Virtual power plant regulation and control method and system for realizing new energy consumption

Abstract

The invention relates to the technical field of virtual power plants, and discloses a virtual power plant regulation and control method and system for realizing new energy consumption, and the system comprises a multi-source heterogeneous data collection module, an intelligent prediction analysis module, a resource aggregation modeling module, an optimization decision module, a block chain cooperation module, and a digital twinborn evaluation module. Multi-time-scale coupling prediction is carried out on new energy output and load demand through the deep space-time convolutional neural network, fluctuation and intermittency characteristics of new energy can be described, short-term and ultra-short-term prediction precision is improved, wind curtailment and light curtailment rate and load reduction risk are reduced, and the prediction efficiency is improved. According to the method, high matching between a virtual power plant scheduling plan and an actual operation condition is guaranteed, a flexible resource feature matrix is constructed, and distributed energy storage, interruptible load and electric vehicle multi-element resources are subjected to refined modeling and aggregation, so that a virtual unit capable of being efficiently scheduled can be formed, the resource utilization efficiency is improved, and the overall scheduling cost is reduced.

Description

Technical Field

[0001] This invention relates to the field of virtual power plant technology, specifically to a virtual power plant control method and system for realizing the consumption of new energy. Background Technology

[0002] Under the context of the Energy Internet, new renewable energy generation will be connected to the grid on a large scale to reduce the proportion of traditional fossil fuel power generation and achieve the low-carbon and clean development of the power grid. As the proportion of renewable energy generation increases, the capacity required for grid regulation and the ramp-up rate of regulating units must also increase, leading to the issue of renewable energy consumption. Virtual power plants can aggregate controllable loads and demand-side resources such as energy storage to participate in multiple scenarios including peak shaving, frequency regulation, and renewable energy consumption, and currently have broad development potential.

[0003] Currently, in the field of renewable energy consumption, the regulation process of virtual power plants faces multiple technical bottlenecks. Due to the volatility and intermittency of renewable energy output, existing forecasting methods are unable to accurately characterize its dynamic characteristics across multiple time scales, resulting in large deviations in short-term and ultra-short-term forecasts. This leads to a low degree of matching between the dispatching plans formulated by virtual power plants and actual operating conditions, resulting in wind and solar power curtailment and insufficient load supply.

[0004] Therefore, a virtual power plant control method and system for realizing the consumption of new energy is proposed to solve the above problems. Summary of the Invention

[0005] To address the shortcomings of existing technologies, this invention provides a virtual power plant control method and system for realizing the consumption of new energy sources. This solves the problem mentioned in the background that the scheduling plan formulated by the virtual power plant does not match the actual operating conditions well, resulting in wind and solar curtailment and insufficient load supply.

[0006] To achieve the above objectives, the present invention provides the following technical solution: a virtual power plant control method and system for realizing renewable energy consumption, comprising: S1. Real-time collection of new energy output data, load demand data and power grid dispatch instructions to generate a dynamic new energy consumption dataset; S2. Based on a deep spatiotemporal convolutional neural network, perform multi-time-scale prediction processing on the dynamic consumption dataset of new energy sources, and output the short-term output prediction curve of new energy sources, the ultra-short-term fluctuation characteristic spectrum and the load elastic response range. S3. Construct a virtual power plant resource pool model, mapping distributed energy storage, interruptible loads, and electric vehicle clusters to virtual machine group parameters, and generating a flexibility resource feature matrix. S4. Based on the short-term output prediction curve of new energy sources and the characteristic matrix of flexible resources, multi-objective optimization matching is performed, and an improved genetic algorithm is used to solve the optimal scheduling strategy to generate the day-ahead control plan. S5. Based on the ultra-short-term fluctuation characteristic spectrum and real-time instructions, rolling corrections are made, and the day-ahead control plan is dynamically adjusted through a fuzzy adaptive controller to output a minute-level control instruction sequence. S6. Establish a risk assessment model and calculate the dynamic ramp margin and risk index based on the load elastic response range and minute-level control command sequence. S7. When the risk index exceeds the threshold, activate the blockchain mechanism to call cross-regional backup resources and generate a collaborative consumption and compensation plan. S8. Execute the control instruction sequence and collaborative absorption compensation scheme through smart contracts, and verify the control effect; S9. Based on the verification of the control effect, construct a digital twin evaluation system to generate a consumption efficiency report and optimization suggestions.

[0007] Preferably, the new energy dynamic consumption dataset in S1 includes new energy power generation potential data, multi-user load data, and market and grid constraint data, and the construction process includes: S11. Collect new energy power generation potential data of photovoltaic power plants and wind farms through meteorological satellite remote sensing data and power prediction system at the site; S12. Collect electricity consumption characteristic data of industrial, commercial and residential loads through smart meter measurement system and demand response management platform; S13. Obtain day-ahead and real-time electricity price signals and cross-sectional transmission constraint data through the electricity market trading system and the power grid energy management system.

