A virtual power plant emergency frequency control method and related equipment
By building a centralized-decentralized hierarchical control architecture in the virtual power plant, the problem of command allocation mismatch caused by the hierarchical difference in response speed of heterogeneous resources is solved, realizing efficient hierarchical issuance of emergency frequency control and rapid response on the resource side, thus improving the emergency frequency control effect of the virtual power plant.
Patent Information
- Application Number
- CN202611380072.2
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2026-09-07
- Publication Date
- 2026-11-17
AI Technical Summary
Existing virtual power plant emergency frequency control methods fail to fully consider the differences in response speed levels of heterogeneous resources, resulting in a mismatch between command allocation and resource response capabilities. Furthermore, offline pre-calculation results lack effective transmission and utilization during the real-time emergency frequency regulation phase, affecting decision-making efficiency.
A centralized-decentralized hierarchical control architecture is established. Through information interaction and command transmission between the power grid control center, the virtual power plant aggregation layer, and the local resource control layer, frequency support capacity and price parameters are pre-calculated and stored. Emergency power support commands are generated in combination with real-time operating status and distributed hierarchically to various distributed resources to achieve rapid execution and closed-loop iterative optimization.
It enables the hierarchical issuance of emergency power support commands and rapid execution on the resource side, improving the response accuracy and overall frequency regulation effect of virtual power plants in emergency frequency control, and ensuring continuous optimization of control commands during the frequency recovery process.
Smart Images

Figure CN123121575A_ABST
Abstract
Description
Technical Field
[0001] This invention relates to the field of power system frequency control technology, specifically to an emergency frequency control method and related equipment for a virtual power plant. Background Technology
[0002] Virtual Power Plants (VPPs) aggregate heterogeneous distributed resources such as distributed generating units, new energy generating units, energy storage systems, electric vehicle clusters, and controllable loads, enabling them to participate as a whole in grid dispatch and electricity market transactions. When grid faults lead to active power deficits, the system frequency faces the risk of dropping. VPPs can provide inertial response and primary frequency regulation support by aggregating various flexible resources, making them an important means of ensuring the frequency security of new power systems.
[0003] Currently, research on VPP participation in power grid dispatching mainly focuses on planning and construction, economic dispatch, and electricity market transactions, with insufficient research on emergency frequency control. In existing emergency frequency control schemes, the power grid control center typically issues a total power command to the VPP, which is then received by the VPP aggregation layer and allocated to its internal distributed resources. However, the VPP aggregates a massive number of heterogeneous and flexible resources. Different types of resources exhibit significant differences in frequency response characteristics such as communication latency, response time constant, power ramp rate, and capacity boundaries. The response speed of each resource can be divided into multiple levels, ranging from milliseconds to seconds or even more than ten seconds. Most existing power allocation methods fail to adequately consider the hierarchical differences in response speed among heterogeneous resources, making it difficult to reasonably decompose fast and slow frequency regulation tasks. This leads to a mismatch between command allocation and the actual response capabilities of resources, hindering the overall frequency regulation performance of the VPP. On the other hand, the VPP frequency support capacity and price parameters obtained in the existing offline pre-calculation stage lack an effective transmission and utilization mechanism in the real-time emergency frequency adjustment stage. The offline calculation results cannot be directly used to guide the generation of real-time power support commands and response settlement after a fault occurs, resulting in low decision-making efficiency for emergency frequency control.
[0004] Therefore, existing technologies urgently need to address the mismatch between instruction allocation and response capabilities caused by insufficient consideration of the hierarchical differences in response speed of heterogeneous resources, as well as the lack of an effective transmission and utilization mechanism for frequency support capacity and price parameters obtained offline pre-calculation during the real-time emergency frequency modulation phase. Summary of the Invention
[0005] The purpose of this invention is to provide a virtual power plant emergency frequency control method and related equipment to overcome the problems existing in the prior art. This invention can realize the hierarchical issuance of emergency power support commands and rapid execution on the resource side, while enabling the offline pre-calculation results to be effectively transmitted and utilized in the real-time emergency frequency regulation stage.
[0006] To achieve the above objectives, the technical solution adopted by the present invention is as follows: In a first aspect, the present invention provides a virtual power plant emergency frequency control method, comprising the following steps: A centralized-decentralized hierarchical control architecture is established, comprising a power grid control center, a virtual power plant aggregation layer, and a local resource control layer connected in sequence; wherein, the power grid control center includes a power grid control center database, and the frequency support capacity and price parameters of each virtual power plant are pre-calculated offline and stored in the power grid control center database; Frequency response parameters of each distributed resource are extracted through the virtual power plant aggregation layer; The system obtains real-time power grid operation status data through the power grid control center. When a power grid fault with active power deficit is detected, it obtains preset frequency security constraints and reads the frequency support capacity and price parameters of each virtual power plant stored in the power grid control center database. The frequency support capacity of each virtual power plant is used as the initial available capacity of each virtual power plant. The frequency response parameters are called and the initial available capacity is corrected in combination with the current real-time operating status of each virtual power plant to obtain the real-time available frequency support capacity. Based on the active power deficit, frequency security constraints and real-time available frequency support capacity, the target support power of the virtual power plant is determined and an emergency power support command is generated. The emergency power support command is sent to the virtual power plant aggregation layer. The virtual power plant aggregation layer allocates the target support power of the virtual power plant based on the current operating status of each distributed resource and the frequency response parameters, and generates the corresponding resource adjustment command. The resource adjustment command is sent to the local resource control layer, which then controls the corresponding distributed resource execution power adjustment according to the resource adjustment command and outputs the corresponding actual response power. The emergency power support command for the next control cycle is updated based on the actual response power to achieve emergency frequency control of the virtual power plant.
[0007] In some embodiments, the frequency support capacity and price parameters of each virtual power plant are pre-calculated offline and stored in the database of the power grid control center, specifically including: Fuzzy sets are constructed based on moment information to quantify the uncertainty of new energy forecasting; With the goal of minimizing the overall system operating cost, five types of market services are integrated: electrical energy, backup, virtual inertia, fast primary frequency regulation, and slow primary frequency regulation. The safety constraints of frequency change rate, minimum fault frequency, quasi-steady-state frequency, and fuzzy sets are written into the model to construct a sub-Bruker optimization model. After equivalent reconstruction of the biblind bar optimization model, a two-stage step-by-step strategy is adopted to solve the problem and output the optimal electrical energy, reserve, inertia and frequency regulation clearing capacity for each time period. Based on the derivation of various service marginal prices using Lagrange dual multipliers, the frequency support capacity and price parameters of each virtual power plant are associated and stored in the database of the power grid control center.
[0008] In some embodiments, constructing a fuzzy set based on moment information specifically includes: Historical prediction error time series samples of wind power output, photovoltaic output, and wind turbine virtual inertia were collected, and wind power output prediction error vector, photovoltaic output prediction error vector, and wind turbine virtual inertia prediction error vector were constructed respectively. The expected values and covariance matrices of the wind power output prediction error vector, photovoltaic power output prediction error vector, and wind turbine virtual inertia prediction error vector are calculated respectively as reference moment information. By introducing the expected relaxation coefficient and the covariance amplification coefficient, a variable moment boundary is constructed that includes the upper and lower fluctuation range of the reference moment information; By utilizing variable moment boundaries and simultaneously introducing first-order moment uncertainty and second-order moment uncertainty, the wind power output prediction error vector, photovoltaic power output prediction error vector, and wind turbine virtual inertia prediction error vector are concatenated into a new energy joint prediction error vector, generating a fuzzy set that simultaneously contains moment uncertainty and coupling correlation between variables.
[0009] In some embodiments, the bipolar bar optimization model is equivalently reconstructed and then solved using a two-stage step-by-step strategy to output the optimal electrical energy, reserve, inertia, and frequency regulation clearing capacity for each time period, specifically including: The nonlinear frequency safety constraints and nonconvex opportunity constraints in the bibliometric bar optimization model are equivalently reconstructed. The nonlinear frequency safety constraints include the frequency change rate safety constraint, the minimum fault frequency safety constraint, and the quasi-steady-state frequency safety constraint. The frequency change rate safety constraint, the minimum fault frequency safety constraint, and the quasi-steady-state frequency safety constraint are transformed into second-order cone forms, and the non-convexity chance constraint is transformed into a linear deterministic constraint, thus transforming the sub-Bruker optimization model into a mixed-integer second-order cone programming model. The mixed-integer second-order cone programming model is decomposed into a unit combination sub-model and an economic dispatch sub-model, and solved using a two-stage step-by-step strategy: the first stage solves the unit combination sub-model, outputs and fixes the 0-1 start-stop state of conventional generator units; the second stage uses the 0-1 start-stop state determined in the first stage as the fixed boundary to solve the economic dispatch sub-model, and outputs the optimal electrical energy, reserve, inertia and frequency regulation clearing capacity for each time period.
[0010] In some embodiments, the step of using the frequency support capacity of each virtual power plant as the initial available capacity of each virtual power plant, calling frequency response parameters, and correcting the initial available capacity in conjunction with the current real-time operating status of each virtual power plant to obtain the real-time available frequency support capacity, determining the target support power of the virtual power plant based on the active power deficit, frequency security constraints, and real-time available frequency support capacity, and generating an emergency power support command specifically includes: The frequency support capacity of each virtual power plant is used as the initial available capacity of each virtual power plant. The frequency support capacity of each virtual power plant includes inertia support capacity, fast primary frequency regulation capacity, and slow primary frequency regulation capacity. By calling the frequency response parameters and combining the current online resource status, current output power, energy storage device charge status, and communication status in the real-time operating status data of the virtual power plant, the inertia support capacity, fast primary frequency regulation capacity, and slow primary frequency regulation capacity are corrected to obtain the real-time available frequency support capacity of the virtual power plant. Based on the active power deficit and frequency security constraints, the total support power required by the system is calculated. Combined with the real-time available frequency support capacity, the target support power of the virtual power plant is allocated proportionally, and an emergency power support command carrying the target power value and response time limit is generated.
[0011] In some embodiments, the allocation of the target support power of the virtual power plant by the virtual power plant aggregation layer based on the current operating status of each distributed resource and in conjunction with frequency response parameters, and the generation of corresponding resource adjustment instructions, specifically includes: The virtual power plant aggregation layer extracts the target support power of the virtual power plant from the emergency power support command, collects the current output, adjustable capacity, remaining regulation margin and communication status of each distributed resource within the virtual power plant, and allocates the target support power of the virtual power plant in combination with frequency response parameters and the unit regulation cost of each distributed resource. The fast frequency regulation power in the target support power of the virtual power plant is allocated to energy storage, photovoltaic units and asynchronous motors, and the slow frequency regulation power in the target support power of the virtual power plant is allocated to distributed units, electric vehicles and air conditioning loads, generating resource regulation commands for each distributed resource.
