Hierarchical scheduling control method and system for virtual power plant

Abstract

The invention relates to a hierarchical scheduling control method and system for a virtual power plant, and the method comprises the steps: recognizing an adjustable resource type in the virtual power plant and an operation main body which the adjustable resource type belongs to, and constructing a hierarchical scheduling system composed of a day-ahead optimization layer, an intra-day coordination layer and a real-time execution layer; in the day-ahead optimization layer, a scheduling optimization model is constructed based on adjustable resources and operation subjects, income weight parameters and scheduling boundary conditions are introduced, income constraints and power constraints of the operation subjects are optimized in a unified manner, and a day-ahead scheduling plan is generated by taking comprehensive income maximization as a target; the intra-day coordination layer adopts a rolling optimization mode to coordinate and correct the day-ahead scheduling result to form an intra-day rescheduling plan; and the real-time execution layer schedules the adjustable resources in real time according to the intra-day rescheduling plan, feeds back execution power, market clearing price, management subject income and load or output prediction error deviation, and is used for dynamically adjusting scheduling optimization model parameters, so that the overall economy of the virtual power plant is improved.

Description

Technical Field

[0001] This application relates to the field of power dispatching technology, and in particular to a hierarchical dispatching control method and system for virtual power plants. Background Technology

[0002] Against the backdrop of high-proportion integration of renewable energy and the gradual advancement of the electricity spot market and ancillary services market, virtual power plants, by aggregating distributed power sources, energy storage devices, and adjustable loads, participate in market bidding and dispatch execution as "power plant-like" entities, which helps to improve the flexibility of the power system and the level of renewable energy consumption.

[0003] Existing virtual power plant dispatching schemes often start from a single operating entity or a single resource type, lacking unified equivalent power modeling and unified constraint descriptions for distributed generation, energy storage, interruptible or peak-shaving loads, and electric vehicle charging loads. This makes it difficult to tap the overall regulation potential within a unified optimization framework. In terms of time scale, many schemes only focus on the day-ahead or intraday phase, failing to form a hierarchical collaborative dispatching system covering the day-ahead optimization layer, intraday coordination layer, and real-time execution layer. This results in day-ahead plans failing to reflect load forecast deviations, renewable energy output fluctuations, and market price changes in a timely manner. Intraday and real-time adjustments rely heavily on experience-based corrections, leading to a disconnect between planning and execution. Regarding multi-operating entity collaboration, a unified objective function is typically used to optimize overall revenue, with revenue weight parameters set statically. The lack of a dynamic adjustment mechanism based on revenue deviations over continuous settlement cycles easily leads to long-term revenue imbalances for some entities. Furthermore, existing technologies lack a unified comprehensive evaluation and offline model update mechanism for performance deviations, electricity price deviations, revenue deviations, and prediction errors. In scenarios such as communication anomalies, frequent disturbances, or rapid ramp-ups, they mainly rely on centralized control. The access side lacks edge control nodes with local optimization capabilities, and the operational safety and adaptive capabilities of virtual power plants still need to be improved. Summary of the Invention

[0004] To at least partially overcome the problem of the lack of hierarchical coordination and benefit coordination mechanisms in virtual power plant scheduling in related technologies, this application provides a hierarchical scheduling control method and system for virtual power plants.

[0005] The proposed solution is as follows: According to a first aspect of the embodiments of this application, a hierarchical scheduling control method for a virtual power plant is provided, comprising: Identify the types of adjustable resources within a virtual power plant and their respective operating entities; Construct a hierarchical scheduling system consisting of a day-ahead optimization layer, an intraday coordination layer, and a real-time execution layer; Based on the adjustable resource types and their respective operating entities, a scheduling optimization model containing each operating entity is constructed at the day-ahead optimization layer. In the scheduling optimization model, corresponding revenue weight parameters and scheduling boundary conditions are introduced for each operating entity, and the revenue constraints and power constraints of each operating entity are incorporated into a unified optimization solution. With the goal of maximizing comprehensive revenue, the day-ahead scheduling plan of the virtual power plant is generated through the scheduling optimization model. Based on the previous day's scheduling plan, the intraday coordination layer uses a rolling optimization method to coordinate and correct the scheduling results for each time period, thereby forming an intraday rescheduling plan; The real-time execution layer performs real-time scheduling of each adjustable resource according to the intraday rescheduling plan, and feeds back the execution power deviation, corresponding market clearing price deviation, revenue deviation of each operating entity, and load or output prediction error of each adjustable resource to the day-ahead optimization layer based on the set time period. Based on the feedback results from the real-time execution layer, the revenue weight parameters and scheduling boundary conditions of each operating entity in the scheduling optimization model are dynamically adjusted, and the day-ahead scheduling plan of the virtual power plant is regenerated through the corrected scheduling optimization model. Execute the intraday scheduling control cycle for the current day.

[0006] Preferably, the method further includes: After completing the intraday scheduling and control cycle for the current day, statistical analysis is conducted on the execution power deviation, market clearing price deviation, revenue deviation of each operating entity, and load or output forecast error for each time period of the current day to form a comprehensive evaluation index for evaluating the scheduling effect of the current day. Based on the comprehensive evaluation index, the revenue weight parameters and scheduling boundary conditions of each operating entity in the scheduling optimization model are updated offline to obtain the updated scheduling optimization model for the next day. The next-day day-ahead scheduling plan for the virtual power plant is generated based on the updated next-day scheduling optimization model.

[0007] Preferably, a scheduling optimization model incorporating each operating entity is constructed at the day-ahead optimization layer, including: The initial value of the revenue weight parameter for each operating entity is determined based on at least one of the following: the proportion of installed capacity in the virtual power plant, the historical power output contribution, and the contractual revenue ratio. The objective function of the scheduling optimization model is to take at least one of the following: revenue from the electricity market, revenue from the ancillary services market, and revenue from flexible load incentives.

[0008] Preferably, the method further includes: The actual output of each adjustable resource is compared with the planned output for the corresponding time period in the day-ahead scheduling plan to obtain the execution power deviation. The actual market clearing price is compared with the day-ahead forecast price to obtain the market clearing price deviation. The actual revenue of each operating entity in the current settlement period is compared with the target revenue to obtain the revenue deviation of each operating entity; The actual load or actual output curve is compared with the predicted curve to obtain the load or output prediction error.