[0008] Preferably, the deep spatiotemporal convolutional neural network in S2 adopts a three-layer heterogeneous structure: The first-layer spatiotemporal convolution module extracts the geographical distribution characteristics of new energy power output; The second-layer gated loop unit captures the temporal correlation of output fluctuations; The third-layer self-attention mechanism integrates meteorological mutation factors and load transfer characteristics.

[0009] Preferably, the flexibility resource feature matrix in S3 includes four-dimensional parameters: The first dimension characterizes the charging and discharging power range and capacity status of the energy storage system. The second dimension quantifies the duration of interruptible load reduction and the cost of compensation. The third dimension aggregates the V2G response rate and power boundary of electric vehicle clusters; The fourth dimension integrates the rapid start-stop characteristics and carbon emission coefficient of small gas turbines; The flexibility resource characteristic matrix is ​​represented by an 8×m matrix as follows: ; in This represents the flexibility resource characteristic matrix of a virtual power plant. Indicates the total number of time periods in the future scheduling cycle. , These represent the time periods. Minimum and maximum charge / discharge power of energy storage systems Indicates the time period The remaining capacity status of the energy storage system Indicates the time period Reduce the compensation costs required per unit of interruptible load. Indicates the time period The maximum discharge power that the electric vehicle cluster can provide, Indicates the time period The rate of climb for a small gas turbine indicates the maximum increase in output it can provide per minute. Indicates the time period The downhill ramp rate of a small gas turbine indicates the maximum reduction in output it can achieve per minute. Indicates the time period The carbon emission coefficient generated per unit of electricity produced by a small gas turbine.

[0010] Preferably, the improved genetic algorithm in S4 includes a triple optimization mechanism: The primary optimization objective is to minimize the renewable energy curtailment rate and load reduction. The second optimization objective is to balance the cost of all flexible resource allocations. The third optimization objective satisfies the power grid security constraint equation.

[0011] Preferably, the fuzzy adaptive controller of S5 adopts a dual closed-loop control structure: The outer ring adjusts the power allocation ratio based on ultra-short-term prediction errors; The inner loop dynamically adjusts the control parameter tuning range based on the frequency deviation signal.

[0012] Preferably, the calculation method for the dynamic ramp margin index in S6 is as follows: Within a 15-minute time window, the difference in matching degree between the actual output fluctuation rate of new energy and the maximum ramping capacity of the virtual power plant is calculated, and the transmission margin of the grid section is superimposed to form a three-dimensional evaluation vector, the mathematical expression of which is: ; in Indicates at time The dynamic ramp margin index, Indicates the virtual power plant at any time The maximum uphill power that can be provided Indicates the virtual power plant at any time The maximum downhill climbing power that can be provided Indicates the time approaching short time window Within, the rate of change in the actual output of new energy sources, Indicates at time Transmission power margin at key sections of the power grid.

[0013] Preferably, the blockchain mechanism in S7 adopts the DPoS optimization algorithm: Select a regional energy hub node as the accountant to verify the reliability of backup resources; The smart contract is triggered to enable the coordinated consumption of power by multiple virtual power plants through a cross-chain communication protocol.

[0014] Preferably, the waste disposal efficiency report and optimization suggestions in S9 are generated through a multi-dimensional weighted evaluation algorithm, including five core indicators: The completion rate of new energy consumption, the contribution of power grid regulation, the reduction in carbon emission intensity, the cost-effectiveness ratio of resource allocation, and user satisfaction with electricity use.

[0015] Preferably, the system includes: The multi-source heterogeneous data acquisition module acquires new energy power generation potential data through a meteorological satellite receiving unit, collects multi-user load data through a broadband measurement unit, and receives market and grid constraint data through a market interface protocol converter, outputting a dynamic new energy consumption dataset. The intelligent prediction and analysis module receives the dynamic consumption dataset of new energy, generates a short-term power output prediction curve of new energy through a deep spatiotemporal convolutional neural network computing engine, and outputs an ultra-short-term fluctuation feature spectrum and load elastic response range using a fluctuation feature extraction algorithm library. The resource aggregation modeling module receives the multi-user load data and new energy power generation potential data, generates schedulable virtual machine group parameters through the virtual machine group equivalent converter, and constructs a flexible resource feature matrix using a flexible resource digital twin. The optimization decision module receives the short-term output forecast curve of the new energy source and the characteristic matrix of the flexible resources, performs optimization matching processing through a multi-objective solver, and generates a day-ahead control plan and a minute-level control instruction sequence using a rolling correction controller. The blockchain collaboration module receives the control instruction sequence, executes instruction verification through cross-regional consensus verification nodes, activates cross-regional backup resource calls using the smart contract execution engine, and outputs a collaborative consumption compensation scheme. The digital twin evaluation module receives the execution effect data and collaborative absorption compensation scheme data of the minute-level control command sequence, displays the control process through a three-dimensional visualization platform, and generates an absorption efficiency evaluation report and strategy optimization suggestions using a strategy self-optimization feedback unit. The strategy optimization suggestions are fed back to the rolling correction controller in the optimization decision module, which is used to adaptively adjust the generation logic of subsequent control instructions. Beneficial effects

