurements, and n is the dimension of the state variables.

[0015] Preferred criteria: The blockchain verification metrics include ① Strategy validity consensus condition: ≥2 / 3 of the nodes must confirm; ② Data upload format: Protobuf encoding is used, including timestamp, data hash, and signature; The resilience assessment of the digital twin includes assessment indicators and dynamic simulation. The assessment indicators include ① Load Guarantee Rate (LSR) = ∑ duration of critical load power supply / total duration; ② System Recovery Time (TRT): the time from the occurrence of a fault to 90% load recovery. The dynamic simulation includes ① using an RTDS real-time simulator with a step size ≤ 50 μs; ② real-time interaction with meteorological data API with an update frequency ≥ 1 time / minute.

[0016] Preferably, the template for the joint carbon energy optimization report includes the following: a. Cover: Scene name, timestamp, strategy version number; b. Abstract: Total carbon emissions, percentage of renewable energy, and radar charts of key indicators; c. Details: Unit output curve, energy storage SOC, carbon emission intensity thermogram; d. Recommendation: Optimization suggestions for the next scheduling cycle.

[0017] The beneficial effects of this invention are as follows: 1. This invention first determines the system stability, carbon emission constraints, and economic objectives. Then, it establishes a dynamic model of source-grid-load-storage and builds a heterogeneous computing platform to support real-time access to multi-source data. It uses FPGA to accelerate electromagnetic transient simulation, improving computational efficiency. It employs blockchain three-level verification (province-region-station) + digital twin dynamic coupling to achieve dynamic evaluation and reliable execution of strategies, ensuring second-level strategy verification and dynamic adjustment. It executes hierarchical control strategies, including second-level frequency regulation, minute-level AGC, and hourly economic dispatch. Finally, it generates a carbon energy joint optimization report to quantify the emission reduction and economic benefits of the strategies.

[0018] 2. In practical applications, this invention shortens response time in typhoon scenarios, improves load guarantee rate (LSR) under extreme weather conditions, optimizes system recovery time (TRT), and effectively enhances grid resilience. Wind curtailment rate and carbon intensity decrease, effectively reducing operating costs. Attached Figure Description

[0019] Figure 1 This is a flowchart illustrating the construction method of a simulation verification platform for a source-grid-load-storage regulation strategy proposed in this invention. Detailed Implementation

[0020] The technical solution of this patent will be further described in detail below with reference to specific embodiments.

[0021] The embodiments of this patent are described in detail below. Examples of these embodiments are shown in the accompanying drawings, wherein the same or similar reference numerals denote the same or similar elements or elements having the same or similar functions throughout. The embodiments described below with reference to the accompanying drawings are exemplary and are only used to explain this patent, and should not be construed as limiting this patent.

[0022] Example 1: A method for constructing a simulation and verification platform for source-grid-load-storage regulation strategy, such as... Figure 1 As shown, it includes the following steps: S1: Demand analysis and target setting; determine system stability, carbon emission constraints and economic targets.

[0023] Furthermore, the sources of demand include historical operational data from the power grid dispatch center, policy documents, and user-side data; Preferably, the historical operating data of the power grid dispatch center includes frequency deviation, load fluctuation, and wind and solar curtailment rates; policy documents include "dual carbon" targets and the requirement that the regional power grid renewable energy penetration rate be ≥50%; and user-side requirements include industrial parks for power supply reliability of ≥99.99%.

[0024] Furthermore, the targets are set in a quantitative manner, including technical indicators, carbon emission indicators, and resilience indicators. Preferably, the technical indicators include frequency deviation ≤ ±0.1Hz, wind curtailment rate ≤ 5%, and energy storage SOC fluctuation ≤ 10%; carbon emission indicators include carbon intensity per unit of electricity ≤ 0.3kgCO2 / kWh; and resilience indicators include critical load guarantee rate ≥ 95% under extreme weather conditions and recovery time ≤ 2h.

[0025] S2: Establish a dynamic model of source-grid-load-storage; Furthermore, the source-grid-load-storage dynamic model includes a thermal power / wind and solar power unit model, a transmission network model, a load response model, and an energy storage converter model.