[0009] Preferably, the revenue weight parameters and scheduling boundary conditions of each operating entity in the day-ahead optimization layer scheduling optimization model are dynamically adjusted, including: The revenue deviation of each business entity is determined based on the feedback results from the real-time execution layer; When the revenue deviation of an operating entity is lower than the preset revenue deviation lower limit for N consecutive settlement cycles, the revenue weight parameter corresponding to that operating entity is increased, and its output upper and lower limits or start-stop number constraints are relaxed accordingly. When the revenue deviation of an operating entity exceeds the preset revenue deviation lower limit for N consecutive settlement cycles, the revenue weight parameter corresponding to that operating entity is reduced, or the adjustable resource output space of other operating entities is allocated first. Each current day has multiple settlement cycles, and N is not less than 5.

[0010] Preferably, the intraday coordination layer, based on the previous day's scheduling plan, employs a rolling optimization approach to coordinate and correct the scheduling results for each time period, forming an intraday rescheduling plan, including: Set the rolling optimization interval and the rolling prediction time domain; the rolling optimization interval is a preset time step, and the rolling prediction time domain covers the current time to the remaining time of the current day or a preset end time. Upon reaching each rolling optimization interval, the actual output and status data of each adjustable resource in the previous rolling optimization interval, as well as the latest load forecast, renewable energy output forecast and market price forecast data, are collected, and the planned values ​​for the corresponding time period in the day-ahead scheduling plan are updated. The scheduling variables of each time period within the rolling prediction time domain after the current time are used as optimization windows. The scheduling variables of each time period within the optimization window are uniformly re-optimized to obtain the updated time-segmented power output plan. In the updated time-segmented output plan, the output results of one or more time periods closest to the current time are selected and locked, and used as a new intraday rescheduling plan to be sent to the real-time execution layer. The output results of the remaining time periods are used as the initial reference plan for the next rolling optimization.

[0011] Preferably, the method further includes: Based on historical operating data and real-time measurement data, load forecasting models, renewable energy output forecasting models, and market price forecasting models are constructed respectively. Based on the feedback results from the real-time execution layer, the parameters of the load forecasting model, renewable energy output forecasting model, and market price forecasting model are corrected.

[0012] Preferably, the method further includes: When the real-time execution layer performs real-time scheduling of each adjustable resource according to the intraday rescheduling plan, an edge control node is configured on the access side of each adjustable resource. When an abnormal scenario occurs, the edge control node performs local optimization control based on a preset power reference value, combined with local voltage, current, frequency and equipment status information. After the abnormal scenario is eliminated, the local execution results are summarized and reported to the intraday coordination layer and the day-ahead optimization layer; The abnormal scenarios include at least: communication abnormalities, frequent disturbances, or rapid climbing scenarios.

[0013] Preferably, identifying the types of adjustable resources within the virtual power plant includes: The adjustable resources within the virtual power plant are divided into at least one or more of the following: distributed generation units, rechargeable and dischargeable energy storage units, interruptible or peak-shifting load units, and electric vehicle charging units. Obtain the operating parameters of various adjustable resources, and construct a unified equivalent power model for adjustable resources based on the operating parameters.

[0014] According to a second aspect of the embodiments of this application, a virtual power plant hierarchical dispatch control system is provided, comprising: Processor and memory; The processor and memory are connected via a communication bus: The processor is used to call and execute the program stored in the memory; The memory is used to store a program, which is at least used to execute a virtual power plant hierarchical scheduling control method as described in any of the above.

[0015] The technical solution provided in this application may include the following beneficial effects: First, by identifying the types of adjustable resources within the virtual power plant and their respective operating entities, foundational data support is provided for the subsequent construction of the dispatch optimization model. These adjustable resource types include distributed generation units, energy storage units, and adjustable load units. This identification process ensures that the model accurately reflects the physical characteristics of different resources and the commercial attributes of the operating entities. Furthermore, a hierarchical dispatch system is constructed, consisting of a day-ahead optimization layer, an intraday coordination layer, and a real-time execution layer. This achieves seamless integration from long-term planning to real-time response, avoiding the disconnect between planning and execution caused by a single time scale in traditional solutions.

[0016] At the day-ahead optimization layer, a scheduling optimization model is constructed based on the adjustable resource types and their respective operating entities. Corresponding revenue weight parameters and scheduling boundary conditions are introduced for each operating entity, and revenue constraints and power constraints are incorporated into a unified optimization solution to generate a day-ahead scheduling plan with the goal of maximizing overall revenue. Specifically, the initial values ​​of the revenue weight parameters can be determined based on the installed capacity ratio, historical power contribution, or contractually agreed revenue proportion, thereby ensuring that the revenue distribution mechanism can dynamically match actual demand and avoid long-term revenue imbalances caused by static weights.

[0017] Through the intraday coordination layer, the scheduling results for each time period are coordinated and corrected using a rolling optimization method based on the day-ahead scheduling plan to form an intraday rescheduling plan. This rolling optimization mechanism dynamically updates the scheduling plan based on the latest load forecast, market price forecast, and renewable energy output forecast data, ensuring that it can respond promptly to fluctuations in new energy sources and market changes. Furthermore, through the real-time execution layer, the real-time scheduling of each adjustable resource is carried out according to the intraday rescheduling plan, and a multi-dimensional closed-loop monitoring mechanism is established based on feedback of execution power deviation, market clearing price deviation, revenue deviation of each operating entity, and load or output forecast errors within a set time period.

[0018] Based on feedback from the real-time execution layer, the revenue weight parameters and scheduling boundary conditions of each operating entity in the scheduling optimization model are dynamically adjusted. The day-ahead scheduling plan is then regenerated using the revised scheduling optimization model. For example, if an operating entity's revenue deviation falls below a preset lower limit for several consecutive settlement cycles, its revenue weight parameter can be increased and the scheduling boundary conditions relaxed, thereby achieving continuous optimization of the revenue distribution mechanism. This executes the intraday scheduling control cycle for the current day, ensuring that the hierarchical scheduling system forms an adaptive optimization closed loop within the daily cycle, significantly improving the virtual power plant's responsiveness to new energy fluctuations and market changes.

[0019] This technical solution achieves multi-timescale collaborative optimization through a hierarchical scheduling system, resolving the disconnect between planning and execution in traditional solutions. Furthermore, the dynamic adjustment mechanism of the revenue weight parameters continuously optimizes the revenue distribution of operating entities based on real-time feedback, avoiding long-term revenue imbalances caused by static weights. In addition, a closed-loop monitoring mechanism is established by providing multi-dimensional deviation data in real-time from the execution layer, enabling the scheduling scheme to adapt promptly to fluctuations in new energy sources and market changes.