[0016] Compared with existing technologies, the present invention provides a virtual power plant control method and system for realizing the consumption of new energy sources, which has the following beneficial effects: 1. In this invention, a deep spatiotemporal convolutional neural network is used to perform multi-timescale coupled prediction of new energy output and load demand, which can characterize the volatility and intermittency of new energy, improve the accuracy of short-term and ultra-short-term prediction, thereby reducing the wind and solar curtailment rate and load reduction risk, and ensuring a high degree of matching between the virtual power plant dispatch plan and the actual operating conditions.

[0017] 2. In this invention, by constructing a flexible resource feature matrix and performing refined modeling and aggregation of distributed energy storage, interruptible loads, and electric vehicle resources, a highly schedulable virtual machine group can be formed, fully exploring and coordinating the response potential of various flexible resources, improving resource utilization efficiency, and reducing overall scheduling costs.

[0018] 3. In this invention, by introducing a rolling optimization correction mechanism and a blockchain collaborative absorption scheme, it is possible to respond in real time to grid commands and new energy power fluctuations, adaptively adjust scheduling strategies and quickly and reliably call cross-regional backup resources. At the same time, relying on a digital twin evaluation system, it forms multi-dimensional performance feedback and closed-loop optimization, thereby enhancing the safety and stability of grid operation and the overall absorption level of new energy. Attached Figure Description

[0019] Figure 1 This is a flowchart of a virtual power plant control method for realizing the consumption of new energy, according to the present invention. Figure 2 This is a schematic diagram of the architecture of a virtual power plant control system for realizing the consumption of new energy sources according to the present invention. Detailed Implementation

[0020] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0021] For specific implementation examples, please refer to: Figure 1-2 A virtual power plant control method and system for realizing renewable energy consumption, comprising: S1. Real-time collection of new energy output data, load demand data and power grid dispatch instructions to generate a dynamic new energy consumption dataset; S2. Based on a deep spatiotemporal convolutional neural network, multi-timescale prediction processing is performed on the dynamic absorption dataset of new energy sources to output the short-term output prediction curve, ultra-short-term fluctuation characteristic spectrum, and load elastic response range of new energy sources. The specific operation includes: the network adopts a three-layer heterogeneous structure. The first layer spatiotemporal convolutional module extracts the spatial geographical distribution characteristics of new energy output. The second layer gated recurrent unit captures the temporal correlation of its output fluctuation. The third layer self-attention mechanism integrates meteorological change factors and load transfer characteristics. Through network training and inference, the network outputs the short-term output prediction curve, ultra-short-term fluctuation characteristic spectrum, and load elastic response range of new energy sources. S3. Construct a virtual power plant resource pool model, mapping distributed energy storage, interruptible loads, and electric vehicle clusters to virtual machine group parameters, and generating a flexibility resource feature matrix. S4. Based on the short-term output forecast curve of new energy and the characteristic matrix of flexible resources, a multi-objective optimization matching is carried out. The short-term output forecast curve of new energy is used as the basic forecast data, and the characteristic matrix of flexible resources is used as the optimization adjustment boundary. A multi-objective optimization model containing triple optimization objectives and grid security constraints is constructed. An improved genetic algorithm is used to solve the optimal scheduling strategy and generate the day-ahead control plan. S5. Based on the ultra-short-term fluctuation characteristic spectrum and real-time instructions, rolling corrections are performed. The daytime control plan is dynamically adjusted through a fuzzy adaptive controller, and a minute-level control instruction sequence is output. The specific operation includes: S51. Input and Error Calculation: Real-time reading of the minute-level fluctuation prediction value of new energy power characterized by the ultra-short-term fluctuation characteristic spectrum, and collection of real-time grid dispatch instructions, calculation of the error between the predicted value and the actual value of new energy power, as well as the deviation between the current control plan and the real-time instructions; S52. Fuzzification Processing: The prediction error and instruction deviation obtained above are used as input variables, and according to the preset fuzzy rule base, they are mapped to the corresponding fuzzy linguistic variables and membership degrees respectively. S53. Fuzzy Reasoning and Decision-Making: Outer Loop Control: Based on the fuzzy quantity of ultra-short-term prediction error, fuzzy reasoning rules are applied to dynamically correct the priority and proportion of various flexible resources in power allocation; Inner loop control: Based on the fuzzy amount of the power grid frequency deviation signal, another set of fuzzy inference rules is applied to adaptively adjust the parameter tuning range of the controller in order to optimize the response speed and stability of the control system. S54. Defuzzification and Command Generation: The fuzzy quantity output after fuzzy inference decision-making is converted into a precise power adjustment quantity through a defuzzification algorithm. This adjustment quantity is applied to the day-ahead control plan to generate precise power control commands for the next few minutes to tens of minutes. S55, Command Sequence Output: Arrange the generated precise power control commands in chronological order, encapsulate them into a minute-level control command sequence that can be issued, and output it to the execution unit; S6. Establish a risk assessment model and calculate the dynamic ramp margin and risk index based on the load elastic response range and minute-level control command sequence. S7. When the risk index exceeds the threshold, activate the blockchain mechanism to call cross-regional backup resources and generate a collaborative consumption and compensation plan. S8. Execute the control instruction sequence and collaborative absorption compensation scheme through smart contracts, and verify the control effect; S9. Based on the verification of the control effect, construct a digital twin evaluation system to generate a consumption efficiency report and optimization suggestions.