[0012] In some embodiments, the step of controlling the corresponding distributed resource execution power adjustment by the local resource control layer according to the resource adjustment instruction and outputting the corresponding actual response power specifically includes: The local resource control layer receives resource adjustment instructions sent by the virtual power plant aggregation layer. The resource adjustment instructions include the target adjustment power and response time requirements of the corresponding distributed resource. Obtain the preset frequency response parameters and local power grid frequency status data of the corresponding distributed resources. The preset frequency response parameters include response time, power ramp rate and output upper and lower limits. The power grid frequency status data includes frequency deviation and frequency change rate. Power regulation is performed based on resource adjustment instructions, response time, power ramp rate, output upper and lower limits, frequency deviation, and frequency change rate, and the corresponding actual response power is output.
[0013] In some embodiments, updating the emergency power support command for the next control cycle based on the actual response power specifically includes: The actual response power of each distributed resource after executing resource regulation commands is uploaded to the virtual power plant aggregation layer, which then aggregates the actual response power of the virtual power plant and sends it to the power grid control center. Based on the deviation between the actual response power and the target support power, the power grid control center adjusts the virtual power plant target support power for the next control cycle and generates an updated emergency power support command until the frequency returns to a safe range.
[0014] Secondly, the present invention provides a virtual power plant emergency frequency control system, comprising: The power grid control center is used to perform offline pre-calculation, collect real-time power grid operation status data, identify fault active power deficit, generate and issue emergency power support instructions, and receive the actual response power of virtual power plants uploaded by each virtual power plant aggregation layer, and iteratively correct the emergency power support instructions for the next control cycle. At least one virtual power plant aggregation layer is connected to the power grid control center. It is used to receive emergency power support instructions, collect frequency response parameters and the current operating status of each distributed resource, allocate the target support power of the virtual power plant, generate corresponding resource adjustment instructions and send them to the local resource control layer, and summarize the actual response power of the virtual power plant and upload it to the power grid control center. Several local resource controllers are connected to the virtual power plant aggregation layer. They are used to read resource adjustment instructions, control the corresponding distributed resources to perform power adjustment, output the corresponding actual response power, and upload it to the virtual power plant aggregation layer.
[0015] Thirdly, the present invention provides a computer device, including a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor executes the computer program to implement the steps of the above-described method for predicting the remaining life of the wind turbine pitch bearing throughout its entire life cycle.
[0016] Fourthly, the present invention provides a computer-readable storage medium storing a computer program that, when executed by a processor, implements the steps of the above-described method for predicting the remaining lifespan of a wind turbine pitch bearing throughout its entire life cycle.
[0017] The above technical solution has the following advantages or beneficial effects: Firstly, this invention provides an emergency frequency control method for virtual power plants. By establishing a centralized-decentralized hierarchical control architecture and setting up an information interaction and command transmission mechanism between the power grid control center, the virtual power plant aggregation layer, and the local resource control layer, it achieves hierarchical issuance of emergency power support commands and rapid execution on the resource side. This method pre-calculates the frequency support capacity and price parameters of each virtual power plant offline and stores them in the power grid control center's database. When a power grid fault with active power deficit is detected, the data is directly read and invoked. This ensures that the offline pre-calculation results can be effectively transmitted and utilized during the real-time emergency frequency regulation phase, avoiding the problem of low decision-making efficiency caused by the inability of offline calculation results to be directly used to guide the generation of real-time power support commands. Meanwhile, upon receiving emergency power support commands, the virtual power plant aggregation layer performs secondary allocation based on the current operating status and frequency response parameters of each internal distributed resource. Fast frequency regulation tasks are assigned to fast-response resources such as energy storage, photovoltaic units, and asynchronous motors, while slow frequency regulation tasks are assigned to slow-response resources such as distributed units, electric vehicles, and air conditioning loads. This ensures that the decomposition of frequency regulation tasks matches the actual response capabilities of each distributed resource, avoiding the mismatch between command allocation and response capabilities caused by insufficient consideration of the hierarchical differences in response speeds of heterogeneous resources. Furthermore, by transmitting the actual response power of each distributed resource layer by layer back to the grid control center and updating the emergency power support commands for the next control cycle accordingly, a closed-loop iterative correction mechanism is formed. This ensures continuous optimization of control commands during frequency recovery, effectively improving the response accuracy and overall frequency regulation effect of the virtual power plant participating in grid emergency frequency control.
[0018] Secondly, this invention provides a virtual power plant emergency frequency control system. Through a three-layer centralized-decentralized hierarchical architecture consisting of a power grid control center, a virtual power plant aggregation layer, and a local resource controller, it achieves efficient hierarchical issuance of emergency power support commands and rapid response execution of distributed resources. The power grid control center can read offline pre-calculation results and iteratively correct commands based on the actual response power. The virtual power plant aggregation layer can adapt and allocate resources according to the real-time status of resources, and the local resource controller achieves precise execution. The three work together to effectively improve the response speed and adjustment accuracy of the virtual power plant in participating in the power grid emergency frequency control.
[0019] Thirdly, the present invention provides a computer device that, through a processor executing a specific computer program, can efficiently implement the steps of the method of the present invention. When performing data processing tasks, the computer device can accurately perform numerical calculations and logical judgments, avoiding errors caused by human factors. At the same time, since the computer program has high stability and reliability, it can ensure the accuracy and consistency of the data processing results.
[0020] Fourthly, the present invention provides a computer-readable storage medium. By programming the steps of the method of the present invention into a computer program and storing it on the computer-readable storage medium, users can easily load these programs onto any compatible computer device and execute them without rewriting or converting the code, which greatly improves the convenience and flexibility of program execution. Attached Figure Description
[0021] Figure 1 This is a schematic diagram of a centralized-decentralized virtual power plant hierarchical control architecture as shown in an embodiment of the present invention; Figure 2 This is a schematic diagram of the predicted power output of new energy sources as shown in an embodiment of the present invention; Figure 3 This is a schematic diagram of the unit combination scheduling results in scenario 1 shown in the embodiment of the present invention; Figure 4 This is a schematic diagram of the unit combination scheduling results in scenario 2 of the present invention; Figure 5 This is a schematic diagram of the unit combination scheduling results in scenario 3 shown in the embodiment of the present invention; Figure 6 These are schematic diagrams of market clearing prices under different scenarios as shown in the embodiments of the present invention; wherein, (a) is a schematic diagram of electricity price, (b) is a schematic diagram of reserve price, (c) is a schematic diagram of inertia price, (d) is a schematic diagram of fast primary frequency regulation price, and (e) is a schematic diagram of slow primary frequency regulation price. Figure 7 This is a schematic diagram illustrating the system scheduling situation according to an embodiment of the present invention; Figure 8 This is a schematic diagram of the internal resource scheduling results of a virtual power plant as shown in an embodiment of the present invention; Figure 9 This is a schematic diagram of market prices under different renewable energy power generation levels as shown in the embodiments of the present invention; wherein, (a) is a schematic diagram of electricity price, (b) is a schematic diagram of reserve price, (c) is a schematic diagram of inertia price, (d) is a schematic diagram of fast primary frequency regulation price, and (e) is a schematic diagram of slow primary frequency regulation price. Figure 10 The diagrams illustrate the supply of inertia and frequency response services under different renewable energy power generation levels in the embodiments of the present invention; wherein, (a) is a diagram of inertia supply, (b) is a diagram of fast primary frequency regulation supply, and (c) is a diagram of slow primary frequency regulation supply. Detailed Implementation
[0022] The present invention will be further described in detail below with reference to specific embodiments. These descriptions are for explanation purposes only and are not intended to limit the scope of the invention.
[0023] To enable those skilled in the art to better understand the present invention, the technical solutions of the present invention will be clearly and completely described below with reference to the accompanying drawings of the embodiments of the present invention. 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 should fall within the scope of protection of the present invention.
[0024] It should be noted that the terms "first," "second," etc., in the specification, claims, and accompanying drawings of this invention are used to distinguish similar objects and are not necessarily used to describe a specific order or sequence. It should be understood that such data can be interchanged where appropriate so that the embodiments of the invention described herein can be implemented in orders other than those illustrated or described herein. Furthermore, the terms "comprising" and "having," and any variations thereof, are intended to cover a non-exclusive inclusion; for example, a process, method, system, product, or apparatus that comprises a series of steps or units is not necessarily limited to those steps or units explicitly listed, but may include other steps or units not explicitly listed or inherent to such processes, methods, products, or apparatus.
[0025] Current research on VPP (Voltage Power Utilization) mainly focuses on power grid planning, electricity markets, and economic dispatch, while further research is needed on emergency frequency control. Emergency control scenarios refer to the control process where, due to a fault in the regional power grid, there is an active power deficit, leading to a risk of frequency drop in the system. VPP aggregates multiple flexible resources and uses inertial response, primary frequency regulation, and other adjustment methods to supplement the deficit power, maintain grid frequency stability, and prevent misjudgments in emergency frequency control decisions. Emergency frequency control methods incorporating VPP face the following challenges: VPP aggregates a massive number of heterogeneous flexible resources, each with different control methods and response characteristics for frequency regulation. Constructing an accurate VPP frequency response model that reflects the characteristics of multiple flexible resources and evaluating VPP frequency response capabilities to support emergency frequency control of the power system is a significant challenge. Furthermore, the VPP frequency response model is complex and high-order with numerous control parameters, and these parameters are time-varying, varying under different power grid operating conditions. This further complicates the emergency frequency control decision-making problem involving VPP. Meanwhile, the VPP frequency support capacity and price parameters obtained in the offline pre-calculation stage lack an effective transmission and utilization mechanism in the real-time emergency frequency regulation stage, which restricts the decision-making efficiency of emergency frequency control.
[0026] Based on the above background, this application provides an emergency frequency control method for virtual power plants. This method establishes a centralized-decentralized hierarchical control architecture to achieve information interaction and command transmission between the power grid control center, the virtual power plant aggregation layer, and the local resource control layer. The frequency support capacity and price parameters of each virtual power plant are pre-calculated offline and stored in the power grid control center database. When a power grid fault with active power deficit is detected, the frequency support capacity and price parameters of each virtual power plant stored in the database are read, and the initial available capacity is corrected by combining frequency response parameters and the current real-time operating status. Emergency power support commands are generated and distributed and executed level by level. This achieves reasonable decomposition and allocation of frequency regulation tasks while fully considering the hierarchical differences in the response speed of heterogeneous resources. Simultaneously, the offline pre-calculation results can be effectively transmitted and utilized during the real-time emergency frequency regulation phase, improving the decision-making efficiency and execution effect of emergency frequency control.
[0027] Example 1: This embodiment provides an emergency frequency control method for a virtual power plant, which is applicable to emergency control scenarios when a regional power grid experiences an active power deficit due to a fault and the system frequency faces the risk of dropping.