[0020] It should be understood that the above general description and the following detailed description are exemplary and explanatory only, and do not limit this application. Attached Figure Description

[0021] The accompanying drawings, which are incorporated in and form part of this specification, illustrate embodiments consistent with this application and, together with the description, serve to explain the principles of this application.

[0022] Figure 1 This is a flowchart illustrating a hierarchical scheduling and control method for a virtual power plant according to an embodiment of this application; Figure 2 This is a schematic diagram of the structure of a hierarchical scheduling and control system for a virtual power plant provided in one embodiment of this application.

[0023] Reference numerals: Processor-21; Memory-22. Detailed Implementation

[0024] Exemplary embodiments will now be described in detail, examples of which are illustrated in the accompanying drawings. When the following description relates to the drawings, unless otherwise indicated, the same numbers in different drawings denote the same or similar elements. The embodiments described in the following exemplary embodiments do not represent all embodiments consistent with this application. Rather, they are merely examples of apparatuses and methods consistent with some aspects of this application as detailed in the appended claims.

[0025] Example 1 This application proposes a hierarchical scheduling and control method for virtual power plants, referring to... Figure 1 ,include: S11: Identify the types of adjustable resources within the virtual power plant and their respective operating entities; S12: Construct a hierarchical scheduling system consisting of a day-ahead optimization layer, an intraday coordination layer, and a real-time execution layer; S13: Based on the adjustable resource type and its operating entity, a scheduling optimization model containing each operating entity is constructed at the day-ahead optimization layer. Corresponding revenue weight parameters and scheduling boundary conditions are introduced for each operating entity, and the revenue constraints and power constraints of each operating entity are incorporated into a unified optimization solution. With the goal of maximizing comprehensive revenue, the day-ahead scheduling plan of the virtual power plant is generated through the scheduling optimization model. S14: The intraday coordination layer uses a rolling optimization method to coordinate and correct the scheduling results of each time period based on the day-ahead scheduling plan, forming an intraday rescheduling plan; S15: The real-time execution layer performs real-time scheduling of each adjustable resource according to the intraday rescheduling plan, and feeds back the execution power deviation, corresponding market clearing price deviation, revenue deviation of each operating entity, and load or output forecast error of each adjustable resource to the day-ahead optimization layer based on the set time period. S16: Based on the feedback results from the real-time execution layer, dynamically adjust the revenue weight parameters and scheduling boundary conditions of each operating entity in the scheduling optimization model, and regenerate the day-ahead scheduling plan of the virtual power plant through the corrected scheduling optimization model; S17: Execute the intraday scheduling control cycle for the current day.

[0026] Preferably, the adjustable resources within the virtual power plant are divided into at least one or more of the following: distributed generation units, rechargeable and dischargeable energy storage units, interruptible or peak-shifting load units, and electric vehicle charging units; the operating parameters of each type of adjustable resource are obtained, and a unified equivalent power model of the adjustable resources is constructed based on the operating parameters.

[0027] Specifically, distributed generation units refer to devices capable of actively providing power output, which can be implemented using photovoltaic modules, wind turbines, or micro gas turbines, aiming to provide a stable power source for virtual power plants. Rechargeable energy storage units refer to devices with energy storage and release capabilities, which can be implemented using lithium-ion battery packs, sodium-sulfur batteries, or flywheel energy storage systems, aiming to balance supply and demand fluctuations through energy time-shifting characteristics. Interruptible or peak-shaving load units refer to load devices capable of reducing or shifting electricity demand during specific periods, which can be implemented using industrial production lines, commercial air conditioning systems, or residential water heaters, aiming to enhance demand-side response flexibility. Electric vehicle charging units refer to mobile loads that support controlled charging behavior, which can be implemented using home charging piles, public fast charging stations, or wireless charging facilities, aiming to optimize overall regulation capabilities by utilizing their dispatch potential.

[0028] In practical applications, operating parameters refer to key indicators reflecting the dynamic characteristics of adjustable resources. These can include installed capacity, upper and lower limits of output, response rate, etc., with the aim of accurately characterizing the actual adjustment capability of the resources. A unified equivalent power model for adjustable resources refers to converting resources with different physical characteristics into a standardized mathematical representation in units of power. This can be achieved through linearization modeling methods or nonlinear function fitting techniques, with the aim of ensuring that heterogeneous resources are incorporated into the same mathematical framework in the optimization solution.

[0029] This technical solution first systematically classifies adjustable resources within a virtual power plant, clarifying the functional characteristics and regulatory behaviors of distributed generation units, rechargeable energy storage units, interruptible or peak-shaving load units, and electric vehicle charging units. Based on this, key operating parameters for each resource type are obtained, such as the installed capacity and upper / lower limits of output for distributed generation units, the charging / discharging efficiency and energy storage range for rechargeable energy storage units, the reduction depth and recovery time for interruptible or peak-shaving load units, and the charging power and scheduling window for electric vehicle charging units. These parameters provide a structured input framework for subsequently constructing a unified equivalent power model. Furthermore, by converting resources with different physical characteristics into a standardized power representation, a unified description of heterogeneous resources within the optimization framework is achieved, effectively solving the optimization fragmentation problem caused by differences in resource types.

[0030] Furthermore, this method, combined with a hierarchical scheduling system comprising a day-ahead optimization layer, an intraday coordination layer, and a real-time execution layer, provides fundamental support for tapping the overall regulation potential of virtual power plants. By aggregating dispersed resources into schedulable entities, it not only improves the feasibility of scheduling plans but also enhances the collaborative optimization capabilities of virtual power plants across multiple time scales.

[0031] In practical applications, a hierarchical scheduling system consisting of a day-ahead optimization layer, an intraday coordination layer, and a real-time execution layer can be constructed in various ways. For example, the day-ahead optimization layer can generate an initial scheduling plan using mathematical programming algorithms, the intraday coordination layer can dynamically update the plan based on rolling window technology, and the real-time execution layer can achieve rapid response through edge computing nodes. Thus, the hierarchical scheduling system can cover the entire chain of needs from long-term planning to real-time adjustments.

[0032] The innovation of this application lies in achieving multi-timescale collaborative optimization through a hierarchical scheduling system, solving the problem of disconnect between planning and execution in traditional schemes. Furthermore, the dynamic adjustment mechanism of the revenue weight parameters can continuously optimize the revenue distribution of operating entities based on real-time feedback results, avoiding long-term revenue imbalances caused by static weights. In addition, a closed-loop monitoring mechanism is established by providing multi-dimensional deviation data in real-time from the execution layer, enabling the scheduling scheme to adapt promptly to fluctuations in new energy sources and market changes.