[0022] The dynamic renewable energy consumption dataset in S1 includes renewable energy generation potential data, diverse user load data, and market and grid constraint data. The construction process includes: S11. Collect new energy power generation potential data of photovoltaic power plants and wind farms through meteorological satellite remote sensing data and power prediction system at the site; S12. Collect electricity consumption characteristic data of industrial, commercial and residential loads through smart meter measurement system and demand response management platform; S13. Obtain day-ahead and real-time electricity price signals and cross-sectional transmission constraint data through the electricity market trading system and the power grid energy management system.

[0023] The deep spatiotemporal convolutional neural network in S2 adopts a three-layer heterogeneous structure: The first-layer spatiotemporal convolution module extracts the geographical distribution characteristics of new energy power output; The second-layer gated loop unit captures the temporal correlation of output fluctuations; The third-layer self-attention mechanism integrates meteorological mutation factors and load transfer characteristics.

[0024] The flexibility resource feature matrix in S3 contains four-dimensional parameters: The first dimension characterizes the charging and discharging power range and capacity status of the energy storage system. The second dimension quantifies the duration of interruptible load reduction and the cost of compensation. The third dimension aggregates the V2G response rate and power boundary of electric vehicle clusters; The fourth dimension integrates the rapid start-stop characteristics and carbon emission coefficient of small gas turbines; The flexibility resource characteristic matrix is ​​represented by an 8×m matrix as follows: ; in This represents the flexibility resource characteristic matrix of a virtual power plant. Indicates the total number of time periods in the future scheduling cycle. , These represent the time periods. Minimum and maximum charge / discharge power of energy storage systems Indicates the time period The remaining capacity status of the energy storage system Indicates the time period Reduce the compensation costs required per unit of interruptible load. Indicates the time period The maximum discharge power that the electric vehicle cluster can provide, Indicates the time period The rate of climb for a small gas turbine indicates the maximum increase in output it can provide per minute. Indicates the time period The downhill ramp rate of a small gas turbine indicates the maximum reduction in output it can achieve per minute. Indicates the time period The carbon emission coefficient generated per unit of electricity produced by a small gas turbine.

[0025] The improved genetic algorithm in S4 includes a triple optimization mechanism: The primary optimization objective is to minimize the renewable energy curtailment rate and load reduction. The second optimization objective is to balance the cost of all flexible resource allocations. The third optimization objective satisfies the power grid security constraint equation; The specific steps for improving the genetic algorithm include: S41. Chromosome Encoding and Population Initialization: Chromosomes are constructed using real-number encoding. Each chromosome contains a scheduling plan sequence for flexible resources, and an initial population is generated randomly. S42. Multi-objective fitness function calculation: Construct the fitness function based on the triple optimization mechanism; its expression is: ; in This represents the fitness value of an individual; a higher value indicates a better solution. , , These are the weighting coefficients. Indicates the rate of curtailment of renewable energy. Indicates the amount of load reduction. This represents the standard deviation of resource allocation costs. Indicates the degree of violation of power grid constraints; Weighting coefficient , , Dynamically adjust based on the real-time operating status of the power grid: Increase when the predicted output of new energy sources is higher. When the cost of resource allocation differs significantly, it increases. Increase when the power grid operation approaches the safety boundary ; S43. Non-dominated ranking and crowding calculation: Perform non-dominated ranking on individuals in the population, divide them into Pareto ranks, and calculate the crowding distance between individuals within the same rank. S44. Elite Retention and Selection Operation: A binary tournament selection method is adopted, with the tournament size set to 2. The Pareto level of an individual is the primary selection criterion, and crowding distance is the secondary criterion. Excellent individuals are retained to the next generation of the population. S45. Crossover and mutation operations: A new offspring population is generated using a simulated binary crossover operator and a polynomial mutation operator, where the crossover distribution index is set to 20 and the mutation distribution index is set to 20. S46. Iteration Termination Judgment: When the maximum number of iterations is reached, output the Pareto optimal solution set; otherwise, return to S42 to continue iterating.