[0026] Furthermore, the parameter calibration method of the source-grid-load-storage dynamic model is as follows: the thermal power unit model uses the least squares method to fit the actual coal consumption curve; the wind and solar power unit model is based on the historical output data of actual power plants and uses the GARCH model to characterize the volatility; the load response model uses the Price Elasticity Matrix (PEM) model, and the elasticity coefficient is obtained by regression from the electricity meter data; the energy storage converter model uses a second-order RC equivalent circuit, and the parameters are identified through HPPC test data.

[0027] S3: Build a heterogeneous computing platform; Furthermore, the heterogeneous computing platform includes a data interface layer, a computing engine layer, and a policy verification layer, etc. The data interface layer supports IEC 61850 and Modbus protocols; the computing engine layer uses FPGA to accelerate electromagnetic transient simulation; and the strategy verification layer integrates a blockchain multi-level verification module and a digital twin inference module.

[0028] Furthermore, the implementation of the blockchain multi-level verification module includes the following aspects: 1) Provincial-level blockchain verification system; The node hardware configuration includes: ① a server cluster equipped with Intel Xeon Scalable processors, with each node configured with SGX enclave memory ≥256MB; ② a secure communication protocol based on the national cryptographic SM9 algorithm with end-to-end encryption and a key update cycle of ≤15 minutes.

[0029] Additionally, the verification logic includes: ① Calculating power volatility: ;in, The rated power of the tie line is used to trigger a strategy review when the R_ fluctuation at three consecutive sampling points is greater than 5% / min. ② Consensus mechanism: Improved HotStuff algorithm, with a view switching timeout set to 3 seconds.

[0030] 2) Local chain verification system; Among them, the node security module includes: ① The distribution network RTU equipment integrates SM2-SM3 dual algorithm chips, with a signature speed of ≥1000 times / second; ② Lightweight zero-knowledge proofs (zk-SNARKs) are performed before data is uploaded to the blockchain, with a proof time of ≤10ms; In addition, voltage qualification rate assessment: ① Sampling standard: At least 3 monitoring points (first end, middle end, and last end) are deployed in each area, with a sampling interval of ≤1 second; ② Qualification judgment: The percentage of voltage deviation within ±7% (220V standard) for a continuous duration is ≥99.7%.

[0031] 3) Station-level chain verification system; Among them, real-time performance is guaranteed: the energy storage PCS controller adopts a time-triggered architecture (TTEthernet) with a clock synchronization accuracy of ≤1μs; Another option is a trusted execution environment: ① The TEE chip must be CC EAL5+ certified and support Dynamic Trust Metric (DRTM); ② Critical instructions (such as charge / discharge start / stop) must be decrypted and executed within the enclave. This achieves multi-timescale collaborative verification ranging from seconds (≤200ms) to hours, solving the problem of high latency (>2 seconds) in traditional single-level verification.

[0032] 4) Cross-chain interaction protocol; The data format adopts the IEEE 2030.5 standard extended fields; In addition, the conflict resolution uses the Shapley value algorithm based on game theory to assign verification weights.

[0033] Furthermore, the digital twin simulation module includes a data input layer, a physical field modeling layer, a strategy simulation layer, and a visualization interaction layer; it can access ECMWF typhoon data (0.1° resolution) and PHM equipment status data in real time, and combine CFD fluid simulation (mesh size ≤50m) to realize dynamic calculation of tower failure probability (automatic triggering of contingency plan when Pfail>30%).

[0034] The data input layer is used to access meteorological data and monitor the status of equipment. Preferably, the meteorological data includes typhoon data and ice storm data, specifically: ① real-time access to ECMWF typhoon forecast data, with a spatial resolution of 0.1°×0.1° and a temporal resolution of 1 hour, including central pressure, movement path and radius of the 7-level wind circle; ② access to freezing rain forecast data from the China Meteorological Administration, with a spatial resolution of 5km×5km, including ambient temperature, precipitation intensity and duration.