[0033] The working principle of this application embodiment is as follows: First, by identifying the types of adjustable resources within the virtual power plant and their respective operating entities, basic data support is provided for the subsequent construction of the scheduling optimization model. The adjustable resource types include distributed generation units, energy storage units, and adjustable load units, etc. This identification process ensures that the model can accurately reflect the physical characteristics of different resources and the commercial attributes of the operating entities. Furthermore, a hierarchical scheduling system consisting of a day-ahead optimization layer, an intraday coordination layer, and a real-time execution layer is constructed, achieving a complete chain connection from long-term planning to real-time response, avoiding the problem of plan and execution disconnect caused by a single time scale in traditional solutions.

[0034] At the day-ahead optimization layer, a scheduling optimization model is constructed based on the adjustable resource types and their respective operating entities. Corresponding revenue weight parameters and scheduling boundary conditions are introduced for each operating entity, and revenue constraints and power constraints are incorporated into a unified optimization solution to generate a day-ahead scheduling plan with the goal of maximizing overall revenue. Specifically, the initial values ​​of the revenue weight parameters can be determined based on the installed capacity ratio, historical power contribution, or contractually agreed revenue proportion, thereby ensuring that the revenue distribution mechanism can dynamically match actual demand and avoid long-term revenue imbalances caused by static weights.

[0035] Through the intraday coordination layer, the scheduling results for each time period are coordinated and corrected using a rolling optimization method based on the day-ahead scheduling plan to form an intraday rescheduling plan. This rolling optimization mechanism dynamically updates the scheduling plan based on the latest load forecast, market price forecast, and renewable energy output forecast data, ensuring that it can respond promptly to fluctuations in new energy sources and market changes. Furthermore, through the real-time execution layer, the real-time scheduling of each adjustable resource is carried out according to the intraday rescheduling plan, and a multi-dimensional closed-loop monitoring mechanism is established based on feedback of execution power deviation, market clearing price deviation, revenue deviation of each operating entity, and load or output forecast errors within a set time period.

[0036] Based on feedback from the real-time execution layer, the revenue weight parameters and scheduling boundary conditions of each operating entity in the scheduling optimization model are dynamically adjusted. The day-ahead scheduling plan is then regenerated using the revised scheduling optimization model. For example, if an operating entity's revenue deviation falls below a preset lower limit for several consecutive settlement cycles, its revenue weight parameter can be increased and the scheduling boundary conditions relaxed, thereby achieving continuous optimization of the revenue distribution mechanism. This executes the intraday scheduling control cycle for the current day, ensuring that the hierarchical scheduling system forms an adaptive optimization closed loop within the daily cycle, significantly improving the virtual power plant's responsiveness to new energy fluctuations and market changes.

[0037] Example 2 This application further proposes that after completing the intraday dispatch control cycle of the current day, statistical analysis be conducted on the execution power deviation, market clearing price deviation, revenue deviation of each operating entity, and load or output prediction error for each time period of the current day to form a comprehensive evaluation index for evaluating the dispatch effect of the current day; based on the comprehensive evaluation index, the revenue weight parameters of each operating entity and the dispatch boundary conditions in the dispatch optimization model are updated offline to obtain the updated dispatch optimization model for the next day; and the next day's day-ahead dispatch plan for the virtual power plant is generated based on the updated dispatch optimization model for the next day.

[0038] Specifically, the comprehensive evaluation index refers to the quantitative assessment result formed by systematically statistically analyzing multi-dimensional data such as execution power deviation, market output price deviation, revenue deviation of each operating entity, and load or output prediction error. It can be implemented using weighted average method, fuzzy comprehensive evaluation method, or principal component analysis method. Its purpose is to comprehensively capture key issues in scheduling execution and provide a reliable basis for model updates. Offline updates refer to the process of deeply optimizing and adjusting the revenue weight parameters and scheduling boundary conditions in the scheduling optimization model in a non-real-time state. This can be achieved through genetic algorithms, particle swarm optimization algorithms, or Bayesian optimization methods. Its purpose is to ensure that the model parameters can adapt to long-term operating characteristics and overcome the shortcomings of limited computing resources and insufficient parameter adjustment in real-time dynamic adjustments.

[0039] In detail, after completing the intraday dispatch control cycle for the current day, the system first statistically analyzes the power deviation, market output price deviation, revenue deviation of each operating entity, and load or output forecast error for each time period of the day. The dispatch effect is quantified through comprehensive evaluation indicators, thus comprehensively reflecting the problems in dispatch execution. Subsequently, based on the formed comprehensive evaluation indicators, the revenue weight parameters of each operating entity and the dispatch boundary conditions in the dispatch optimization model are updated offline. This process allows for in-depth optimization of model parameters in a non-real-time state, ensuring that the model can adapt to long-term operating characteristics. Finally, based on the updated next-day dispatch optimization model, the next-day day-ahead dispatch plan for the virtual power plant is generated, effectively integrating historical experience into the formulation of the new plan, making the dispatch plan more forward-looking and adaptable. In this process, the offline update mechanism and the real-time dynamic adjustment mechanism complement each other, solving the problem of underutilization of historical data and improving the revenue stability and dispatch accuracy of the virtual power plant in long-term operation.

[0040] Example 3 This application further proposes to construct a dispatch optimization model that includes each operating entity at the day-ahead optimization layer, including: determining the initial value of the revenue weight parameter of each operating entity based on at least one of the following: the proportion of installed capacity of each operating entity in the virtual power plant, the historical output contribution, and the contractual revenue ratio; and using at least one of the following: electricity market revenue, ancillary service market revenue, and flexible load incentive revenue as the comprehensive revenue objective function of the dispatch optimization model.

[0041] Specifically, the installed capacity share refers to the proportion of physical regulation capacity of each operating entity within the virtual power plant. This can be achieved by statistically analyzing the ratio of the rated power of each entity's equipment to the total installed capacity of the virtual power plant, aiming to quantify the potential regulation capacity of each entity. Historical output contribution can be understood as the actual output performance of each operating entity over past operating cycles. This can be achieved by analyzing reliable output records in historical operating data, aiming to reflect the actual operational reliability of each entity. Contractual revenue share refers to the revenue distribution ratio specified by each operating entity in the commercial agreement. This can be achieved by analyzing the revenue distribution rules in the contract terms, aiming to ensure the binding force of the commercial contract. The comprehensive revenue objective function refers to an optimization objective encompassing multiple revenue sources. This can be achieved by selectively combining electricity market revenue, ancillary service market revenue, and flexible load incentive revenue, aiming to comprehensively reflect the actual operating mechanism of the electricity market.