[0026] The S5's fuzzy adaptive controller employs a dual closed-loop control structure: The outer ring adjusts the power allocation ratio based on ultra-short-term prediction errors; The inner loop dynamically adjusts the control parameter tuning range based on the frequency deviation signal.

[0027] The calculation method for the dynamic ramp margin index in S6 is as follows: Within a 15-minute time window, the difference in matching degree between the actual output fluctuation rate of new energy and the maximum ramping capacity of the virtual power plant is calculated, and the transmission margin of the grid section is superimposed to form a three-dimensional evaluation vector, the mathematical expression of which is: ; in Indicates at time The dynamic ramp margin index, Indicates the virtual power plant at any time The maximum uphill power that can be provided Indicates the virtual power plant at any time The maximum downhill climbing power that can be provided Indicates the time approaching short time window Within, the rate of change in the actual output of new energy sources, Indicates at time Transmission power margin at key sections of the power grid.

[0028] The blockchain mechanism in S7 uses the DPoS optimized algorithm: Select a regional energy hub node as the accountant to verify the reliability of backup resources; Smart contract triggering enables collaborative consumption of power from multiple virtual power plants through cross-chain communication protocols; The specific steps of the DPoS optimization algorithm include: S71. Election of Ledger Nodes: Based on the historical scheduling performance data and communication reliability indicators of regional energy hub nodes, calculate the comprehensive reputation value of each node, and select the N nodes with the highest comprehensive reputation values ​​as the set of ledger node for the current consensus period. S72. Resource Verification and Block Generation: The ledger node performs multiple rounds of voting to verify the callable capacity, response rate, and contract validity of cross-regional backup resources. The verified resource information and scheduling instructions are packaged to generate a new block. S73, Cross-chain communication trigger: The on-chain state of the target virtual power plant is obtained through the cross-chain communication relay bridge. When the local blockchain generates a new block containing the instructions to call up backup resources, the cross-chain smart contract is automatically triggered to initiate a collaborative consumption request to the target virtual power plant. S74. State Synchronization and Confirmation: The smart contract executor on the target virtual power plant chain receives the request and verifies its legality. After execution, it feeds back the confirmation result to the source blockchain through the cross-chain protocol, thus completing the global state synchronization.

[0029] The S9's waste disposal efficiency report and optimization suggestions are generated using a multi-dimensional weighted evaluation algorithm, including five core indicators: New energy consumption completion rate, grid regulation support contribution, carbon emission intensity reduction, resource utilization cost-effectiveness ratio, and user electricity satisfaction. The specific operational steps of the multi-dimensional weighted evaluation algorithm include: S91. Standardization of indicator data: Obtain the original data of the five core indicators respectively, and use the extreme value standardization method to normalize the data of each indicator to the [0,1] interval to form a standardized indicator vector. S92. Combined weight calculation: The subjective weight of each indicator is calculated using the analytic hierarchy process (AHP), and the objective weight of each indicator is calculated using the entropy weight method. The subjective and objective weights are then integrated using the multiplication integration method to generate a comprehensive weight coefficient. S93. Comprehensive Score Calculation: The standardized indicator vector and the comprehensive weight coefficient are weighted and summed to generate a comprehensive score for new energy consumption efficiency. The calculation formula is as follows: ; in For the first The weight of each indicator For the first The standardized value of the indicator; S94. Evaluation Report Generation: Based on the comprehensive score, efficiency levels are divided, and based on the comparative analysis of the scores of each sub-indicator and the weight coefficients, optimization suggestions are generated for prediction accuracy, resource scheduling strategy, and response command execution.