[0028] S1. Establish a centralized-decentralized hierarchical control architecture, which includes a power grid control center, a virtual power plant aggregation layer, and a local resource control layer connected in sequence. The power grid control center includes a power grid control center database, and the frequency support capacity and price parameters of each virtual power plant are pre-calculated offline and stored in the power grid control center database.
[0029] In this embodiment, the frequency support capacity and price parameters of each virtual power plant are pre-calculated offline and stored in the database of the power grid control center, specifically including: A fuzzy set is constructed based on moment information to quantify the uncertainty of new energy forecasting. With the goal of minimizing the overall system operating cost, five types of market services—electricity, reserve, virtual inertia, fast primary frequency regulation, and slow primary frequency regulation—are integrated. Frequency change rate safety constraints, minimum fault frequency safety constraints, quasi-steady-state frequency safety constraints, and fuzzy sets are incorporated into the model to construct a sub-Blu-ray optimization model. After equivalent reconstruction of the sub-Blu-ray optimization model, a two-stage step-by-step strategy is used to solve it, outputting the optimal electricity, reserve, inertia, and frequency regulation clearing capacity for each time period. Marginal prices for various services are derived based on Lagrange dual multipliers, and the frequency support capacity and price parameters of each virtual power plant are associated and stored in the power grid control center database.
[0030] Specifically, constructing fuzzy sets based on moment information includes: Historical prediction error time-series samples of wind power output, photovoltaic output, and wind turbine virtual inertia were collected to construct prediction error vectors for wind power output, photovoltaic output, and wind turbine virtual inertia, respectively. The expected values and covariance matrices of the prediction error vectors for wind power output, photovoltaic output, and wind turbine virtual inertia were calculated as reference moment information. The expected relaxation coefficient and covariance amplification coefficient were introduced to construct a variable moment boundary that includes the upper and lower fluctuation ranges of the reference moment information. Using the variable moment boundary, first-order moment uncertainty and second-order moment uncertainty were introduced simultaneously. The prediction error vectors for wind power output, photovoltaic output, and wind turbine virtual inertia were concatenated into a new energy joint prediction error vector, generating a fuzzy set that simultaneously includes moment uncertainty and the coupling correlation between variables.
[0031] Specifically, after equivalent reconstruction of the sub-Bruker optimization model, a two-stage step-by-step strategy is adopted to solve it, outputting the optimal electrical energy, reserve, inertia, and frequency regulation clearing capacity for each time period, including: The nonlinear frequency security constraints and nonconvex opportunity constraints in the subbulb bar optimization model are equivalently reconstructed. The nonlinear frequency security constraints include frequency change rate security constraints, minimum fault frequency security constraints, and quasi-steady-state frequency security constraints. The frequency change rate security constraints, minimum fault frequency security constraints, and quasi-steady-state frequency security constraints are transformed into second-order cone forms, and the nonconvex opportunity constraints are transformed into linear deterministic constraints, thus transforming the subbulb bar optimization model into a mixed-integer second-order cone programming model. The mixed-integer second-order cone programming model is decomposed into a unit combination sub-model and an economic dispatch sub-model, and solved using a two-stage step-by-step strategy: the first stage solves the unit combination sub-model, outputs and fixes the 0-1 start-stop state of conventional generator units; the second stage uses the 0-1 start-stop state determined in the first stage as a fixed boundary to solve the economic dispatch sub-model, outputting the optimal electrical energy, reserve, inertia, and frequency regulation clearing capacity for each time period.
[0032] In this embodiment, the frequency support capacity and price parameters of each virtual power plant are pre-calculated offline and stored in the database of the power grid control center. This can also be referred to as the offline pre-calculation process for the dispatch cycle. This part is completed in advance during the normal dispatch cycle (when no grid fault occurs), providing capacity parameters, price signals, and reserved capacity for emergency control after a fault.
[0033] S11, based on moment information, constructs fuzzy sets to quantify the uncertainty of multiple predictions for new energy.
[0034] Specifically, new energy units include large wind turbines not aggregated by VPPs and distributed photovoltaic units aggregated by VPPs. Because there are discrepancies between the actual active power output of wind and solar power and the predicted values, it is necessary to construct a fuzzy set containing torque information and considering uncertainty correlations to describe the uncertainty of new energy prediction errors.
[0035] Specifically, a fuzzy set method based on moment information is adopted to collect time series samples of historical prediction errors of wind power output, photovoltaic power output, and wind turbine virtual inertia, and calculate the error benchmark expectation matrix and covariance matrix; set the expectation relaxation coefficient and covariance amplification coefficient to construct the variable moment interval, and generate a set of feasible probability distributions containing the coupling correlation of three types of random variables.
[0036] S12, combining frequency security constraints, build a multi-service joint clearing and splitting optimization model, and define the objective function and all constraints.
[0037] Specifically, the optimization objective function is to minimize the overall system operating cost; five types of market services are integrated, including electrical energy, reserve, virtual inertia, and fast and slow primary frequency regulation; three types of frequency safety limits, namely system frequency change rate, minimum fault frequency, and quasi-steady-state frequency, are written into the model as hard constraints, and unit operation constraints, distributed resource regulation constraints, and uncertainty opportunity constraints are superimposed to construct a complete distributed bar optimization model.
[0038] S13 performs an equivalent reconstruction of the original non-convex model and uses a two-stage step-by-step strategy to solve the reconstructed model.
[0039] Specifically, the nonlinear frequency safety constraints and nonconvex opportunity constraints within the model are equivalently reconstructed to transform the original model into a mixed integer second-order cone programming model. A two-stage solution process is adopted: the first stage solves the unit combination sub-model, outputs and fixes the 0-1 start-stop state of conventional generator units; the second stage locks the start-stop variables to solve the economic dispatch sub-model, outputting the optimal electrical energy, reserve, inertia, and frequency regulation clearing capacity for each time period.
[0040] S14 derives the market prices of various frequency modulation services based on Lagrange dual multipliers, completes simulation verification, and stores the frequency support capacity of each VPP.
[0041] Extract the Lagrange dual multipliers corresponding to the model constraints to derive the marginal service prices for power, reserve, virtual inertia, fast primary frequency regulation, and slow primary frequency regulation; build multi-scenario simulation examples to complete quantitative evaluation from the dimensions of market operation performance, adaptability to different proportions of renewable energy generation, and model effectiveness; store the frequency support capacity and price parameters of each virtual power plant in the database of the power grid control center, with the frequency support capacity serving as the reserved capacity before a fault and the upper limit of call after a fault, and the price parameters used for service settlement and performance evaluation.
[0042] In this embodiment, a centralized-decentralized hierarchical control architecture is established, which is triggered when a power grid fault causes an active power deficit, and the results of offline pre-calculation are used to achieve fast closed-loop control.
[0043] A three-tiered centralized-decentralized hierarchical control architecture—grid control center-VPP aggregation layer-local resource controller—is pre-built for emergency control scenarios, clarifying the communication and interaction rules for command issuance and data upload between layers; based on the frequency response characteristics of wind turbines, photovoltaics, energy storage, electric vehicle V2G, and air conditioning flexible loads, parameters such as communication latency, response time, power ramp rate, output upper and lower limits, and energy state safety range are extracted, and fast frequency regulation resources and slow frequency regulation resources are divided according to response time.
[0044] The power grid control center issues power control commands, system minimum frequency thresholds, and quasi-steady-state frequency thresholds based on power deficits and the frequency response capability of the VPP, and monitors the power grid frequency deviation in real time.
[0045] VPP aggregation layer: Composed of various VPP substations. Each substation receives the maximum support power, steady-state power, control response time, power ramp-up time and ramp-up rate of internal resources, and summarizes and reports the maximum regulating power, steady-state support power, ramp-up time and ramp-up rate of its own substation. At the same time, based on the power support amount issued by the control center and the operating status of each resource, the power command is allocated to various flexible resources within the station.
[0046] Local resource control layer: including distributed units, new energy units (grid-type), ES system (equipped with grid-type inverter and VSG control module), cluster EV (with cluster control module added), controllable air conditioning load (with frequency converter and load control module added), etc., which act as control terminals to perform power regulation.
[0047] Based on the frequency response characteristics of various resources, this embodiment extracts the following parameters: communication delay, response time, power ramp-up rate, output upper and lower limits, and energy state safety range. Resources are then divided into fast frequency regulation resources (response time ≤ 1s, including energy storage, photovoltaics, and asynchronous motors) and slow frequency regulation resources (response time ≤ 10s, including distributed units, EV clusters, and air conditioning loads) according to their response time. These parameters are pre-configured in the controllers at each level for use in case of faults.
[0048] S2 extracts the frequency response parameters of each distributed resource through the virtual power plant aggregation layer; obtains real-time power grid operation status data through the power grid control center; when a power grid fault active power deficit is detected, obtains the preset frequency security constraint; and reads the frequency support capacity and price parameters of each virtual power plant stored in the database of the power grid control center.
[0049] In this embodiment, the power grid control center collects real-time data on the entire network's frequency, regional load output, renewable energy power generation, and basic resource parameters of each VPP. It also monitors the system's rate of frequency change (RoCoF) and frequency deviation in real time.
[0050] In this embodiment, based on the system frequency change rate, minimum frequency, and quasi-steady-state frequency deviation in the real-time operating status data of the power grid, it is determined whether the power grid has experienced a frequency anomaly. When it is determined that a power grid fault has caused an active power deficit, the corresponding active power deficit is determined based on the state of the frequency anomaly, and preset frequency safety constraints are invoked.
[0051] In this embodiment, the power grid control center collects real-time frequency of the entire network, regional load output, power generation of new energy power plants, and basic resource parameters of each VPP in real time; it monitors the system frequency change rate in real time, and when a fault is detected that causes an active power deficit, it retrieves three types of safety limits preset by the system: maximum allowable RoCoF, minimum fault frequency, and quasi-steady-state frequency, and reads the offline clearing and capacity reference results obtained by module one.
[0052] S3 uses the frequency support capacity of each virtual power plant as the initial available capacity of each virtual power plant, calls the frequency response parameters, and corrects the initial available capacity in combination with the current real-time operating status of each virtual power plant to obtain the real-time available frequency support capacity. Based on the active power deficit, frequency security constraints and real-time available frequency support capacity, the target support power of the virtual power plant is determined, and an emergency power support command is generated.
[0053] In this embodiment, S3 specifically includes: S31, the frequency support capacity of each virtual power plant is used as the initial available capacity of each virtual power plant, wherein the frequency support capacity of each virtual power plant includes inertia support capacity, fast primary frequency regulation capacity and slow primary frequency regulation capacity.
[0054] S32 calls the frequency response parameters and, in conjunction with the current online resource status, current output power, energy storage device charge status, and communication status in the real-time operating status data of the virtual power plant, corrects the inertia support capacity, fast primary frequency regulation capacity, and slow primary frequency regulation capacity to obtain the real-time available frequency support capacity of the virtual power plant.