[0042] In detail, when constructing the scheduling optimization model at the current optimization layer, the initial values ​​of the revenue weight parameters for each operating entity are scientifically set using at least one of the following: installed capacity ratio, historical output contribution, and contractual revenue ratio. This avoids unfairness caused by subjective and arbitrary settings. For example, the installed capacity ratio can quantify the potential adjustment capacity of each entity, the historical output contribution reflects the reliable contribution in actual operation, and the contractual revenue ratio ensures that commercial agreements are respected. This integration of multi-dimensional criteria makes the weight allocation more aligned with the actual operation scenario of the virtual power plant, providing a reasonable starting point for the subsequent dynamic adjustment mechanism. Secondly, when constructing the comprehensive revenue objective function, at least one of the following is included: electricity market revenue, ancillary service market revenue, and flexible load incentive revenue. This breaks through the limitations of a single revenue source and closely links the optimization objective to the actual operation mechanism of the electricity market. For example, electricity market revenue covers basic electricity trading, ancillary service market revenue reflects the value of value-added services such as frequency regulation and peak shaving, and flexible load incentive revenue reflects the additional revenue brought by demand response. By selectively combining these revenue sources, the objective function can dynamically adapt to market changes and fully capture the diversified revenue potential of virtual power plants, thereby supporting the day-ahead dispatch plan to achieve true overall revenue maximization on a fair basis.

[0043] Furthermore, the aforementioned scheme effectively addresses the issues of unscientific parameter setting and incomplete objectives in the scheduling optimization model by precisely setting the initial values ​​of revenue weight parameters and systematically defining the comprehensive revenue objective function. When determining the initial values ​​of revenue weight parameters, at least one of the following is considered: the installed capacity proportion of each operating entity in the virtual power plant, its historical output contribution, and the proportion of contractually agreed revenue. This avoids arbitrary subjective settings and ensures that the initial weights objectively reflect the physical contribution capacity, historical operating performance, and commercial contractual constraints of each entity. When constructing the comprehensive revenue objective function, at least one of the following is incorporated: electricity market revenue, ancillary service market revenue, and flexible load incentive revenue. This overcomes the limitations of a single revenue source and closely links the optimization objective to the actual operating mechanism of the electricity market. Through these technical solutions, the optimization process ensures both fairness and comprehensive revenue, avoiding the neglect of key revenue channels that could lead to suboptimal overall revenue, thereby improving the overall operating efficiency and fairness of the virtual power plant.

[0044] Example 4 This application further proposes the following technical solutions: comparing the actual output of each adjustable resource with the planned output for the corresponding time period in the day-ahead scheduling plan to obtain the execution power deviation; comparing the actual market clearing price with the day-ahead forecast price to obtain the market clearing price deviation; comparing the actual revenue of each operating entity with the target revenue in the current settlement period to obtain the revenue deviation of each operating entity; and comparing the actual load or actual output curve with the forecast curve to obtain the load or output forecast error.

[0045] Specifically, the execution power deviation refers to the difference between the output power of each adjustable resource during actual operation and the planned value for the corresponding time period in the day-ahead scheduling plan. It can be calculated by collecting real-time operating data of each adjustable resource and combining it with the power setpoint in the day-ahead scheduling plan, aiming to quantify the degree of deviation between scheduling execution and the plan. Market clearing price deviation can be understood as the difference between the actual market clearing price and the day-ahead predicted price. It can be achieved by obtaining real-time price data from the market price monitoring system and comparing it with the day-ahead predicted price, aiming to capture the actual impact of market price fluctuations on scheduling decisions. Revenue deviation of each operating entity refers to the gap between the actual revenue and the target revenue of each operating entity within a specific settlement period. It can be calculated by combining the target revenue model and actual revenue data of each operating entity through the revenue settlement module, aiming to ensure the timeliness and target orientation of revenue assessment. Load or output forecast error refers to the deviation between the actual load or output curve and the forecast curve. It can be achieved by using a curve fitting algorithm to conduct a comprehensive comparative analysis of the actual curve and the forecast curve, aiming to fully reflect the accuracy and fluctuation characteristics of the forecast model.

[0046] In detail, the above technical solution provides a precise data feedback foundation for the hierarchical dispatch system of virtual power plants by establishing a standardized deviation and error calculation framework. First, based on the day-ahead dispatch plan as the sole benchmark, comparing the actual output of each adjustable resource with the planned output accurately quantifies the deviation between dispatch execution and the plan, avoiding misjudgments caused by using historical data or other reference values, thus providing an objective basis for performance evaluation of the real-time execution layer. Second, using the day-ahead forecast price as a reference, comparing the actual market clearing price with the forecast price effectively captures the actual impact of market price fluctuations on dispatch decisions, making deviation calculation closely linked to market expectations at the time of dispatch plan generation, providing targeted data support for revenue model correction. Based on this, by limiting the time frame and binding target revenue, comparing the actual revenue of each operating entity with the target revenue ensures the timeliness and target orientation of revenue assessment, providing a direct basis for dynamically adjusting revenue weight parameters and maintaining the fairness of multi-entity collaboration. Finally, by employing a holistic curve comparison rather than single-point calculation, the actual load or output curve is compared with the predicted curve. This comprehensively reflects the accuracy and fluctuation characteristics of the prediction model, providing multi-dimensional error information for parameter correction and scheduling plan optimization, and enhancing the system's adaptability to the fluctuating nature of new energy resources. Through this technical solution, the problem of model adjustment failure caused by ambiguous data definitions is resolved, ensuring the consistency and accuracy of feedback data, thereby improving the dynamic adjustment effect of the scheduling optimization model.

[0047] Example 5 This application further proposes dynamic adjustments to the revenue weight parameters and scheduling boundary conditions of each operating entity in the day-ahead optimization layer scheduling optimization model, including: determining the revenue deviation of each operating entity based on the feedback results of the real-time execution layer; increasing the revenue weight parameter corresponding to the operating entity when the revenue deviation of the operating entity is lower than the preset revenue deviation lower limit for N consecutive settlement cycles, and correspondingly relaxing its output upper and lower limits or start / stop frequency constraints; decreasing the revenue weight parameter corresponding to the operating entity when the revenue deviation of the operating entity is higher than the preset revenue deviation lower limit for N consecutive settlement cycles, or prioritizing the allocation of adjustable resource output space of other operating entities; wherein each current day has multiple settlement cycles, and N is not less than 5.