[0030] The system includes: The multi-source heterogeneous data acquisition module acquires new energy power generation potential data through a meteorological satellite receiving unit, collects multi-user load data through a broadband measurement unit, and receives market and grid constraint data through a market interface protocol converter, outputting a dynamic new energy consumption dataset. The intelligent predictive analysis module receives dynamic renewable energy consumption datasets, generates short-term renewable energy output prediction curves through a deep spatiotemporal convolutional neural network computing engine, and outputs ultra-short-term fluctuation feature spectra and load elastic response ranges using a fluctuation feature extraction algorithm library. The operation process of the fluctuation feature extraction algorithm library includes: I. The wavelet packet decomposition algorithm is used to perform multi-level decomposition of the time series data of new energy power to extract the energy distribution characteristics in different frequency bands; II. Use spectral clustering algorithm to identify the energy distribution characteristics and classify them into stationary fluctuation modes, gradually changing fluctuation modes, and abrupt fluctuation modes; III. Based on historical data, statistically analyze the duration, slope of change, and amplitude probability distribution of each wave mode to generate a feature spectrum to describe the wave characteristics; The resource aggregation modeling module receives diverse user load data and new energy power generation potential data, generates schedulable virtual machine group parameters through a virtual machine group equivalent converter, and constructs a flexible resource feature matrix using a flexible resource digital twin. The specific operation process of the virtual machine group equivalent converter includes: Ⅰ. Feature Vector Extraction: For various flexible resources, the adjustable power upper limit, adjustable power lower limit, rated capacity, maximum ramp rate, minimum continuous running time, and response delay time are extracted as equal-value clustering feature vectors. II. Dynamic Weighted Clustering: The k-means clustering algorithm based on Euclidean distance is adopted, and the feature vectors are dynamically weighted according to the importance of resource type and scheduling cost, so that distributed resources with similar characteristics are aggregated into several virtual machine groups. III. Equivalent Parameter Mapping: The adjustable power upper and lower limits of all resources in each virtual machine group are algebraically summed to form the total adjustable power range of the virtual machine group; the maximum values ​​of the minimum continuous running time and response latency of the resources in the group are respectively used as the minimum continuous running time and response latency of the virtual machine group. The optimization decision-making module receives the short-term output forecast curve of new energy sources and the characteristic matrix of flexible resources. It performs optimization matching processing through a multi-objective solver and uses a rolling correction controller to generate day-ahead control plans and minute-level control instruction sequences. The specific operation process of the rolling correction controller includes: I. Error Calculation: Receive the ultra-short-term fluctuation characteristic spectrum and the real-time grid dispatch command, and calculate the prediction error of new energy power and the deviation of the grid command; II. Parameter Tuning: Based on the prediction error of new energy power and the deviation of grid command, the proportional, integral, and derivative parameters of the fuzzy adaptive controller are dynamically adjusted using fuzzy logic rules; III. Instruction Generation: The day-ahead control plan is continuously optimized using the tuned fuzzy adaptive controller to generate minute-level control instruction sequences; IV. Safety Verification: The sequence of control commands is sent to the power grid safety constraint verification unit for over-limit detection to ensure that the commands meet the requirements for safe system operation; The blockchain collaboration module receives a sequence of control instructions, verifies the execution of instructions through cross-regional consensus verification nodes, activates the cross-regional backup resource call using the smart contract execution engine, and outputs a collaborative consumption compensation scheme. The digital twin evaluation module receives execution effect data of minute-level control command sequences and collaborative absorption compensation scheme data, displays the control process through a 3D visualization platform, and generates absorption efficiency evaluation reports and strategy optimization suggestions using the strategy self-optimization feedback unit. Among them, the strategy optimization suggestions are fed back to the rolling correction controller in the optimization decision module, which is used to adaptively adjust the generation logic of subsequent control instructions.

[0031] Based on the above-disclosed content and in conjunction with practical applications, the operation steps of the virtual power plant control method and system for realizing renewable energy consumption according to the present invention are as follows: Step 1: Multi-source data sensing and aggregation The system collects and aggregates power output data from new energy power generation equipment, load demand data from diverse users, and dispatch instructions issued by the power grid dispatching agency in real time through various deployed sensing and communication units, which together form a dynamic new energy consumption dataset for subsequent in-depth analysis.

[0032] Step 2: Multi-timescale collaborative prediction By using deep spatiotemporal convolutional neural networks to intelligently analyze the dynamic consumption dataset of new energy sources, high-precision short-term output prediction curves of new energy sources, ultra-short-term fluctuation characteristic spectra for characterizing minute-level fluctuations, and load elastic response ranges are generated in parallel, providing accurate input basis for optimization decisions.

[0033] Step 3: Flexible Resource Aggregation Modeling A dynamic aggregation model for virtual power plant resource pools is constructed, which maps distributed energy storage, interruptible loads, and heterogeneous resources of electric vehicle clusters into schedulable virtual machine groups with standardized parameters through an equivalent transformation algorithm, and generates a flexible resource feature matrix that comprehensively characterizes their response capabilities.

[0034] Step 4: Multi-objective optimization decision generation Based on the new energy forecast curve and the flexibility resource characteristic matrix, an improved genetic algorithm is used to solve the problem. This algorithm simultaneously optimizes the three objectives of maximizing consumption, cost economy and grid security, and finally generates the day-ahead control plan of the virtual power plant.

[0035] Step 5: Rolling Correction and Real-time Command Generation Based on the ultra-short-term fluctuation characteristic spectrum and real-time grid dispatch instructions, a fuzzy adaptive controller is used to dynamically and continuously optimize and correct the day-ahead plan on a minute-by-minute basis. This process specifically includes: first, calculating the prediction error of renewable energy power and the instruction deviation; then, dynamically tuning the controller parameters using fuzzy logic rules; subsequently, using the tuned controller to generate a minute-by-minute real-time control instruction sequence; and finally, sending this instruction sequence to a safety verification unit to ensure that it meets all safety constraints related to line power, node voltage, and system frequency before outputting it.