[0055] S33 calculates the total support power required by the system based on the active power deficit and frequency security constraints, and allocates the target support power of the virtual power plant proportionally according to the real-time available frequency support capacity, and generates an emergency power support command carrying the target power value and response time limit.
[0056] In this embodiment, the power grid control center reads the pre-stored aggregated frequency regulation capability parameters of each virtual power plant and calls the pre-configured frequency response parameters such as power ramp rate, response time, and communication delay. Combining the current output of each VPP, the online status of the equipment, the energy storage status, the adjustable margin, and the communication status, the real-time available capacity at the time of the fault is corrected. The control center combines the frequency safety limit and the equivalent inertia required by the fault active power deficit calculation system, as well as the total support power for fast and slow frequency regulation, to generate an emergency power support command carrying the frequency safety threshold, target regulation power, and frequency regulation response time limit.
[0057] S4 sends the emergency power support command to the virtual power plant aggregation layer. The virtual power plant aggregation layer allocates the target support power of the virtual power plant based on the current operating status of each distributed resource and the frequency response parameters, and generates the corresponding resource adjustment command.
[0058] In this embodiment, S4 specifically includes: The virtual power plant aggregation layer extracts the target support power of the virtual power plant from the emergency power support command, collects the current output, adjustable capacity, remaining regulation margin and communication status of each distributed resource within the virtual power plant, and allocates the target support power of the virtual power plant in combination with frequency response parameters and the unit regulation cost of each distributed resource. The fast frequency regulation power in the target support power of the virtual power plant is allocated to energy storage, photovoltaic units and asynchronous motors, and the slow frequency regulation power in the target support power of the virtual power plant is allocated to distributed units, electric vehicles and air conditioning loads, generating resource regulation commands for each distributed resource.
[0059] In this embodiment, the VPP aggregation layer receives an emergency power support command and simultaneously collects the current output, adjustable capacity, and remaining adjustment margin of all distributed resources within the station. Combining the preset frequency regulation response speed and unit adjustment cost of each resource, the total target power is allocated hierarchically. The 1-second fast frequency regulation task is allocated to energy storage, photovoltaic units, and asynchronous motors, while the 10-second slow frequency regulation task is allocated to distributed units, electric vehicles, and temperature-controlled air conditioners. This generates detailed power adjustment commands for each device.
[0060] S5 sends the resource adjustment command to the local resource control layer, which then controls the corresponding distributed resource to adjust its execution power according to the command and outputs the corresponding actual response power.
[0061] In this embodiment, S5 specifically includes: S51, the local resource control layer receives a resource adjustment instruction sent by the virtual power plant aggregation layer. The resource adjustment instruction includes the target adjustment power and response time requirement of the corresponding distributed resource.
[0062] S2, obtain the preset frequency response parameters and local power grid frequency status data of the corresponding distributed resource. The preset frequency response parameters include response time, power ramp rate and output upper and lower limits. The power grid frequency status data includes frequency deviation and frequency change rate.
[0063] S3 performs power regulation based on resource adjustment instructions, response time, power ramp rate, output upper and lower limits, frequency deviation, and frequency change rate, and outputs the corresponding actual response power.
[0064] In this embodiment, each distributed resource is equipped with a local controller that reads preset frequency response parameters and detailed power adjustment instructions, and performs power adjustment in combination with locally measured frequency deviation and frequency change rate. When inertia and fast primary frequency regulation resources meet preset trigger conditions, they automatically respond. Distributed units with inertial response capabilities output inertia support, and energy storage, photovoltaics, controllable flexible loads, distributed units and electric vehicles output primary frequency regulation active power to collaboratively compensate for the active power deficit during grid faults.
[0065] S6 updates the emergency power support command for the next control cycle based on the actual response power, thereby realizing emergency frequency control of the virtual power plant.
[0066] In this embodiment, S6 specifically includes: S61, upload the actual response power of each distributed resource after executing the resource adjustment command to the virtual power plant aggregation layer, and the virtual power plant aggregation layer summarizes the actual response power of the virtual power plant and sends it to the power grid control center; S62, the power grid control center adjusts the virtual power plant target support power for the next control cycle based on the deviation between the actual response power and the target support power, and generates an updated emergency power support command until the frequency is restored to a safe range.
[0067] In this embodiment, the underlying local controller uploads the actual adjusted output of resources to the VPP aggregation layer in real time. All VPP aggregation layers summarize the total internal frequency regulation output and upload it to the power grid control center. The control center compares the real-time monitored frequency with the preset safety threshold, iteratively corrects the emergency power support command for the next control cycle, and cyclically completes the fault-layered closed-loop frequency regulation control. The actual response speed, actual delivered power, and command tracking deviation of each VPP recorded during the fault period, together with the service price obtained from Module 1, are used for response service settlement and performance evaluation, and are also used to correct the VPP available capacity parameters for the next scheduling cycle.
[0068] This embodiment utilizes information interaction and command transmission between the power grid control center, the virtual power plant aggregation layer, and the local resource control layer in a centralized-decentralized hierarchical control architecture. The frequency support capacity of each virtual power plant, calculated offline, is stored in the power grid control center's database and read and corrected after a fault occurs. Emergency power support commands are then generated and distributed level by level to the virtual power plant aggregation layer and the local resource control layer for execution. The commands for the next control cycle are updated based on actual response power feedback. This achieves a reasonable decomposition and allocation of frequency regulation tasks while fully considering the hierarchical differences in the response speeds of heterogeneous resources. Simultaneously, the offline pre-calculation results are effectively transmitted and utilized during the real-time emergency frequency regulation phase, improving the decision-making efficiency and execution effect of emergency frequency control.
[0069] Example 2: Based on Embodiment 1 above, this embodiment provides a detailed explanation of the virtual power plant multi-market joint clearing and pricing method, which involves pre-calculating the frequency support capacity and price parameters of each virtual power plant offline and storing them in the power grid control center database. This method can be used in the pre-calculation stage before a fault in emergency frequency control to optimize the configuration of the frequency support capacity and determine the service price of each virtual power plant, providing pre-stored frequency support capacity and price parameters for the control method described in Embodiment 1.
[0070] This embodiment proposes a centralized-decentralized hierarchical architecture for VPP participation in emergency frequency control, comprising two control layers: a VPP aggregation layer and a local control layer. See the architecture diagram below. Figure 1 , contains A VPP subsite.
[0071] VPP aggregates various adjustable DERs, including distributed generation units, renewable energy units, ES systems, EVs, controllable loads, and other DERs that can provide inertial or primary frequency response. Distributed generation units need to remain online, relying on their own mechanical inertia to provide inertial response, and converting mechanical kinetic energy into electrical energy through existing speed governors to achieve primary frequency response. New energy generating units must be configured as grid-connected types; the ES system must be equipped with grid-connected inverters, virtual synchronous generator control modules, and state of charge monitoring and control modules to achieve primary frequency response through power regulation. The frequency regulation power generated by its droop control must be within the capacity range set by the VPP, and the use of energy storage power and energy margin must be coordinated between power dispatch and frequency response services. Individual EVs cannot participate independently due to their small capacity, so they are aggregated by the VPP. A cluster control module and a charging dispatch system must be added to coordinate the charging behavior of multiple EVs and connect to the VPP platform to provide primary frequency response services. Adjustable loads such as variable frequency air conditioners must have continuously adjustable characteristics. Inverters and load control modules must be added to change power consumption by adjusting the operating frequency of the motor or compressor, thereby achieving primary frequency response.
[0072] The power grid control center issues specific power control command values based on the power deficit and the frequency response capability of the VPP. System minimum frequency threshold and quasi-steady-state frequency threshold Real-time monitoring of grid frequency deviation and real-time acquisition of the maximum regulation power that all VPPs within the grid can provide. Steady-state support power The lowest system frequency after VPP participates in emergency frequency control according to the given power command under the anticipated fault condition. and quasi-steady-state frequency .
[0073] The main body of the VPP aggregation layer consists of each VPP substation. Each VPP substation receives the maximum support power that each control resource can provide. steady-state power Control response time Power ramp-up time and climbing rate And accordingly, report the maximum regulating power that this substation can provide. Steady-state support power Climbing time and climbing rate On the other hand, based on the power support amount issued by the power grid control center The VPP substation will display the power command value based on the operating status of each control resource within the substation. Resources are allocated flexibly and centrally within the station.
[0074] The main components of the local control layer include distributed generating units, new energy generating units, ES power stations, clustered EVs, and controllable loads. When a regional active power deficit fault occurs, these control resources, acting as control terminals, will participate in emergency frequency control to provide power support and ensure grid frequency security.
[0075] The power grid control center needs to issue power support instructions for emergency frequency control scenarios. These instructions include both the system's minimum frequency threshold and the quasi-steady-state frequency threshold. This is frequency safety control for active power deficits under fault scenarios, requiring the VPP to ensure that the minimum frequency and quasi-steady-state frequency meet safety constraints under a given support power. Emergency frequency control after a fault not only allocates power but also simultaneously meets safety requirements such as the minimum frequency and quasi-steady-state frequency.
[0076] The US electricity market has widely adopted a joint clearing mechanism for the energy and reserve markets, with the New York Independent System Operator (NYISO) being a prime example. Because the available capacity of frequency response services is closely related to the unit configuration, it has a significant coupling relationship with energy and reserve services. Therefore, unified clearing of energy, reserve, and frequency response services can effectively improve the overall operating efficiency of the system.
[0077] This invention constructs a multi-market joint framework involving Virtual Power Providers (VPPs). Within a VPP, various types of heterogeneous distributed resources are aggregated, including distributed generator units, renewable energy units, ES systems, electric vehicles (EVs), and flexible loads, and internally coordinated and scheduled based on submitted operating parameters. At the market level, VPPs and conventional synchronous generators, as independent market entities, participate in the energy market, reserve market, and frequency ancillary services market, submitting their available aggregated capacity to the system operator. The joint clearing result must satisfy the overall feasible operating domain constraints of the VPP to ensure the feasibility of internal resource scheduling. Under a given emergency control scenario, the system operator collects bidding information from various market participants, performs unified optimization and clearing of energy, reserve, inertia, and frequency regulation services, and feeds back the scheduling plan and market settlement results to each participant. Through this mechanism, VPPs can participate in multi-type frequency service transactions in an aggregated form, thereby achieving the collaborative utilization of distributed resources at the system level.
[0078] In this embodiment, a fuzzy set based on moment information is first established to quantify the uncertainty of the new energy power output prediction error. The new energy units of this invention are divided into large wind turbine units not aggregated by VPPs and distributed photovoltaic units aggregated by VPPs. There is a deviation between the actual active power output of wind and photovoltaic units and the predicted values, resulting in prediction errors with uncertainty. Taiwan large wind turbine units The random value of active power output at time t, the first The random values of active power output of the polymerized photovoltaic within each VPP can be expressed as follows: (1); (2); In the formula: It is the first Taiwan large wind turbine units Actual power at any given moment; It is the first Taiwan large wind turbine units Power prediction at any given time; It is the first Taiwan large wind turbine units Power prediction error at time step; It is the first A VPP polymer photovoltaic in Actual power at any given moment; It is the first A VPP polymer photovoltaic in Power prediction at any given time; It is the first A VPP polymer photovoltaic in Power prediction error at time t.