[0048] Specifically, the revenue weight parameter refers to the key parameter used to adjust the priority and revenue distribution ratio of each operating entity in the scheduling optimization model. It can be implemented by introducing a weighted algorithm based on historical data or a machine learning model. The scheduling boundary conditions can be understood as a set of operational constraints on each operating entity, such as the output range and the number of start-stop cycles. They can be implemented using dynamic programming methods or linear programming techniques. The purpose of introducing the above features is to solve the problem of long-term revenue imbalance for some operating entities caused by the static setting of the revenue weight parameter through a dynamic adjustment mechanism.

[0049] In detail, this scheme quantifies the actual revenue deviation of each operating entity through feedback results from the real-time execution layer, ensuring that adjustments are based on actual operating conditions rather than a static model. For operating entities whose revenue deviation is below a preset lower limit for N consecutive settlement cycles, their scheduling priority and operational flexibility are improved by increasing their revenue weight parameters and relaxing scheduling boundary conditions, thereby improving their revenue situation and maintaining their participation enthusiasm. For operating entities whose revenue deviation is above a preset lower limit for N consecutive settlement cycles, their revenue weight parameters are reduced or their output space is reallocated to avoid excessive revenue concentration and promote fair cooperation among multiple entities. In addition, setting N to be no less than 5 effectively filters short-term market fluctuations and accidental interference, ensuring that adjustments are based on stable revenue trends and enhancing the robustness and adaptability of the dynamic mechanism. Through the above technical solution, not only is the problem of static setting of revenue weight parameters and scheduling boundary conditions solved, but the overall collaborative efficiency and operational safety of the virtual power plant are also improved.

[0050] Example 6 This application further proposes a rolling optimization approach based on the day-ahead scheduling plan, using an intraday coordination layer to coordinate and correct the scheduling results for each time period, forming an intraday rescheduling plan. This includes: setting a rolling optimization interval and a rolling forecast time domain; the rolling optimization interval is a preset time step, and the rolling forecast time domain covers the current time to the remaining time of the current day or a preset end time; upon reaching each rolling optimization interval, collecting the actual output and status data of each adjustable resource within the previous rolling optimization interval, as well as the latest load forecast, renewable energy output forecast, and market price forecast data, and updating the planned values ​​for the corresponding time periods in the day-ahead scheduling plan; using multiple time periods after the current time but within the rolling forecast time domain as optimization windows, uniformly re-optimizing the scheduling variables for each time period within the optimization windows to obtain an updated time-segmented output plan; in the updated time-segmented output plan, selecting and locking the output results of one or more time periods closest to the current time as a new intraday rescheduling plan and issuing it to the real-time execution layer, with the output results of the remaining time periods serving as the initial reference plan for the next rolling optimization.

[0051] Specifically, the rolling optimization interval refers to the time step in which the system periodically triggers the optimization process. It can be implemented using a fixed time interval (e.g., 15 minutes, 30 minutes, etc.) or a dynamic time interval based on event triggers. Its purpose is to ensure that the optimization frequency meets real-time requirements while avoiding excessive computational burden due to excessive frequency. The rolling prediction time domain refers to the time range extending from the current moment to a future moment. It can be implemented using a complete time domain covering the current moment to a preset end time or a time domain length dynamically adjusted according to actual needs. Its purpose is to provide a sufficient time view to balance real-time performance and foresight. The optimization window refers to multiple consecutive time periods selected within the rolling prediction time domain. It can be implemented using a fixed number of time periods or a time period range dynamically adjusted based on prediction accuracy. Its purpose is to ensure that the scheduling variables for each time period can be optimized in a coordinated manner as a whole, avoiding the fragmentation problem caused by segmented independent adjustments.

[0052] In detail, the above scheme enables the system to periodically trigger correction processes by setting rolling optimization intervals and rolling forecast time domains, while providing a complete time view for optimization, thereby effectively responding to load or output fluctuations. Upon reaching each rolling optimization interval, the actual output and status data of each adjustable resource within the previous rolling optimization interval are collected. Combined with the latest load forecast, renewable energy output forecast, and market price forecast data, the planned values ​​for the corresponding time periods in the day-ahead dispatch plan are dynamically corrected. This ensures that the rescheduling plan closely matches the current operating status of the system, reducing execution errors caused by forecast deviations. By uniformly re-optimizing the dispatch variables for each time period within the optimization window, a smooth transition in output between time periods and full exploitation of resource potential are achieved. Selecting and locking the output results of one or more time periods closest to the current time and issuing them for execution ensures the stability and operability of the instructions while retaining results for longer-term time periods as a reference for subsequent optimization, thus achieving a close connection between intraday rescheduling plans and real-time execution. Furthermore, the above scheme, combined with the mechanisms of the day-ahead optimization layer and the real-time execution layer, further improves the adaptability and execution effectiveness of the virtual power plant dispatch plan through dynamic updates and feedback mechanisms.

[0053] Example 7 This application further proposes to construct load forecasting models, renewable energy output forecasting models, and market price forecasting models based on historical operating data and real-time measurement data, respectively; and to revise the parameters of the load forecasting models, renewable energy output forecasting models, and market price forecasting models based on the feedback results of the real-time execution layer.

[0054] Specifically, load forecasting models are mathematical models used to predict future load demand within a virtual power plant. They can be implemented using time series analysis, machine learning algorithms, or deep learning networks, aiming to capture the long-term regularities of load changes and reflect instantaneous changes in the system's state. Renewable energy output forecasting models can be understood as specialized models for predicting the power output of renewable energy sources such as wind and solar power. They can be implemented through a combination of physical modeling and data-driven approaches, or simply through statistical models trained on historical data, aiming to improve the accuracy of predicting the volatility of renewable energy. Furthermore, market price forecasting models are tools used to predict future electricity price trends in the power market. They can be implemented using methods such as regression analysis, support vector machines, or neural networks, aiming to provide highly reliable price input data for optimized dispatching.