[0036] Step Six: Cross-Regional Collaborative Emergency Response When the assessment finds that the local adjustment capacity is insufficient, the mechanism based on blockchain consensus is immediately activated. Through smart contracts, it automatically and reliably verifies and calls cross-regional backup resources, generates and executes an emergency collaborative absorption and compensation plan to cope with extreme working conditions.

[0037] Step 7: Full-process digital twin evaluation and optimization Construct a digital twin evaluation system for regulation effectiveness to comprehensively evaluate multiple dimensions of indicators, including renewable energy absorption rate, grid support, carbon emission reduction benefits, economic efficiency, and user satisfaction, throughout the entire regulation process. Generate quantitative evaluation reports and strategy optimization suggestions to form a closed-loop feedback loop and drive continuous system optimization.

[0038] It should be noted that, in this document, relational terms such as "first" and "second" are used only to distinguish one entity or operation from another, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Furthermore, the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such a process, method, article, or apparatus. Without further limitations, an element defined by the phrase "comprising one..." does not exclude the presence of other identical elements in the process, method, article, or apparatus that includes said element.

[0039] Although embodiments of the invention have been shown and described, it will be understood by those skilled in the art that various changes, modifications, substitutions and alterations can be made to these embodiments without departing from the principles and spirit of the invention, the scope of which is defined by the appended claims and their equivalents.

Claims

1. A virtual power plant control method for realizing the consumption of new energy sources, characterized in that: include: S1. Real-time collection of new energy output data, load demand data and power grid dispatch instructions to generate a dynamic new energy consumption dataset; S2. Based on a deep spatiotemporal convolutional neural network, perform multi-time-scale prediction processing on the dynamic consumption dataset of new energy sources, and output the short-term output prediction curve of new energy sources, the ultra-short-term fluctuation characteristic spectrum and the load elastic response range. S3. Construct a virtual power plant resource pool model, mapping distributed energy storage, interruptible loads, and electric vehicle clusters to virtual machine group parameters, and generating a flexibility resource feature matrix. S4. Based on the short-term output prediction curve of new energy sources and the characteristic matrix of flexible resources, multi-objective optimization matching is performed, and an improved genetic algorithm is used to solve the optimal scheduling strategy to generate the day-ahead control plan. S5. Based on the ultra-short-term fluctuation characteristic spectrum and real-time instructions, rolling corrections are made, and the day-ahead control plan is dynamically adjusted through a fuzzy adaptive controller to output a minute-level control instruction sequence. S6. Establish a risk assessment model and calculate the dynamic ramp margin and risk index based on the load elastic response range and minute-level control command sequence. S7. When the risk index exceeds the threshold, activate the blockchain mechanism to call cross-regional backup resources and generate a collaborative consumption and compensation plan. S8. Execute the control instruction sequence and collaborative absorption compensation scheme through smart contracts, and verify the control effect; S9. Based on the verification of the control effect, construct a digital twin evaluation system to generate a consumption efficiency report and optimization suggestions.

2. The virtual power plant control method for realizing renewable energy consumption according to claim 1, characterized in that: The new energy dynamic consumption dataset in S1 includes new energy power generation potential data, multi-user load data, and market and grid constraint data. The construction process includes: S11. Collect new energy power generation potential data of photovoltaic power plants and wind farms through meteorological satellite remote sensing data and power prediction system at the site; S12. Collect electricity consumption characteristic data of industrial, commercial and residential loads through smart meter measurement system and demand response management platform; S13. Obtain day-ahead and real-time electricity price signals and cross-sectional transmission constraint data through the electricity market trading system and the power grid energy management system.

3. The virtual power plant control method for realizing renewable energy consumption according to claim 1, characterized in that: The deep spatiotemporal convolutional neural network in S2 adopts a three-layer heterogeneous structure: The first-layer spatiotemporal convolution module extracts the geographical distribution characteristics of new energy power output; The second-layer gated loop unit captures the temporal correlation of output fluctuations; The third-layer self-attention mechanism integrates meteorological mutation factors and load transfer characteristics.