[0079] Will time The prediction error of the typhoon generator sets constitutes a column vector. Summing all its elements yields the total wind power prediction error. Similarly, time The prediction errors of each VPP polymerized photovoltaic unit constitute a column vector. The total prediction error of photovoltaic power is obtained by summing the results. Furthermore, the prediction error vectors of wind power and photovoltaics are concatenated to obtain the joint prediction error vector for new energy sources. The total prediction error of the system is the sum of the total prediction errors of wind power and photovoltaic power, that is... .
[0080] This invention employs an affine control strategy to characterize the actual power output characteristics of a synchronous generator. The active power output setpoint of the synchronous generator is an affine function of the total prediction error of the system. The unit can adjust its output proportionally to the total prediction error of new energy sources through an adjustable participation factor. The random values of active power output for conventional large synchronous generators and aggregated distributed units within a VPP can be expressed as follows: (3); (4); In the formula: It is the first A conventional synchronous generator in Actual power output at any given time; It is the first A conventional synchronous generator in The planned power value at any given time; It is the first A conventional synchronous generator in Adjustable participation factors at any given time; It is the first Within a VPP, aggregated distributed units Actual power output at any given time; It is the first Within a VPP, aggregated distributed units The planned power value at any given time; It is the first Within a VPP, aggregated distributed units Adjustable participation factors at any given time.
[0081] Large wind turbines equipped with power controllers possess the technical potential to provide virtual inertia and rapid frequency regulation support, making them a crucial resource for ensuring system frequency security. However, due to the randomness of wind speed, the available inertia capacity of wind turbines is uncertain; therefore, it is... The maximum inertia constant at time t is modeled as a random variable, and its expression is: (5); In the formula: It is the first Taiwan large wind turbine The actual value of the moment of inertia at any given time; It is the first Taiwan large wind turbine Predicted inertia value at time; It is the first Taiwan large wind turbine The random error term of inertia at time t.
[0082] The system power injection uncertainty considered in this invention originates solely from the prediction error of new energy output. In the bibliometric optimization model, the probability distribution of uncertain variables is usually unknown, and their distribution boundaries are difficult to accurately characterize; therefore, there is no need to predefine a specific distribution form. Existing research often uses fuzzy sets based on fixed moments to describe the probability distribution of uncertain variables, but the fixed moment parameters are difficult to accurately match the statistical characteristics of actual uncertainties. Therefore, this invention constructs a fuzzy set containing variable moment information and considering the correlation of uncertainties, its general form being: (6); In the formula: It is a fuzzy set of uncertain variables, used to describe the set of all possible probability distributions; It is an uncertain variable The probability density function; It is an uncertain variable The first moment; It is the expected baseline value; It is a pre-defined covariance matrix; It is the relaxation coefficient of the expected deviation; It is an uncertain variable The actual covariance matrix; It is the covariance amplification factor.
[0083] The main causes of frequency security events in power systems are sudden load surges or sudden generator outages, which can lead to an imbalance in the system's active power. The system's dynamic frequency response process can be divided into two stages: inertial response and primary frequency response. At the moment of disturbance, the inertial response initiates first, its core function being to suppress the increase in the rate of frequency change. Some resources are equipped with control loops whose response speeds are much faster than those of turbine generator governors and valve mechanisms, while the response delays of governors and valve mechanisms are relatively greater. Therefore, this embodiment divides primary frequency regulation into fast primary frequency regulation and slow primary frequency regulation based on the resource's response speed, resulting in different service pricing. This embodiment assumes that both fast and slow primary frequency regulation exhibit a linear growth characteristic with a preset ramp rate, where fast primary frequency regulation... The output of the full response capacity is completed at all times, while slow primary frequency modulation is completed in... Achieve full response at all times.
[0084] Following a frequency security event, the combined frequency response characteristics of all market participants can be described by a swing equation, expressed as: (7); In the formula: yes The total inertia of the system at any given moment; It is the system's rated frequency; yes Frequency deviation of the time system; yes Total capacity of the slow primary frequency modulation system; yes Total capacity of the fast primary frequency modulation system; yes The net power deficit that occurs in the system at any given time.
[0085] Typically, three thresholds—frequency change rate, frequency extreme level, and quasi-steady-state frequency level—are used as safety standards for the dynamic evolution of system frequency. Therefore, the dynamic frequency safety constraints corresponding to these three thresholds collectively determine the minimum demand for inertia and primary frequency modulation (PMM) auxiliary services. Based on the established system dynamic frequency response model, the minimum demand for inertia and PMM auxiliary service resources required to meet the dynamic frequency safety constraints can be calculated.
[0086] (1) Frequency change rate constraint.
[0087] At the initial moment of a frequency security event, the system frequency change rate is determined only by two factors: the net power deficit and the total system inertia. The constraint expression is as follows: (8); In the formula: It is the safe upper limit threshold for the system frequency change rate; yes The dual variables (Lagrange multipliers) of the RoCoF safety constraint at time t.
[0088] (2) Frequency extreme value level constraint.
[0089] After a severe disturbance occurs, to prevent the system frequency from dropping to the low-frequency load shedding threshold and triggering a cascading power outage, strict constraints must be imposed on the frequency valley value. Considering the system inertia and the dynamic characteristics of fast and slow primary frequency regulation, the constraint expression for the frequency valley value is as follows: (9); In the formula: This is the maximum permissible amplitude of the system frequency drop; This refers to the delivery time of the enhanced FM service; This refers to the delivery time of a frequency modulation service.
[0090] It should be noted that this equation represents a nonlinear constraint. To facilitate subsequent optimization and dual variable analysis, it can be further transformed into a second-order cone form. The transformed standard second-order cone programming form is as follows: (10); In the formula: The first vector component of the second-order cone constraint at the lowest frequency point is in The dual variable at time (Lagrange multiplier); The second vector component of the second-order cone constraint at the lowest frequency point is in The dual variable at time (Lagrange multiplier); The scalar part on the right end of the second-order cone constraint at the lowest frequency point is... The dual variable at time (Lagrange multiplier).
[0091] The dual variable corresponding to this second-order cone constraint must satisfy the following condition: (11); (3) Quasi-steady-state frequency level constraint.
[0092] Once the system frequency drops to its lowest point, the fast primary frequency regulation (CFM) and slow primary frequency regulation (SFM) capacities begin to continuously provide active power support to the system. Their combined response capacity must be greater than the system's power loss, and the constraint expression is as follows: (12); The inertia support service for the power system can be provided jointly by conventional synchronous generators, large wind turbines, and distributed units aggregated by VPPs. The formula for calculating the total system inertia is: (13); In the formula: It is the first The inertial time constant of a conventional synchronous generator; It is the first The upper limit of the power output of a conventional synchronous generator; It is the first A conventional synchronous generator in The binary variable representing the running state at any given moment; It is the first Inertial time constant of aggregated distributed units within a VPP; It is the first The maximum power output of aggregated distributed units within a VPP; It is the equivalent inertia time constant associated with disturbances or power loss events.
[0093] The fast primary frequency regulation service is provided collaboratively by large wind turbines and VPP-aggregated asynchronous motors, ES units, and photovoltaic units. The total fast primary frequency regulation capacity of the system is calculated as follows: (14); In the formula: It is the first Taiwan large wind turbine units Enhanced frequency modulation capacity provided at all times; It is the first Each VPP internal polymer photovoltaic unit in Enhanced frequency modulation capacity provided at all times; It is the first Agile asynchronous motors within VPPs Enhanced frequency modulation capacity provided at all times; It is the first The VPP contains polymer ES units in Enhanced frequency modulation capacity provided at all times.
[0094] Conventional synchronous generators are the core supporting resource for slow primary frequency regulation services. In addition, distributed units, EV clusters, and air conditioning loads aggregated by VPP also have the technical capability to provide slow primary frequency regulation services. The formula for calculating the total slow primary frequency regulation capacity of the system is as follows: (15); In the formula: It is the first A conventional synchronous generator in The frequency modulation capacity provided at any given time; It is the first Within a VPP, aggregated distributed units The frequency modulation capacity provided at any given time; It is the first Within a VPP, aggregated EV clusters The frequency modulation capacity provided at any given time; It is the first The aggregated air conditioning load within a VPP is The frequency modulation capacity provided at any given time.
[0095] In this embodiment, a split-bar decision-making model is used to address the participation of VPPs in the energy, reserve, and frequency response markets in emergency control scenarios. The model aims to minimize the total operating cost of resources within the scheduling cycle to achieve joint market clearing, and the pricing strategy is derived based on the Lagrange duality theorem and KKT conditions.
[0096] All market participants jointly participate in the electricity, reserve, and frequency response markets. The core objective of clearing is to minimize the total resource operating cost within the scheduling cycle. The objective function is constructed as follows: (16); (17); (18); (19); In the formula: It is the total operating cost of resources within the scheduling period; This is the total operating cost of a conventional synchronous generator (for other market participants); This is the total operating cost of a large wind turbine unit; This is the total operating cost of the VPP; It is the first The unit power generation cost coefficient of a conventional synchronous generator; It is the first The no-load operating cost of a conventional synchronous generator; It is the first The cost factor for a unit of slow primary frequency regulation capacity provided by a conventional synchronous generator; It is the first Cost coefficient per unit upward standby capacity of a conventional synchronous generator; It is the first A conventional synchronous generator in Upward reserve capacity at all times; It is the first Cost coefficient per unit of downside reserve capacity of a conventional synchronous generator; It is the first A conventional synchronous generator in Downward reserve capacity at any time; It is the first A conventional synchronous generator in The cost of powering on at any given moment; It is the first A conventional synchronous generator in Downtime costs at all times; It is the first The cost factor for providing the inertial time constant of a large wind turbine unit; It is the first The cost factor for a unit of rapid primary frequency regulation capacity of a large wind turbine unit; It is the first Unit power generation cost coefficient of aggregated distributed units within a VPP; It is the first Unit upward standby capacity cost factor for aggregated distributed units within a VPP; It is the first A VPP in Upward reserve capacity at all times; It is the first Unit downside standby capacity cost factor for aggregated distributed units within a VPP; It is the first A VPP in Downward reserve capacity at any time; It is the first The cost factor for a unit of slow primary frequency regulation capacity is provided by the aggregated distributed units within a VPP; It is the first The cost factor for a unit of slow primary frequency regulation capacity is provided by the aggregated EV cluster within a VPP. It is the first The cost factor for providing a unit slow primary frequency regulation capacity within a VPP aggregated air conditioning load; It is the first Each VPP (Voltage Power Plant) provides a cost factor per unit of fast primary frequency regulation capacity; It is the first The cost factor for a unit of fast primary frequency regulation capacity is provided by the aggregated asynchronous motor within a VPP; It is the first Each VPP aggregated ES unit provides a cost factor per unit of fast primary frequency modulation capacity.