[0055] In detail, the above scheme, by constructing separate load forecasting, renewable energy output forecasting, and market price forecasting models, ensures independent optimization space for different forecasting objects and avoids mutual interference between multi-dimensional forecasting tasks, thereby significantly improving the basic accuracy of each forecasting model. Based on this, the parameters of the above models are dynamically corrected using actual operational error information such as execution power deviation and market clearing price deviation fed back from the real-time execution layer. This allows the forecasting models to adapt to changes in system operating status in a timely manner and eliminates the cumulative effect of historical forecasting errors. This mechanism fully utilizes the advantage of historical operating data in capturing long-term regularities, while combining the characteristics of real-time measurement data reflecting instantaneous changes in system status, providing highly timely and accurate input data support for intraday rolling optimization. Furthermore, by dynamically adjusting the forecasting model parameters, the problem of insufficient accuracy caused by static forecasting models is effectively solved, ensuring the accuracy and responsiveness of intraday rescheduling plan generation.

[0056] In summary, the above technical solutions not only solve the problem that forecast data cannot adapt to changes in system operation in a timely manner, but also significantly improve the accuracy and adaptability of virtual power plant scheduling through dynamic correction mechanisms, laying a solid foundation for the efficient operation of the hierarchical collaborative scheduling system.

[0057] Example 8 This application further proposes that when the real-time execution layer performs real-time scheduling of each adjustable resource according to the intraday rescheduling plan, edge control nodes are configured on the access side of each adjustable resource. When abnormal scenarios occur, the edge control nodes perform local optimization control based on preset power reference values, combined with local voltage, current, frequency and equipment status information. After the abnormal scenario is eliminated, the local execution results are summarized and reported to the intraday coordination layer and the day-ahead optimization layer. Abnormal scenarios include at least: communication abnormalities, frequent disturbances or rapid ramping scenarios.

[0058] In practical applications, an edge control node refers to a distributed control unit with local computing and decision-making capabilities, which can be implemented using embedded controllers, industrial computers, or intelligent terminals. Specifically, the preset power reference value refers to a baseline value set based on historical scheduling plans or safety boundaries. Its purpose is to provide a stable basis for local optimization under abnormal scenarios and avoid system instability caused by blind adjustments. Local voltage, current, frequency, and equipment status information are key parameters reflecting the real-time operating status of the power grid. These can be acquired through sensors, measuring devices, or data acquisition modules to accurately capture abnormal fluctuations and adjust power output accordingly.

[0059] Specifically, this solution, by configuring edge control nodes on each adjustable resource access side, enables the activation of a localized response mechanism without relying on central commands in abnormal scenarios such as communication anomalies, frequent disturbances, or rapid ramp-ups. The edge control nodes perform local optimization control based on preset power reference values, combining locally sensed voltage, current, frequency, and equipment status information to effectively maintain grid voltage and frequency stability, preventing equipment overload or system crashes. After the abnormal scenario is resolved, the local execution results are summarized and reported to the intraday coordination layer and the day-ahead optimization layer. This feedback mechanism incorporates local decision-making data from the abnormal period into the upper-layer model, used to correct subsequent scheduling plans and revenue weight parameters, thereby improving the system's adaptability to similar scenarios. Furthermore, the design for communication anomalies, frequent disturbances, or rapid ramp-ups ensures that the solution covers core risk points in grids with high renewable energy penetration, guarantees the continuity of real-time scheduling, and strengthens the overall robustness of the virtual power plant.

[0060] In summary, the above technical solutions not only solve the scheduling interruption problem caused by the failure of centralized control in abnormal scenarios, but also improve the safety and adaptability of virtual power plant operation.

[0061] Example 9 This application also discloses a hierarchical scheduling and control system for virtual power plants, including: Processor 21 and memory 22; Processor 21 and memory 22 are connected via a communication bus: The processor 21 is used to call and execute the program stored in the memory 22; The memory 22 is used to store a program, which is used to execute at least one of the virtual power plant hierarchical scheduling control methods described above.

[0062] It is understood that the same or similar parts in the above embodiments can be referred to each other, and the contents not described in detail in some embodiments can be referred to the same or similar contents in other embodiments.

[0063] It should be noted that in the description of this application, the terms "first," "second," etc., are used for descriptive purposes only and should not be construed as indicating or implying relative importance. Furthermore, in the description of this application, unless otherwise stated, "a plurality of" means at least two.

[0064] Any process or method described in the flowchart or otherwise herein can be understood as representing a module, segment, or portion of code comprising one or more executable instructions for implementing a particular logical function or process, and the scope of the preferred embodiments of this application includes additional implementations in which functions may be performed not in the order shown or discussed, including substantially simultaneously or in reverse order depending on the function involved, as will be understood by those skilled in the art to which embodiments of this application pertain.

[0065] It should be understood that various parts of this application can be implemented using hardware, software, firmware, or a combination thereof. In the above embodiments, multiple steps or methods can be implemented using software or firmware stored in memory and executed by a suitable instruction execution system. For example, if implemented in hardware, as in another embodiment, it can be implemented using any one or a combination of the following techniques known in the art: discrete logic circuits having logic gates for implementing logical functions on data signals, application-specific integrated circuits (ASICs) having suitable combinational logic gates, programmable gate arrays (PGAs), field-programmable gate arrays (FPGAs), etc.

[0066] Those skilled in the art will understand that all or part of the steps of the methods described in the above embodiments can be implemented by a program instructing related hardware. The program can be stored in a computer-readable storage medium, and when executed, it includes one or a combination of the steps of the method embodiments.

[0067] Furthermore, the functional units in the various embodiments of this application can be integrated into a processing module, or each unit can exist physically separately, or two or more units can be integrated into a module. The integrated module can be implemented in hardware or as a software functional module. If the integrated module is implemented as a software functional module and sold or used as an independent product, it can also be stored in a computer-readable storage medium.

[0068] The storage media mentioned above can be read-only memory, disk, or optical disk, etc.

[0069] In the description of this specification, the references to terms such as "one embodiment," "some embodiments," "example," "specific example," or "some examples," etc., indicate that a specific feature, structure, material, or characteristic described in connection with that embodiment or example is included in at least one embodiment or example of this application. In this specification, the illustrative expressions of the above terms do not necessarily refer to the same embodiment or example. Furthermore, the specific features, structures, materials, or characteristics described may be combined in any suitable manner in one or more embodiments or examples.

[0070] Although embodiments of this application have been shown and described above, it is understood that the above embodiments are exemplary and should not be construed as limiting this application. Those skilled in the art can make changes, modifications, substitutions and variations to the above embodiments within the scope of this application.