4. The virtual power plant control method for realizing renewable energy consumption according to claim 1, characterized in that: The flexibility resource feature matrix in S3 contains four-dimensional parameters: The first dimension characterizes the charging and discharging power range and capacity status of the energy storage system. The second dimension quantifies the duration of interruptible load reduction and the cost of compensation. The third dimension aggregates the V2G response rate and power boundary of electric vehicle clusters; The fourth dimension integrates the rapid start-stop characteristics and carbon emission coefficient of small gas turbines; The flexibility resource characteristic matrix is ​​represented by an 8×m matrix as follows: ； in This represents the flexibility resource characteristic matrix of a virtual power plant. Indicates the total number of time periods in the future scheduling cycle. , These represent the time periods. Minimum and maximum charge / discharge power of energy storage systems Indicates the time period The remaining capacity status of the energy storage system Indicates the time period Reduce the compensation costs required per unit of interruptible load. Indicates the time period The maximum discharge power that the electric vehicle cluster can provide, Indicates the time period The rate of climb for a small gas turbine indicates the maximum increase in output it can provide per minute. Indicates the time period The downhill ramp rate of a small gas turbine indicates the maximum reduction in output it can achieve per minute. Indicates the time period The carbon emission coefficient generated per unit of electricity produced by a small gas turbine.

5. The virtual power plant control method for realizing renewable energy consumption according to claim 1, characterized in that: The improved genetic algorithm in S4 includes a triple optimization mechanism: The primary optimization objective is to minimize the renewable energy curtailment rate and load reduction. The second optimization objective is to balance the cost of all flexible resource allocations. The third optimization objective satisfies the power grid security constraint equation.

6. The virtual power plant control method for realizing renewable energy consumption according to claim 1, characterized in that: The fuzzy adaptive controller of S5 adopts a dual closed-loop control structure: The outer ring adjusts the power allocation ratio based on ultra-short-term prediction errors; The inner loop dynamically adjusts the control parameter tuning range based on the frequency deviation signal.

7. The virtual power plant control method for realizing renewable energy consumption according to claim 1, characterized in that: The calculation method for the dynamic ramp margin index in S6 is as follows: Within a 15-minute time window, the difference in matching degree between the actual output fluctuation rate of new energy and the maximum ramping capacity of the virtual power plant is calculated, and the transmission margin of the grid section is superimposed to form a three-dimensional evaluation vector, the mathematical expression of which is: ； in Indicates at time The dynamic ramp margin index, Indicates the virtual power plant at any time The maximum uphill power that can be provided Indicates the virtual power plant at any time The maximum downhill climbing power that can be provided Indicates the time approaching short time window Within, the rate of change in the actual output of new energy sources, Indicates at time Transmission power margin at key sections of the power grid.

8. The virtual power plant control method for realizing renewable energy consumption according to claim 1, characterized in that: The blockchain mechanism in S7 uses the DPoS optimization algorithm: Select a regional energy hub node as the accountant to verify the reliability of backup resources; The smart contract is triggered to enable the coordinated consumption of power by multiple virtual power plants through a cross-chain communication protocol.

9. The virtual power plant control method for realizing renewable energy consumption according to claim 1, characterized in that: The S9's waste disposal efficiency report and optimization suggestions are generated using a multi-dimensional weighted evaluation algorithm, and include five core indicators: The completion rate of new energy consumption, the contribution of power grid regulation, the reduction in carbon emission intensity, the cost-effectiveness ratio of resource allocation, and user satisfaction with electricity use.

10. A virtual power plant control system for realizing renewable energy consumption, used to implement the virtual power plant control method for realizing renewable energy consumption according to any one of claims 1-9, characterized in that: The system includes: The multi-source heterogeneous data acquisition module acquires new energy power generation potential data through a meteorological satellite receiving unit, collects multi-user load data through a broadband measurement unit, and receives market and grid constraint data through a market interface protocol converter, outputting a dynamic new energy consumption dataset. The intelligent prediction and analysis module receives the dynamic consumption dataset of new energy, generates a short-term power output prediction curve of new energy through a deep spatiotemporal convolutional neural network computing engine, and outputs an ultra-short-term fluctuation feature spectrum and load elastic response range using a fluctuation feature extraction algorithm library. The resource aggregation modeling module receives the multi-user load data and new energy power generation potential data, generates schedulable virtual machine group parameters through the virtual machine group equivalent converter, and constructs a flexible resource feature matrix using a flexible resource digital twin. The optimization decision module receives the short-term output forecast curve of the new energy source and the characteristic matrix of the flexible resources, performs optimization matching processing through a multi-objective solver, and generates a day-ahead control plan and a minute-level control instruction sequence using a rolling correction controller. The blockchain collaboration module receives the control instruction sequence, executes instruction verification through cross-regional consensus verification nodes, activates cross-regional backup resource calls using the smart contract execution engine, and outputs a collaborative consumption compensation scheme. The digital twin evaluation module receives the execution effect data and collaborative absorption compensation scheme data of the minute-level control command sequence, displays the control process through a three-dimensional visualization platform, and generates an absorption efficiency evaluation report and strategy optimization suggestions using a strategy self-optimization feedback unit. The strategy optimization suggestions are fed back to the rolling correction controller in the optimization decision module, which is used to adaptively adjust the generation logic of subsequent control instructions.

Images