[0097] The operational constraints for each market participant are as follows: (1) Operating constraints of conventional synchronous generators.
[0098] Conventional synchronous generators provide power systems with reserve services, inertial support, and slow primary frequency regulation services. Therefore, their operation must meet reserve capacity limits, slow primary frequency regulation capacity limits, and other operational constraints.
[0099] Reserve supply constraints. To ensure sufficient regulation capacity under uncertain power disturbances, conventional synchronous generator sets must provide both upward and downward reserve capacity.
[0100] (20); (twenty one); (twenty two); (twenty three); (twenty four); In the formula, It is the first The upper limit of the standby power of a conventional synchronous generator; It is the first The lower limit of the standby power of a conventional synchronous generator.
[0101] Generator output limits. The output power of the generator set, together with the allocated values for reserve capacity and maximum generating frequency limits, shall meet the following operating constraints: (25); (26); (27); Climbing constraints: (28); (29); In the formula: It is the first The maximum power of the conventional synchronous generator is reduced to the lowered ramp rate limit; It is the first The maximum power of the conventional synchronous generator is increased to the maximum ramp rate limit.
[0102] Startup cost constraints: (30); (31); Downtime cost constraints: (32); (33); (2) Wind turbine operation constraints.
[0103] Large wind turbines provide enhanced frequency response services by reserving an active power margin. This margin must be within the allowable curtailment power range and meet uncertainty constraints at the confidence level. The constraints are as follows: (34); (35); In the formula: It is the first Taiwan large wind turbine units Wind curtailment power at all times; It is the first Maximum curtailment rate coefficient for large-scale wind turbine units; It is the risk tolerance of the power margin constraint of wind turbine units.
[0104] The inertia capacity that a wind farm can provide is non-negative, and at a specified confidence level, the dispatched inertia capacity must not exceed its actual maximum available inertia capacity: (36); (37); In the formula: It is the first Taiwan large wind turbine units The inertia provided at any given moment.
[0105] (3) VPP operation constraints.
[0106] VPP can integrate various flexible resources, and the operational limitations of each resource are as follows: The output power of distributed generation units is limited by their generating capacity. These units are capable of simultaneously serving as backup power sources and restoring power frequencies.
[0107] (38); (39); (40); (41); (42); (43); (44); (45); (46); In the formula: It is the first Reserve power limit for aggregated distributed units within a VPP; It is the first Reserve power limit of aggregated distributed units within a VPP; It is the first The upper limit of slow primary frequency regulation capacity of aggregated distributed units within a VPP.
[0108] Photovoltaic units contribute to primary frequency regulation through grid-connected or grid-linked droop control, and the available primary frequency regulation capacity is limited by the reduced power margin.
[0109] (47); (48); In the formula: It is the first Each VPP internal polymer photovoltaic unit in Constant power consumption; It is the first Maximum curtailment rate coefficient for each VPP polymer photovoltaic unit; It is the risk tolerance for curtailment power constraints of polymerized photovoltaic units within the VPP.
[0110] The asynchronous motor is modeled as a controllable motor-driven load that can provide rapid primary frequency regulation by temporarily reducing power.
[0111] (49); In the formula: It is the first The rated power of the asynchronous motor aggregated within each VPP.
[0112] The ES system participates in both energy dispatch and the provision of frequency regulation reserve capacity, and its operational constraints are as follows: (50); (51); (52); (53); (54); (55); In the formula: It is the first Within a VPP, aggregated EV clusters The charging power at any given time; It is the first Within a VPP, aggregated EV clusters Maximum charging power limit at any given time; It is the first Within a VPP, aggregated EV clusters Discharge power at any given moment; It is the first Within a VPP, aggregated EV clusters Maximum discharge power limit at any given time; It is the first Within a VPP, aggregated EV clusters The maximum capacity of slow primary frequency modulation at any given moment; It is the first Within a VPP, aggregated EV clusters The energy of a moment; It is the first Within a VPP, aggregated EV clusters The change in battery level due to changes in the EV's grid connection status at any given moment; It is the first Charge and discharge efficiency of aggregated EV clusters within a VPP; It is the time step of the scheduling period; It is the first Within a VPP, aggregated EV clusters The lower limit of energy at any given time; It is the first Within a VPP, aggregated EV clusters The upper limit of energy at any given moment.
[0113] The air conditioning load participates in frequency regulation through controllable power adjustment. Its charging equivalent power limit and primary frequency regulation capacity are as follows: (56); (57); (58); (59); (60); (61); In the formula: It is the first The aggregated air conditioning load within a VPP is Power consumption at any given moment; It is the first The upper limit of power consumption of aggregated air conditioning load within a VPP; It is the first The upper limit of slow primary frequency regulation capacity of aggregated air conditioning load within a VPP; It is the first The aggregated air conditioning load within a VPP is Energy state at any given moment; It is the first State retention factor of aggregated air conditioning load energy within a VPP; It is the first Energy conversion efficiency of aggregated air conditioning load within VPP; It is the first External disturbance terms of aggregated air conditioning load within a VPP; It is the first The aggregated air conditioning load within a VPP is The upper limit of energy at any given moment.
[0114] (4) Power balance constraints The net power injection of VPP is defined as: (62); Positive values indicate that power is injected into the system, while negative values correspond to purchasing electricity from the grid.
[0115] At the system level, the power balance requirement is that the total load demand equals the total power generation provided by conventional generators, wind farms, and VPPs. (63); After separating the deterministic and uncertain components, the equivalent constraint on power balance is: (64); (65); (66); In the formula: Is the backup participation factor balance constraint in The dual variable at time (Lagrange multiplier); and These represent the uncertainty bias (or extreme scenario probability / weight) allocated to conventional units and VPPs, respectively.
[0116] The solution method is as follows: Compared to deterministic market joint clearing models based on power prediction, joint clearing models based on uncertain power fuzzy sets introduce random variables, increasing computational complexity. The chance-constrained sub-Brussels bar optimization model established in the previous section is a non-convex problem, and its compact mathematical form is: (67); Because the chance constraint includes random variables Such constraints are inherently non-convex and require equivalent reconstruction. Based on the equivalent transformation method of the split-bar chance constraint, the above chance constraint can be transformed into a convex constraint, in the form of: (68); in The safety factor for the fuzzy bar is calculated based on the fuzzy set parameters and the confidence level, using the following formula: (69); Based on this convex transformation method, the random constraints of the split bar in the model can all be transformed into equivalent deterministic convex constraints.
[0117] The pricing framework proposed in this embodiment adopts a two-stage optimization solution process, realizing unit combination decision-making and marginal pricing calculation in stages, ensuring the efficiency of model solution and the economy of pricing.
[0118] The first stage involves solving the unit combination model with frequency constraints. Under the premise of satisfying grid constraints and frequency security constraints, the optimal unit start-up and shutdown state is determined with the goal of minimizing the total system operating cost. This state is then passed on as a boundary condition to the subsequent pricing stage.
[0119] The second stage involves fixing the unit combination variables and solving the economic dispatch model with frequency constraints. Compared to the unit combination model with frequency constraints, this stage eliminates start-up and shutdown costs and their corresponding constraints, treating unit start-up and shutdown decisions as known parameters. This allows the model to achieve marginal pricing according to economic dispatch principles, thus establishing reasonable price signals for various service products in the market.
[0120] During the clearing process, the prices of various services can be obtained using Lagrange multipliers, and the settlement price can be represented by the partial derivative of the Lagrange function with respect to the required service. Based on the Lagrange function, the shadow prices of different services can be obtained.
[0121] The price of electricity and the price of the reserve participation factor are determined by the dual variables of the system power balance constraint and the reserve allocation constraint, respectively.
[0122] The price of inertia is determined by the marginal cost of frequency safety constraints, comprehensively reflecting the demand for inertia resources from both frequency change rate constraints and frequency valley constraints. The calculation formula is as follows: (70); In the formula: yes The price of system inertia at any given moment.
[0123] Both the enhanced frequency response price and the primary frequency response price are determined by the combined marginal effect of the minimum frequency constraint and the quasi-steady-state frequency constraint, reflecting the marginal value of the two types of frequency response resources, respectively. The calculation formulas are as follows: (71); (72); In the formula: yes The price of a single, rapid frequency adjustment; yes The price of a single slow frequency modulation at any given time; yes The dual variable (Lagrange multiplier) corresponding to the quasi-steady-state frequency security constraint at any given time.
[0124] This invention establishes a clear information exchange and command transmission mechanism between the system control center, the VPP aggregation layer, and the local control layer through a centralized-distributed hierarchical control architecture involving virtual power plants in emergency frequency control. This enables the hierarchical issuance of power support commands and rapid execution on the resource side. For typical resources such as distributed units, new energy units, ES, EV clusters, and controllable loads, considering factors such as communication latency, control response time constant, ramp-up process, and capacity boundaries, the invention models the frequency response characteristics of these resources, providing a model foundation for the aggregation layer's capacity summarization and control allocation. Building upon this foundation, a multi-market joint clearing framework for energy, reserve, and frequency response services involving VPPs is further established. A fuzzy set based on moment information is introduced to describe the uncertainty of new energy prediction errors. Three types of security constraints—frequency change rate, frequency minimum point, and quasi-steady-state frequency—are used to characterize the system's emergency frequency security requirements. Feasible optimized clearing and marginal pricing calculations are achieved through model reconstruction and a two-stage solution process, forming price signals for frequency security services such as inertia and primary frequency regulation. The time-segmented inertia, fast primary frequency regulation, and slow primary frequency regulation capacity of each virtual power plant determined by the joint clearing serve as the reserved capacity before a fault and the upper limit for call-up after a fault. The corresponding price parameters are used for capacity compensation, actual response settlement, and performance evaluation.
[0125] Example 3: Based on the above embodiments, this embodiment systematically evaluates the market performance, analyzes the impact of new energy power generation levels, and verifies the model effectiveness of the virtual power plant multi-market joint clearing and pricing method provided in each of the above embodiments 2.