Claims

1. A hierarchical scheduling and control method for a virtual power plant, characterized in that, include: Identify the types of adjustable resources within a virtual power plant and their respective operating entities; Construct a hierarchical scheduling system consisting of a day-ahead optimization layer, an intraday coordination layer, and a real-time execution layer; Based on the adjustable resource types and their respective operating entities, a scheduling optimization model containing each operating entity is constructed at the day-ahead optimization layer. In the scheduling optimization model, corresponding revenue weight parameters and scheduling boundary conditions are introduced for each operating entity, and the revenue constraints and power constraints of each operating entity are incorporated into a unified optimization solution. With the goal of maximizing comprehensive revenue, the day-ahead scheduling plan of the virtual power plant is generated through the scheduling optimization model. Based on the previous day's scheduling plan, the intraday coordination layer uses a rolling optimization method to coordinate and correct the scheduling results for each time period, thereby forming an intraday rescheduling plan; The real-time execution layer performs real-time scheduling of each adjustable resource according to the intraday rescheduling plan, and feeds back the execution power deviation, corresponding market clearing price deviation, revenue deviation of each operating entity, and load or output prediction error of each adjustable resource to the day-ahead optimization layer based on the set time period. Based on the feedback results from the real-time execution layer, the revenue weight parameters and scheduling boundary conditions of each operating entity in the scheduling optimization model are dynamically adjusted, and the day-ahead scheduling plan of the virtual power plant is regenerated through the corrected scheduling optimization model. Execute the intraday scheduling control cycle for the current day.

2. The method according to claim 1, characterized in that, The method further includes: After completing the intraday scheduling and control cycle for the current day, statistical analysis is conducted on the execution power deviation, market clearing price deviation, revenue deviation of each operating entity, and load or output forecast error for each time period of the current day to form a comprehensive evaluation index for evaluating the scheduling effect of the current day. Based on the comprehensive evaluation index, the revenue weight parameters and scheduling boundary conditions of each operating entity in the scheduling optimization model are updated offline to obtain the updated scheduling optimization model for the next day. The next-day day-ahead scheduling plan for the virtual power plant is generated based on the updated next-day scheduling optimization model.

3. The method according to claim 1, characterized in that, The current optimization layer constructs a scheduling optimization model that includes all operating entities, including: The initial value of the revenue weight parameter for each operating entity is determined based on at least one of the following: the proportion of installed capacity in the virtual power plant, the historical power output contribution, and the contractual revenue ratio. The objective function of the scheduling optimization model is to take at least one of the following: revenue from the electricity market, revenue from the ancillary services market, and revenue from flexible load incentives.

4. The method according to claim 1, characterized in that, The method further includes: The actual output of each adjustable resource is compared with the planned output for the corresponding time period in the day-ahead scheduling plan to obtain the execution power deviation. The actual market clearing price is compared with the day-ahead forecast price to obtain the market clearing price deviation. The actual revenue of each operating entity in the current settlement period is compared with the target revenue to obtain the revenue deviation of each operating entity; The actual load or actual output curve is compared with the predicted curve to obtain the load or output prediction error.

5. The method according to claim 1, characterized in that, The revenue weight parameters and scheduling boundary conditions of each operating entity in the current optimization layer scheduling optimization model are dynamically adjusted, including: The revenue deviation of each business entity is determined based on the feedback results from the real-time execution layer; When the revenue deviation of an operating entity is lower than the preset revenue deviation lower limit for N consecutive settlement cycles, the revenue weight parameter corresponding to that operating entity is increased, and its output upper and lower limits or start-stop number constraints are relaxed accordingly. When the revenue deviation of an operating entity exceeds the preset revenue deviation lower limit for N consecutive settlement cycles, the revenue weight parameter corresponding to that operating entity is reduced, or the adjustable resource output space of other operating entities is allocated first. Each current day has multiple settlement cycles, and N is not less than 5.

6. The method according to claim 1, characterized in that, Based on the day-ahead scheduling plan, the intraday coordination layer uses a rolling optimization approach to coordinate and correct the scheduling results for each time period, forming an intraday rescheduling plan, including: Set the rolling optimization interval and the rolling prediction time domain; the rolling optimization interval is a preset time step, and the rolling prediction time domain covers the current time to the remaining time of the current day or a preset end time. Upon reaching each rolling optimization interval, the actual output and status data of each adjustable resource in the previous rolling optimization interval, as well as the latest load forecast, renewable energy output forecast and market price forecast data, are collected, and the planned values ​​for the corresponding time period in the day-ahead scheduling plan are updated. The scheduling variables of each time period within the rolling prediction time domain after the current time are used as optimization windows. The scheduling variables of each time period within the optimization window are uniformly re-optimized to obtain the updated time-segmented power output plan. In the updated time-segmented output plan, the output results of one or more time periods closest to the current time are selected and locked, and used as a new intraday rescheduling plan to be sent to the real-time execution layer. The output results of the remaining time periods are used as the initial reference plan for the next rolling optimization.

7. The method according to claim 6, characterized in that, The method further includes: Based on historical operating data and real-time measurement data, load forecasting models, renewable energy output forecasting models, and market price forecasting models are constructed respectively. Based on the feedback results from the real-time execution layer, the parameters of the load forecasting model, renewable energy output forecasting model, and market price forecasting model are corrected.

8. The method according to claim 1, characterized in that, The method further includes: When the real-time execution layer performs real-time scheduling of each adjustable resource according to the intraday rescheduling plan, an edge control node is configured on the access side of each adjustable resource. When an abnormal scenario occurs, the edge control node performs local optimization control based on a preset power reference value, combined with local voltage, current, frequency and equipment status information. After the abnormal scenario is eliminated, the local execution results are summarized and reported to the intraday coordination layer and the day-ahead optimization layer; The abnormal scenarios include at least: communication abnormalities, frequent disturbances, or rapid climbing scenarios.

9. The method according to claim 1, characterized in that, Identify the types of adjustable resources within the virtual power plant, including: The adjustable resources within the virtual power plant are divided into at least one or more of the following: distributed generation units, rechargeable and dischargeable energy storage units, interruptible or peak-shifting load units, and electric vehicle charging units. Obtain the operating parameters of various adjustable resources, and construct a unified equivalent power model for adjustable resources based on the operating parameters.

10. A hierarchical dispatching and control system for a virtual power plant, characterized in that, include: Processor and memory; The processor and memory are connected via a communication bus: The processor is used to call and execute the program stored in the memory; The memory is used to store a program, which is at least used to execute the virtual power plant hierarchical scheduling control method according to any one of claims 1-9.

Images