[0035] In a further preferred embodiment, the equipment status is specifically obtained through the PHM system to obtain the remaining life of the transformer, and the monitoring parameters include dissolved gas in oil (H2≥150ppm warning), winding hot spot temperature (θh≥105℃ warning) and partial discharge quantity (≥20pC / pulse warning).

[0036] The physical field modeling layer includes typhoon scene modeling and ice storm scene modeling; Preferably, the typhoon scenario modeling includes: ① Wind field reconstruction: using CFD fluid dynamics simulation, with a computational domain size of 20km×20km×1km and a mesh size ≤50m; ② Tower failure criteria: ; In this context, For wind load, N); is the ultimate bearing capacity of the tower (18kN for a typical 110kV tower); k is the shape parameter.

[0037] Further preferred aspects of ice storm scenario modeling include: ① Icing growth model: ② Insulator flashover criterion: Flashover warning is triggered when the ice thickness is ≥15mm and the humidity is >90%.

[0038] Here, α is the collision coefficient (taken as 0.8); I(t) is the precipitation intensity (mm / h); and ρice is the ice density (taken as 0.9 g / cm³).

[0039] The strategy deduction layer is a dynamic contingency plan generation layer, which includes the generation of typhoon scenarios and ice storm scenarios. Specifically, ① when a typhoon scenario is generated: when... Automatic startup when >30%: Contingency Plan A: Reduce the output of wind farms upstream of the typhoon's path to 50% of their rated capacity; Contingency Plan B: Pre-set tree trimming instructions for the power transmission corridor; Preferably, the tree trimming height is ≥ the distance of the conductor to the ground + 3m.

[0040] ② When generating an ice disaster scenario: Automatically trigger when the predicted ice thickness is >20mm: Contingency Plan C: Activate the DC de-icing device with a current density ≥3A / mm²; Contingency Plan D: Switch to backup line (N-2 verification passed).

[0041] The visualization and interaction layer includes a 3D geographic information system display and risk level output; Preferably, the content displayed by the 3D geographic information system includes the typhoon path prediction error zone (±50km confidence interval) and the tower failure heat map (red: Pfail>50%, yellow: 30-50%, green: <30%). Preferably, the comprehensive risk index R in the risk level output is 0.6×Pfail+0.4×Flashover_prob.

[0042] S4: Design a tiered control strategy; Furthermore, the design of the hierarchical control strategy includes second-level frequency regulation, minute-level AGC, and hour-level economic scheduling; The second-level frequency adjustment employs an event-triggered distributed consensus algorithm. Preferably, the response time for second-level frequency modulation is ≤200ms from the time the frequency exceeds the limit to the time the command is issued.

[0043] The minute-level AGC employs an improved distributed ADMM algorithm: ; In this context, Let P be the power generation cost function of the i-th generating unit; i The active power output of the i-th unit is MW; λ i Let z be the Lagrange multiplier vector; z be the global consistency variable; and ρ be the penalty factor.

[0044] Preferably, the convergence condition for the minute-level AGC is that the power allocation difference between two adjacent iterations is <1MW.

[0045] The hourly economic scheduling considers a stochastic optimization model with carbon constraints: ; ; In this context, Let γ be the power generation cost of unit g during time period t, in yuan; γ is the carbon price weighting coefficient, in yuan / kgCO2; The bus carbon intensity during time period t is expressed in kgCO2 / kWh. Let g be the output of unit g during time period t, in MW; Let g be the output of unit g during time period t, in MW; This represents the reserve capacity for time period t.

[0046] Preferably, the hourly economic scheduling scenario reduction uses K-means clustering to reduce the number of wind and solar power output scenarios from 1000 to 10 representative scenarios.

[0047] S5: Build a multi-scenario testing environment; Furthermore, the test environment covers scenarios such as normal weather, extreme weather, and cyberattacks.

[0048] S6: Verify the effectiveness of the strategy through blockchain consensus and dynamically evaluate resilience indicators using digital twins; adopt joint verification using blockchain and digital twins.

[0049] Furthermore, the blockchain verification metrics include: ① Strategy validity consensus condition: requires confirmation from ≥2 / 3 of the nodes; ② Data on-chain format: uses Protobuf encoding, including timestamp, data hash, and signature.