[0126] Referring to the types and scale of resources aggregated in a virtual power plant, and combining the basic structure of the VPP constructed in this paper, the internal resource capacity is appropriately adjusted and simplified to form the resource allocation scheme in the example. In this example, the VPP aggregates diverse resources such as distributed generators, asynchronous motors, ES units, photovoltaic units, EV clusters, and air conditioning loads. The data used in the example are shown in Tables 1 and 2, and the renewable energy output is as follows: Figure 2As shown, the system's rated frequency is set to 50Hz, and the simulation assumes that the power disturbance in each time period is 5% of the load. Regarding frequency safety constraints, the lowest frequency point is 49.2Hz, and the maximum frequency change rate is limited to 1Hz / s. Fast primary frequency regulation and slow primary frequency regulation must complete full response output within 1s and 10s, respectively. The proposed pricing framework explicitly considers the impact of renewable energy prediction errors, setting the tolerance parameters for each risk constraint to 0.05, and fixing the fuzzy set parameters to 0.1 and 1.1, respectively. Monte Carlo simulation is used to generate samples of renewable energy prediction uncertainties, assuming that the prediction error follows a Gaussian distribution with a mean of 0 and a standard deviation of 5% of the corresponding predicted value. This assumption also applies to the prediction of photovoltaic output and the inertia of large wind turbines. All simulations were performed on a workstation equipped with an AMD Ryzen 7 7840H processor (3.80 GHz) and 32GB of memory, running Windows 11. The model was implemented using MATLAB R2019a and solved using Gurobi 11.0.
[0127]
[0128]
[0129] To analyze the pricing effects under different service participation scenarios, three comparative scenarios were designed: Scenario 1: Only synchronous generators provide inertia, fast primary frequency regulation, and slow primary frequency regulation services; Scenario 2: Synchronous generators and wind turbines jointly provide inertia and frequency response services; Scenario 3: Synchronous generators, wind turbines, and VPPs jointly participate in the supply of inertia and frequency response services.
[0130] Figures 3-5 The results of unit combination scheduling for conventional synchronous generator units are presented under different scenarios. Green marks indicate units online, and white marks indicate units offline. In Scenario 1, to ensure sufficient frequency support capacity, a large number of thermal power units need to be kept online continuously, reflecting the limitations of the adjustment flexibility of traditional single resources. In Scenario 2, after allowing wind turbine units to participate in frequency services, the system's dependence on thermal power units decreases. In Scenario 3, after adding VPP participation, some conventional thermal power units do not need to be kept operational to meet system safety requirements. These results indicate that expanding the scope of frequency service providers can significantly improve system flexibility and optimize the economy of unit combination scheduling strategies.
[0131] Figure 6 The data presents the price signal trends for various services under different scenarios. From scenario 1 to scenario 3, as the supply entities of inertia and frequency response services gradually expand, the scarcity of frequency security products decreases, resulting in a systematic downward trend in the prices of inertia, fast primary frequency modulation, and slow primary frequency modulation.
[0132] Figures 7-8 The system scheduling status and internal resource scheduling results within the VPP were demonstrated during the scheduling cycle. Wind power output provides a stable supply to meet load demand, while synchronous generators assume the main power supply responsibility during peak load periods. The VPP participates in system balance regulation through bidirectional power exchange, outputting power to the system during peak load periods and absorbing power from the system during off-peak periods, thereby improving system operational flexibility and ensuring supply-demand balance. Within the VPP, resources such as photovoltaic units, ES units, distributed generators, and EV clusters achieve optimized energy allocation through coordinated scheduling.
[0133] Table 3 summarizes the revenue, cost, and profit of different resource types under the three scenarios. Compared with Scenario 1, the operating cost of conventional synchronous generators in Scenario 2 and Scenario 3 decreased by 6.36% and 11.53%, respectively. This is due to the wind turbines and VPPs sharing some of the regulation pressure after participating in ancillary services. The wind turbines gained additional revenue by providing inertia and primary frequency response services, and the total revenue steadily increased, demonstrating a clear incentive effect. The net profit of the VPP increased from RMB 10.95 × 10³ to RMB 26.34 × 10³, and its economic feasibility improved with the increase in the depth of ancillary service participation. At the same time, the total payment of the system to all resources in Scenario 3 decreased by 4.28% compared with Scenario 1, indicating that the proposed framework can improve the overall economic efficiency of the market.
[0134]
[0135] Building upon Scenario 3, this study further analyzes the impact of renewable energy generation levels on market outcomes. Renewable energy output is scaled up to 60%, 80%, 100%, and 120% of the baseline level to explore its impact on market prices and ancillary service supply. Figure 9The data reveals the market price response characteristics under different renewable energy generation levels. In the electricity market, higher renewable energy penetration typically reduces marginal generation costs because wind power output replaces some of the high-fuel-cost output of thermal power units. This effect is particularly pronounced in several periods of the 120% renewable energy scenario, manifested as a significant decrease in energy prices. However, this cost reduction is accompanied by increased output volatility and forecast uncertainty, leading to increased system reserve demand and consequently, a rise in reserve prices as renewable energy generation levels increase. When renewable energy output is low, more synchronous generators need to be connected to the grid to meet load demand, resulting in ample system inertia and inertia prices approaching zero. As the proportion of renewable energy increases, thermal power units are gradually retired or reduced in grid connection, reducing the supply of traditional inertia. In this case, providing virtual inertia through inverter-based resources can ensure frequency security while reducing system costs and carbon emissions. The pricing of fast and slow primary frequency regulation is closely related to the flexibility of adjustment margins: during periods of surplus power generation, wind turbines can use the curtailed wind power for frequency response services, alleviating the pressure on other reserve resources and thus suppressing the prices of fast and slow primary frequency regulation; while during periods of power shortage, wind power output is mainly used for energy balance, and the frequency response price is determined by the cost of other alternative resources.
[0136] Figure 10 This demonstrates the supply of inertia and frequency response services under different renewable energy generation levels, and explains... Figure 9 The price change pattern in the system is as follows. VPPs maintain a nearly constant inertia contribution across various renewable energy generation levels. This is mainly because the distributed units aggregated within them do not consider unit combination decisions, and their inertia supply capacity is unaffected by grid connection status. Regarding fast primary frequency regulation supply, VPPs are the primary provider in all scenarios, while wind turbines prioritize energy balance and only participate in fast primary frequency regulation supply when renewable energy output is sufficient and there is room for wind curtailment. Regarding slow primary frequency regulation supply, the contribution of synchronous generators first decreases and then increases with the increase in renewable energy penetration: initially, increased renewable energy output leads to a decrease in the dispatch level of thermal power units, reducing the adjustment margin and decreasing slow primary frequency regulation supply; subsequently, as renewable energy output further increases, the dispatch level of thermal power units continues to decrease, opportunity costs decrease, and the slow primary frequency regulation supply capacity recovers somewhat. Since the total supply of fast and slow primary frequency regulation must meet system emergency backup requirements, the slow primary frequency regulation supply of VPPs shows the opposite trend to that of synchronous generators, i.e., it first increases and then decreases.
[0137] Table 4 compares the economic performance and robustness of the proposed method based on biblical chance constraints with deterministic models and the method based on Gaussian distribution chance constraints. The maximum empirical violation probability is used as the reliability evaluation index, defined as the maximum empirical violation probability of all biblical chance constraints calculated based on the out-of-sample validation dataset. The results show that the proposed method has the lowest maximum empirical violation probability, indicating superior reliability and robustness in dealing with uncertainty. Although the more conservative scheduling strategy leads to increased reserve costs, the method brings higher returns and profits to market participants, demonstrating that uncertainty has been effectively internalized into the pricing process. In contrast, the Gaussian distribution-based method exhibits a significantly higher maximum empirical violation probability, reflecting its insufficient robustness when the distribution assumptions do not match the actual uncertainty.
[0138]
[0139] Example 4: This embodiment provides a virtual power plant emergency frequency control system for implementing the virtual power plant emergency frequency control method described in embodiments 1-3 above, including a power grid control center, at least one virtual power plant aggregation layer, and several local resource controllers.
[0140] The power grid control center is used to perform offline pre-calculation, collect real-time power grid operating status data, identify active power deficits due to faults, generate and issue emergency power support commands, and receive the actual response power of virtual power plants uploaded by the aggregation layer of each virtual power plant, iteratively correcting the emergency power support commands for the next control cycle. The power grid control center includes a power grid control center database, in which the frequency support capacity and price parameters of each virtual power plant are pre-calculated offline and stored.
[0141] At least one virtual power plant aggregation layer is connected to the power grid control center. This aggregation layer receives emergency power support commands, collects frequency response parameters and the current operating status of each distributed resource, allocates the target support power of the virtual power plant, generates corresponding resource adjustment commands, and sends them to the local resource control layer. The aggregation layer also aggregates the actual response power of each distributed resource within the virtual power plant, obtains the actual response power of the virtual power plant, and uploads it to the power grid control center.
[0142] Several local resource controllers are connected to the virtual power plant aggregation layer. Each local resource controller is configured to correspond one-to-one with a distributed resource, and is used to read resource regulation commands, control the corresponding distributed resource to perform power regulation, output the corresponding actual response power, and upload the actual response power to the virtual power plant aggregation layer. The local resource controllers pre-store the frequency response parameters of the corresponding distributed resources, including response time, power ramp-up rate, and upper and lower output limits.
[0143] This embodiment achieves a centralized-decentralized hierarchical control architecture through the cooperation of a three-layer structure consisting of a power grid control center, a virtual power plant aggregation layer, and a local resource controller. This architecture enables the hierarchical issuance of emergency power support commands and their rapid execution on the resource side, as well as the layer-by-layer transmission and closed-loop iterative correction of the actual response power.
[0144] Example 5: This embodiment provides a computer device, which includes a processor and a memory. The memory stores a computer program, which includes program instructions. The processor executes the program instructions stored in the memory. The processor may be a Central Processing Unit (CPU), or other general-purpose processors, digital signal processors (DSPs), application-specific integrated circuits (ASICs), field-programmable gate arrays (FPGAs), or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, etc. The processor is the computing and control core of the terminal, suitable for implementing one or more instructions, specifically suitable for loading and executing one or more instructions in a computer storage medium to implement a corresponding method flow or corresponding function. The processor described in this embodiment can be used to perform related operations of a virtual power plant emergency frequency control method.
[0145] This embodiment provides a computer-readable storage medium, which is a memory device in a computer device used to store programs and data. It is understood that the computer-readable storage medium here can include both the built-in storage medium in the computer device and extended storage media supported by the computer device. The computer-readable storage medium provides storage space containing the terminal's operating system; and this storage space also contains one or more instructions suitable for loading and execution by a processor. These instructions can be one or more computer programs (including program code). It should be noted that the computer-readable storage medium here can be high-speed RAM (Random Access Memory) or non-volatile memory, such as at least one disk storage device. The processor can load and execute one or more instructions stored in the computer-readable storage medium to implement the corresponding steps of the virtual power plant emergency frequency control method in this embodiment.
[0146] Those skilled in the art will understand that embodiments of the present invention can be provided as methods, systems, or computer program products. Therefore, the present invention can take the form of a completely hardware embodiment, a completely software embodiment, or an embodiment combining software and hardware aspects. Moreover, the present invention can take the form of a computer program product implemented on one or more computer-usable storage media (including, but not limited to, disk storage and optical storage) containing computer-usable program code.