[0050] Furthermore, the resilience assessment of the digital twin includes assessment indicators and dynamic simulation. The assessment indicators include ① Load Guarantee Rate (LSR) = ∑ Critical Load Power Supply Duration / Total Duration; ② System Recovery Time (TRT): the time from the occurrence of a fault to 90% load recovery. The dynamic simulation includes ① using an RTDS real-time simulator with a step size ≤ 50 μs; ② real-time interaction with meteorological data API with an update frequency ≥ 1 time / minute.

[0051] S7: Generate a joint carbon energy optimization report; quantify the emission reduction and economic benefits of the strategy.

[0052] A preferred template for a joint carbon energy optimization report includes the following: a. Cover: Scene name, timestamp, strategy version number; b. Abstract: Total carbon emissions, percentage of renewable energy, and radar charts of key indicators; c. Details: Unit output curve, energy storage SOC, carbon emission intensity thermogram; d. Recommendations: Optimization suggestions for the next scheduling cycle, such as suggesting that energy storage charging be brought forward by 1 hour.

[0053] In this embodiment, the system stability, carbon emission constraints, and economic objectives are first determined. Then, a dynamic model of source-grid-load-storage is established, and a heterogeneous computing platform is built to support real-time access to multi-source data. FPGA is used to accelerate electromagnetic transient simulation and improve computing efficiency. Blockchain three-level verification (province-region-station) + digital twin dynamic coupling are adopted to realize dynamic evaluation and reliable execution of strategies, ensuring second-level strategy verification and dynamic adjustment. Layered control strategies are executed, including second-level frequency regulation, minute-level AGC, and hourly economic dispatch. Finally, a carbon energy joint optimization report is generated to quantify the emission reduction and economic benefits of the strategy.

[0054] This invention shortens response time in typhoon scenarios, improves load guarantee rate (LSR) under extreme weather conditions, optimizes system recovery time (TRT), and effectively enhances grid resilience. It also reduces wind curtailment rate and carbon intensity, effectively lowering operating costs.

[0055] Example 2: A method for constructing a simulation and verification platform for source-grid-load-storage regulation strategy, such as... Figure 1 As shown, to conduct verification tests for extreme weather scenarios, this embodiment supplements Embodiment 1 as follows: The extreme weather scenarios in the multi-scenario test environment include typhoon scenarios and ice storm scenarios. For the typhoon scenario, the wind field model adopts the Jensen wake model, and the turbulence intensity is set to IEC Class A. Tower failure: When the wind speed > 42 m / s, the failure probability is calculated according to the Weibull distribution. Additionally, for the ice storm scenario, the icing growth adopts the Makkonen model, with an ambient temperature of -5℃ to 0℃ and a liquid water content of 0.1~0.3 g / m³.

[0056] Preferably, the network attack scenario includes a False Data Injection attack and a defense verification. The False Data Injection attack targets the SCADA system state estimation module; the attack magnitude is voltage amplitude tampering ≤10% and phase angle ≤5°. The defense verification uses a χ² detector with the following detection threshold settings: ; Where α = 0.01, m is the number of measurements, and n is the dimension of the state variables.

[0057] This embodiment covers verification tests for extreme weather scenarios such as N-1 faults and network attacks (False Data Injection) during use.

[0058] Experimental Example 1: I. Platform Deployment Background: 1. Regional Overview: A power grid in a coastal province with a total installed capacity of 65GW (12GW of wind power, 8GW of photovoltaic power, 40GW of thermal power, and 5GW of energy storage). 2. Core issue: Frequent typhoons lead to increased grid failure rates, and it is necessary to verify the effectiveness of the "wind-solar-storage coordinated disaster mitigation and control strategy".

[0059] II. System Modeling: 1. Thermal power unit modeling: Parameters of a 1000MW supercritical unit: Boiler time constant T b =95s, turbine time constant T t =6.2s; Cost function: C(P)=0.038P²+126P (unit: yuan / h) 2. Wind power cluster modeling: The Jensen wake model is used to calculate the wind field changes during the passage of a typhoon: when the rated wind speed is 12 m / s, the cut-in wind speed is 3 m / s, and the wind speed at the center of the typhoon is 42 m / s, the wind speed of the downstream wind turbines decreases to 23 m / s (computation domain 20 km × 20 km).

[0060] 3. Energy storage system modeling: Lithium iron phosphate battery energy storage station: capacity 200MW / 400MWh, initial SOC 65%; converter control bandwidth 550Hz, response delay ≤45ms.

[0061] III. Strategy Validation: 1. Typhoon scenario simulation (digital twin module) ① Input data: ECMWF typhoon track forecast (28.5°N ± 0.1°N, central pressure 935hPa); Tower parameters: 110kV double-circuit line, F cr =18kN; ② Calculation results: The probability of tower failure in the eyewall area of ​​the typhoon is Pfail=52% (red alert); ③ Automatic execution plan: Reduce the output of upstream wind farms to 780MW (originally 1560MW); activate energy storage discharge (SOC decreases from 65% to 42%).

[0062] 2. Blockchain Verification (Three-Level Chain Collaboration) Verification level Verification content Verification results Provincial chain DC tie-line power fluctuation 4.2% / min (not exceeding the 5% threshold) Earth-level chain Key bus voltage qualification rate 99.82% (meets the standard) Site-level links SOC adjustment delay 38ms (meets target) Consensus result: 87% of nodes passed the policy validity verification.

[0063] 3. Economic evaluation The carbon energy optimization report is as follows: index Before optimization After optimization Wind curtailment rate 8.7% 2.1% Carbon strength 0.68kg / kWh 0.59kg / kWh Total cost 4.82 million yuan 4.1 million yuan 4. Extreme scenario testing ①N-1 fault test: Simulated 500kV main transformer trip: Digital twin prediction: Overload of 3 220kV lines (load rate > 105%); Actual execution: The blockchain triggered emergency energy storage support, eliminating the overload within 5 seconds.

[0064] ② Network attack defense: Injecting false data: Bus voltage amplitude was altered by +8%; The χ² detector alarms (J(x)=35.2>χ² threshold value 28.3). The system automatically switches to the redundant measurement channel.

[0065] 5. Key performance indicators are the same as in 3. Carbon Energy Optimization Report.

[0066] 6. Verification of Technological Innovations ① Blockchain-Digital Twin Linkage: The entire process from typhoon warning to strategy execution took 8.3 seconds (traditional methods take 3-5 minutes); ② Convergence of ADMM algorithm Minute-level AGC optimization iteration count: Scene Traditional methods This patent normal 15 times 9 times typhoon Non-convergence 12 times IV. Conclusion confirms: In complex disaster scenarios, the platform can improve power grid resilience by 32% and reduce operating costs by 14.9%.

[0067] The above description is only a preferred embodiment of the present invention, but the scope of protection of the present invention is not limited thereto. Any equivalent substitutions or modifications made by those skilled in the art within the scope of the technology disclosed in the present invention, based on the technical solution and inventive concept of the present invention, should be covered within the scope of protection of the present invention.

Claims

1. A method for constructing a simulation and verification platform for source-grid-load-storage regulation strategy, characterized in that, Includes the following steps: S1: Requirements Analysis and Goal Setting; S2: Establish a dynamic model of source-grid-load-storage; S3: Build a heterogeneous computing platform; S4: Design a tiered control strategy; S5: Build a multi-scenario testing environment; S6: Verify the effectiveness of the strategy through blockchain consensus and dynamically assess resilience indicators using digital twins; S7: Generate a joint carbon energy optimization report; The sources of the demand include historical operating data from the power grid dispatch center, policy documents, and user-side data; the target setting is specifically a quantitative target, including technical indicators, carbon emission indicators, and resilience indicators. The source-grid-load-storage dynamic model includes a thermal power / wind and solar power unit model, a transmission network model, a load response model, and an energy storage converter model; The heterogeneous computing platform comprises a data interface layer, a computing engine layer, and a policy verification layer. The data interface layer supports IEC 61850 and Modbus protocols; The computing engine layer uses FPGA to accelerate electromagnetic transient simulation; The strategy verification layer integrates a blockchain multi-level verification module and a digital twin inference module.