[0147] This invention is described with reference to flowchart illustrations and / or block diagrams of methods, apparatus (systems), and computer program products according to embodiments of the invention. It should be understood that each block of the flowchart illustrations and / or block diagrams, and combinations of blocks in the flowchart illustrations and / or block diagrams, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, special-purpose computer, embedded processor, or other programmable data processing apparatus to produce a machine, such that the instructions, which execute via the processor of the computer or other programmable data processing apparatus, generate instructions for implementing the flowchart illustrations and / or block diagrams. Figure 1 One or more processes and / or boxes Figure 1 A device that provides the functions specified in one or more boxes.
[0148] These computer program instructions may also be stored in a computer-readable storage medium that can direct a computer or other programmable data processing device to function in a particular manner, such that the instructions stored in the computer-readable storage medium produce an article of manufacture including instruction means, which are implemented in a process Figure 1 One or more processes and / or boxes Figure 1 The function specified in one or more boxes.
[0149] These computer program instructions may also be loaded onto a computer or other programmable data processing equipment to cause a series of operational steps to be performed on the computer or other programmable equipment to produce a computer-implemented process, thereby providing instructions that execute on the computer or other programmable equipment for implementing the process. Figure 1 One or more processes and / or boxes Figure 1 The steps of the function specified in one or more boxes.
[0150] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention, and not to limit them; although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some or all of the technical features therein; and these modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the scope of the technical solutions of the embodiments of the present invention.
Claims
1. A virtual power plant emergency frequency control method, characterized in that, Includes the following steps: A centralized-decentralized hierarchical control architecture is established, comprising a power grid control center, a virtual power plant aggregation layer, and a local resource control layer connected in sequence; wherein, the power grid control center includes a power grid control center database, and the frequency support capacity and price parameters of each virtual power plant are pre-calculated offline and stored in the power grid control center database; Frequency response parameters of each distributed resource are extracted through the virtual power plant aggregation layer; The system obtains real-time power grid operation status data through the power grid control center. When a power grid fault with active power deficit is detected, it obtains preset frequency security constraints and reads the frequency support capacity and price parameters of each virtual power plant stored in the power grid control center database. The frequency support capacity of each virtual power plant is used as the initial available capacity of each virtual power plant. The frequency response parameters are called and the initial available capacity is corrected in combination with the current real-time operating status of each virtual power plant to obtain the real-time available frequency support capacity. Based on the active power deficit, frequency security constraints and real-time available frequency support capacity, the target support power of the virtual power plant is determined and an emergency power support command is generated. The emergency power support command is sent to the virtual power plant aggregation layer. The virtual power plant aggregation layer allocates the target support power of the virtual power plant based on the current operating status of each distributed resource and the frequency response parameters, and generates the corresponding resource adjustment command. The resource adjustment command is sent to the local resource control layer, which then controls the corresponding distributed resource execution power adjustment according to the resource adjustment command and outputs the corresponding actual response power. The emergency power support command for the next control cycle is updated based on the actual response power to achieve emergency frequency control of the virtual power plant.
2. The virtual power plant emergency frequency control method according to claim 1, characterized in that, The frequency support capacity and price parameters of each virtual power plant are pre-calculated offline and stored in the database of the power grid control center, specifically including: Fuzzy sets are constructed based on moment information to quantify the uncertainty of new energy forecasting; With the goal of minimizing the overall system operating cost, five types of market services are integrated: electrical energy, backup, virtual inertia, fast primary frequency regulation, and slow primary frequency regulation. The safety constraints of frequency change rate, minimum fault frequency, quasi-steady-state frequency, and fuzzy sets are written into the model to construct a sub-Bruker optimization model. After equivalent reconstruction of the biblind bar optimization model, a two-stage step-by-step strategy is adopted to solve the problem and output the optimal electrical energy, reserve, inertia and frequency regulation clearing capacity for each time period. Based on the derivation of various service marginal prices using Lagrange dual multipliers, the frequency support capacity and price parameters of each virtual power plant are associated and stored in the database of the power grid control center.
3. The virtual power plant emergency frequency control method according to claim 2, characterized in that, The construction of fuzzy sets based on moment information specifically includes: Historical prediction error time series samples of wind power output, photovoltaic output, and wind turbine virtual inertia were collected, and wind power output prediction error vector, photovoltaic output prediction error vector, and wind turbine virtual inertia prediction error vector were constructed respectively. The expected values and covariance matrices of the wind power output prediction error vector, photovoltaic power output prediction error vector, and wind turbine virtual inertia prediction error vector are calculated respectively as reference moment information. By introducing the expected relaxation coefficient and the covariance amplification coefficient, a variable moment boundary is constructed that includes the upper and lower fluctuation range of the reference moment information; By utilizing variable moment boundaries and simultaneously introducing first-order moment uncertainty and second-order moment uncertainty, the wind power output prediction error vector, photovoltaic power output prediction error vector, and wind turbine virtual inertia prediction error vector are concatenated into a new energy joint prediction error vector, generating a fuzzy set that simultaneously contains moment uncertainty and coupling correlation between variables.
4. The virtual power plant emergency frequency control method according to claim 2, characterized in that, The bipolar bar optimization model is equivalently reconstructed and solved using a two-stage step-by-step strategy, outputting the optimal electrical energy, reserve, inertia, and frequency regulation clearing capacity for each time period, specifically including: The nonlinear frequency safety constraints and nonconvex opportunity constraints in the bibliometric bar optimization model are equivalently reconstructed. The nonlinear frequency safety constraints include the frequency change rate safety constraint, the minimum fault frequency safety constraint, and the quasi-steady-state frequency safety constraint. The frequency change rate safety constraint, the minimum fault frequency safety constraint, and the quasi-steady-state frequency safety constraint are transformed into second-order cone forms, and the non-convexity chance constraint is transformed into a linear deterministic constraint, thus transforming the sub-Bruker optimization model into a mixed-integer second-order cone programming model. The mixed-integer second-order cone programming model is decomposed into a unit combination sub-model and an economic dispatch sub-model, and solved using a two-stage step-by-step strategy: the first stage solves the unit combination sub-model, outputs and fixes the 0-1 start-stop state of conventional generator units; the second stage uses the 0-1 start-stop state determined in the first stage as the fixed boundary to solve the economic dispatch sub-model, and outputs the optimal electrical energy, reserve, inertia and frequency regulation clearing capacity for each time period.
5. The virtual power plant emergency frequency control method according to claim 1, characterized in that, The process involves using the frequency support capacity of each virtual power plant as its initial available capacity, calling frequency response parameters, and correcting the initial available capacity based on the current real-time operating status of each virtual power plant to obtain the real-time available frequency support capacity. Based on the active power deficit, frequency security constraints, and the real-time available frequency support capacity, the target support power for the virtual power plant is determined, and an emergency power support command is generated. Specifically, this includes: The frequency support capacity of each virtual power plant is used as the initial available capacity of each virtual power plant. The frequency support capacity of each virtual power plant includes inertia support capacity, fast primary frequency regulation capacity, and slow primary frequency regulation capacity. By calling the frequency response parameters and combining the current online resource status, current output power, energy storage device charge status, and communication status in the real-time operating status data of the virtual power plant, the inertia support capacity, fast primary frequency regulation capacity, and slow primary frequency regulation capacity are corrected to obtain the real-time available frequency support capacity of the virtual power plant. Based on the active power deficit and frequency security constraints, the total support power required by the system is calculated. Combined with the real-time available frequency support capacity, the target support power of the virtual power plant is allocated proportionally, and an emergency power support command carrying the target power value and response time limit is generated.
6. The virtual power plant emergency frequency control method according to claim 1, characterized in that, The virtual power plant aggregation layer allocates the target support power of the virtual power plant based on the current operating status of each distributed resource and the frequency response parameters, and generates corresponding resource adjustment instructions, specifically including: The virtual power plant aggregation layer extracts the target support power of the virtual power plant from the emergency power support command, collects the current output, adjustable capacity, remaining regulation margin and communication status of each distributed resource within the virtual power plant, and allocates the target support power of the virtual power plant in combination with frequency response parameters and the unit regulation cost of each distributed resource. The fast frequency regulation power in the target support power of the virtual power plant is allocated to energy storage, photovoltaic units and asynchronous motors, and the slow frequency regulation power in the target support power of the virtual power plant is allocated to distributed units, electric vehicles and air conditioning loads, generating resource regulation commands for each distributed resource.
7. The virtual power plant emergency frequency control method according to claim 1, characterized in that, The local resource control layer controls the corresponding distributed resource execution power adjustment according to the resource adjustment instruction, and outputs the corresponding actual response power, specifically including: The local resource control layer receives resource adjustment instructions sent by the virtual power plant aggregation layer. The resource adjustment instructions include the target adjustment power and response time requirements of the corresponding distributed resource. Obtain the preset frequency response parameters and local power grid frequency status data of the corresponding distributed resources. The preset frequency response parameters include response time, power ramp rate and output upper and lower limits. The power grid frequency status data includes frequency deviation and frequency change rate. Power regulation is performed based on resource adjustment instructions, response time, power ramp rate, output upper and lower limits, frequency deviation, and frequency change rate, and the corresponding actual response power is output.
8. The virtual power plant emergency frequency control method according to claim 1, characterized in that, The emergency power support command for updating the next control cycle based on the actual response power specifically includes: The actual response power of each distributed resource after executing resource regulation commands is uploaded to the virtual power plant aggregation layer, which then aggregates the actual response power of the virtual power plant and sends it to the power grid control center. Based on the deviation between the actual response power and the target support power, the power grid control center adjusts the virtual power plant target support power for the next control cycle and generates an updated emergency power support command until the frequency returns to a safe range.
9. A virtual power plant emergency frequency control system, characterized in that, The method for implementing the virtual power plant emergency frequency control method according to any one of claims 1-8 includes: The power grid control center is used to perform offline pre-calculation, collect real-time power grid operation status data, identify fault active power deficit, generate and issue emergency power support instructions, and receive the actual response power of virtual power plants uploaded by each virtual power plant aggregation layer, and iteratively correct the emergency power support instructions for the next control cycle. At least one virtual power plant aggregation layer is connected to the power grid control center. It is used to receive emergency power support instructions, collect frequency response parameters and the current operating status of each distributed resource, allocate the target support power of the virtual power plant, generate corresponding resource adjustment instructions and send them to the local resource control layer, and summarize the actual response power of the virtual power plant and upload it to the power grid control center. Several local resource controllers are connected to the virtual power plant aggregation layer. They are used to read resource adjustment instructions, control the corresponding distributed resources to perform power adjustment, output the corresponding actual response power, and upload it to the virtual power plant aggregation layer.
10. A computer device comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, characterized in that, When the processor executes the computer program, it implements the steps of the virtual power plant emergency frequency control method as described in any one of claims 1-8.
11. A computer-readable storage medium storing a computer program, characterized in that, When the computer program is executed by the processor, it implements the steps of the virtual power plant emergency frequency control method as described in any one of claims 1-8.