2. The method for constructing a simulation verification platform for source-grid-load-storage regulation strategy according to claim 1, characterized in that, The implementation of the blockchain multi-level verification module includes the following aspects: 1) Provincial-level blockchain verification system; The node hardware configuration includes: ① a server cluster equipped with Intel Xeon Scalable processors, with each node configured with SGX enclave memory ≥256MB; ② a secure communication protocol based on the national cryptographic SM9 algorithm with end-to-end encryption and a key update cycle of ≤15 minutes. Additionally, the verification logic includes: ① Calculating power volatility: ;in, The rated power of the tie line is used to trigger a strategy review when the R_ fluctuation at three consecutive sampling points is greater than 5% / min. ② Consensus mechanism: Improved HotStuff algorithm, with view switching timeout set to 3 seconds; 2) Local chain verification system; Among them, the node security module includes: ① The distribution network RTU equipment integrates SM2-SM3 dual algorithm chips, with a signature speed of ≥1000 times / second; ② Lightweight zero-knowledge proof is performed before data is uploaded to the chain, with a proof time of ≤10ms; Additionally, voltage compliance rate assessment: ① Sampling standards: At least 3 monitoring points must be deployed in each area, namely the beginning, middle and end points, with a sampling interval of ≤1 second; ② Qualification criteria: The duration of voltage deviation within ±7% accounts for ≥99.7% of the total duration. 3) Station-level chain verification system; Among them, real-time performance is guaranteed: the energy storage PCS controller adopts a time-triggered architecture with a clock synchronization accuracy of ≤1μs; Another trusted execution environment: ① The TEE chip must be certified by CC EAL5+ and support dynamic trust measurement; ② Critical instructions must be decrypted and executed within the enclave. 4) Cross-chain interaction protocol; The data format adopts the IEEE 2030.5 standard extended fields; In addition, the conflict resolution uses the Shapley value algorithm based on game theory to assign verification weights.

3. The method for constructing a simulation verification platform for source-grid-load-storage regulation strategy according to claim 1, characterized in that, The digital twin simulation module includes a data input layer, a physical field modeling layer, a strategy simulation layer, and a visualization and interaction layer. The data input layer is used to access meteorological data and monitor equipment status. The meteorological data includes typhoon data and ice storm data. Specifically: ① Real-time access to ECMWF typhoon forecast data, with a spatial resolution of 0.1°×0.1° and a temporal resolution of 1 hour, including central pressure, movement path, and radius of the 7-level wind circle; ② Access to freezing rain forecast data from the Central Meteorological Observatory, with a spatial resolution of 5km×5km, including ambient temperature, precipitation intensity, and duration. In addition, the equipment status is specifically obtained through the PHM system to obtain the remaining life of the transformer, and the monitoring parameters include dissolved gas in oil, winding hot spot temperature, and partial discharge.

4. The method for constructing a source-grid-load-storage regulation strategy simulation and verification platform according to claim 3, characterized in that, The physical field modeling layer includes typhoon scenario modeling and ice storm scenario modeling; among them, the typhoon scenario modeling includes: ① Wind field reconstruction: using CFD fluid dynamics simulation, with a computational domain size of 20km×20km×1km and a mesh size ≤50m; ② Tower failure criteria: ; In addition, the content of ice disaster scenario modeling includes: ① Icing growth model: ② Insulator flashover criterion: Flashover warning is triggered when the ice thickness is ≥15mm and the humidity is >90%.

5. The method for constructing a source-grid-load-storage regulation strategy simulation and verification platform according to claim 4, characterized in that, The strategy deduction layer is a dynamic contingency plan generation layer, which includes the generation of typhoon scenarios and ice storm scenarios. Specifically, ① when the typhoon scenario is generated: when Automatic startup when >30%: Contingency Plan A: Reduce the output of wind farms upstream of the typhoon's path to 50% of their rated capacity; Contingency Plan B: Pre-set tree trimming instructions for the power transmission corridor, with tree trimming height ≥ conductor-to-ground distance + 3m; ② When generating an ice disaster scenario: Automatically trigger when the predicted ice thickness is >20mm: Contingency Plan C: Activate the DC de-icing device with a current density ≥3A / mm²; Contingency Plan D: Switch to the backup line.

6. The method for constructing a source-grid-load-storage regulation strategy simulation and verification platform according to claim 5, characterized in that, The visualization interaction layer includes a 3D geographic information system display and a risk level output. The 3D geographic information system display includes the typhoon path prediction error zone and the tower failure heat map. The risk level output includes a comprehensive risk index R = 0.6 × Pfail + 0.4 × Flashover_prob.

7. The method for constructing a simulation verification platform for source-grid-load-storage regulation strategy according to claim 1, characterized in that, The design of the hierarchical control strategy includes second-level frequency regulation, minute-level AGC, and hour-level economic dispatch. The second-level frequency adjustment adopts an event-triggered distributed consensus algorithm, and the response time of the second-level frequency adjustment is ≤200ms from the frequency exceeding the limit to the command being issued. The minute-level AGC uses an improved distributed ADMM algorithm: ,in, Let P be the power generation cost function of the i-th generating unit; i The active power output of the i-th unit is MW; λ i Let z be the Lagrange multiplier vector; z be the global consistency variable; ρ be the penalty factor; the convergence condition of the minute-level AGC is that the difference in power allocation between two adjacent iterations is <1MW; The hourly economic scheduling model considers carbon constraints and is a stochastic optimization model. ； ； in, Let γ be the power generation cost of unit g during time period t, in yuan; γ is the carbon price weighting coefficient, in yuan / kgCO2; The bus carbon intensity during time period t is expressed in kgCO2 / kWh. Let g be the output of unit g during time period t, in MW; Let g be the output of unit g during time period t, in MW; Reserve capacity for time period t; The hourly economic scheduling scenario reduction uses K-means clustering to reduce the number of wind and solar power output scenarios from 1000 to 10 representative scenarios.

8. The method for constructing a simulation and verification platform for source-grid-load-storage regulation strategy according to claim 1, characterized in that, The test environment covers normal weather, extreme weather, and cyberattack scenarios. The extreme weather scenarios include typhoon scenarios and ice storm scenarios. In the typhoon scenario, the wind field model adopts the Jensen wake model, and the turbulence intensity is set to IEC Class A. Tower failure: when the wind speed is >42m / s, the failure probability is calculated according to the Weibull distribution. In the ice storm scenario, the icing growth adopts the Makkonen model, the ambient temperature is -5℃~0℃, and the liquid water content is 0.1~0.3g / m³. The network attack scenario includes a False Data Injection attack and defense verification. The False Data Injection attack targets the SCADA system state estimation module; its attack magnitude is defined as voltage amplitude tampering ≤10% and phase angle ≤5°. The defense verification employs a χ² detector with the following detection threshold settings: ; Where α = 0.01, m is the number of measurements, and n is the dimension of the state variables.

9. The method for constructing a simulation verification platform for source-grid-load-storage regulation strategy according to claim 1, characterized in that, The blockchain verification metrics include: ① Strategy validity consensus condition: confirmation from ≥2 / 3 of the nodes is required; ② Data upload format: Protobuf encoding is used, including timestamp, data hash, and signature. The resilience assessment of the digital twin includes assessment indicators and dynamic simulation. The assessment indicators include ① Load Guarantee Rate (LSR) = ∑ duration of critical load power supply / total duration; ② System Recovery Time (TRT): the time from the occurrence of a fault to 90% load recovery. The dynamic simulation includes ① using an RTDS real-time simulator with a step size ≤ 50 μs; ② real-time interaction with meteorological data API with an update frequency ≥ 1 time / minute.

10. The method for constructing a simulation verification platform for source-grid-load-storage regulation strategy according to claim 1, characterized in that, The template for the joint carbon energy optimization report includes the following: a. Cover: Scene name, timestamp, strategy version number; b. Abstract: Total carbon emissions, percentage of renewable energy, and radar charts of key indicators; c. Details: Unit output curve, energy storage SOC, carbon emission intensity thermogram; d. Recommendation: Optimization suggestions for the next scheduling cycle.

Images