A multi-stage optimization control method for virtual power plant considering supply load

Abstract

This application relates to the field of virtual power plant optimal dispatch technology, specifically to a multi-stage optimal control method for virtual power plants considering supply load. The method includes: acquiring grid dispatch instructions, real-time status data of each dispatchable resource within the virtual power plant, and multi-stage planning data; determining the matching degree between each dispatchable resource and the grid dispatch instructions; determining the coordination index of each dispatchable resource; determining the dynamic adjustment weight of each dispatchable resource based on its matching degree and coordination index; and generating and issuing control instructions for each dispatchable resource based on the power to be allocated and the dynamic adjustment weight of each dispatchable resource. This method can improve the rationality of resource allocation.

Description

Technical Field

[0001] This application relates to the field of virtual power plant optimization scheduling technology, specifically to a multi-stage optimization control method for virtual power plants that considers supply load. Background Technology

[0002] With the high proportion of renewable energy sources, such as wind and solar power, being connected to the grid, their inherent intermittency, volatility, and uncertainty have brought unprecedented challenges to the safe, stable, and economical operation of the power system. Virtual power plants, as a key technology for integrating massive, dispersed, and heterogeneous distributed energy resources, are considered one of the core solutions for improving the grid's ability to absorb renewable energy and enhancing system flexibility and resilience.

[0003] In existing technologies, the optimal control of virtual power plants typically employs a multi-stage framework, including a day-ahead stage, an intraday rolling correction stage, and a real-time control stage. The day-ahead stage formulates an economically optimal dispatch plan based on forecast data, the intraday rolling correction stage adjusts the plan based on ultra-short-term forecasts, and the real-time control stage responds to dispatch commands issued by the power grid.

[0004] However, the optimization objectives may differ at different stages. For example, the day-ahead stage focuses on economic efficiency, while the real-time stage focuses on command tracking accuracy. Inconsistent optimization objectives can lead to a lack of effective coordination between stages. Inappropriate resource allocation in the real-time control stage can result in unsatisfactory overall control performance and affect the operational reliability of the virtual power plant in resource control. Summary of the Invention

[0005] To address the technical problem of unsatisfactory overall control performance caused by unreasonable resource allocation during the real-time control phase of a virtual power plant, this application aims to provide a multi-stage optimization control method for a virtual power plant that considers the supply load. The specific technical solution adopted is as follows:

[0006] This application provides a multi-stage optimization control method for a virtual power plant considering supply and load, comprising: acquiring grid dispatch instructions, real-time status data of each dispatchable resource within the virtual power plant, and multi-stage planning data, wherein the dispatchable resources include supply-side resources and load-side resources, and the multi-stage planning data includes day-ahead planning data and intraday rolling correction planning data; the grid dispatch instructions include the power to be allocated; determining the matching degree between each dispatchable resource and the grid dispatch instructions based on the grid dispatch instructions and the real-time status data of each dispatchable resource; determining a coordination index for each dispatchable resource based on the matching degree between each dispatchable resource and the grid dispatch instructions, the real-time status data of each dispatchable resource, and the multi-stage planning data, wherein the coordination index characterizes the degree of consistency between the real-time status data of the dispatchable resource and the multi-stage planning data; determining the dynamic adjustment weight of each dispatchable resource based on the matching degree and the coordination index; and generating and issuing control instructions for each dispatchable resource based on the power to be allocated and the dynamic adjustment weight of each dispatchable resource.

[0007] Optionally, determining the matching degree between each dispatchable resource and the grid dispatch command based on the grid dispatch command and the real-time status data of each dispatchable resource includes: determining the current adjustment direction based on the grid dispatch command and extracting the command feature vector from it, wherein the current adjustment direction is the power adjustment direction indicated by the grid dispatch command at the current moment, and the command feature vector includes at least the power change amplitude, the direction change frequency, and the duration of continuous same direction; extracting the capability feature vector of each dispatchable resource in the current adjustment direction based on the real-time status data of each dispatchable resource, wherein the capability feature vector includes at least the adjustable power range, response speed, and duration of the dispatchable resource in the current adjustment direction; and determining the matching degree between each dispatchable resource and the grid dispatch command based on the command feature vector and the capability feature vector of each dispatchable resource.

[0008] Optionally, the determination of the matching degree between each schedulable resource and the grid dispatch command based on the instruction feature vector and the capability feature vector of each schedulable resource includes: determining the difference between the instruction feature vector and the capability feature vector of each schedulable resource based on the difference between the power change amplitude and the adjustable power range, the difference between the direction change frequency and the response speed, and the difference between the continuous same-direction duration and the allowable duration; determining the changing trend of the adjustable power range of each schedulable resource in the current adjustment direction based on the real-time status data of each schedulable resource; and determining the matching degree between each schedulable resource and the grid dispatch command based on the difference between the instruction feature vector and the capability feature vector of each schedulable resource, and the changing trend of the adjustable power range of each schedulable resource in the current adjustment direction.

[0009] Optionally, the real-time status data of each dispatchable resource includes the available power in the current adjustment direction, the maximum adjustable power in the current adjustment direction, and the current measured power. The planned data for the day-ahead phase includes the planned power sequence for the entire day, and the planned data for the intraday rolling correction phase includes the short-term planned power sequence. Based on the matching degree between each dispatchable resource and the grid dispatch command, the real-time status data of each dispatchable resource, and the multi-stage planned data, the coordination index of each dispatchable resource is determined, including: determining the adjustment direction demand characteristics of each dispatchable resource based on the current adjustment direction, the available power of each dispatchable resource in the current adjustment direction, the maximum adjustable power of each dispatchable resource in the current adjustment direction, and the matching degree between each dispatchable resource and the grid dispatch command; determining the state deviation degree of each dispatchable resource based on the current measured power, the planned power sequence for the entire day, and the short-term planned power sequence; and determining the coordination index of each dispatchable resource based on the adjustment direction demand characteristics and the state deviation degree.

[0010] Optionally, the above-mentioned determination of the regulation direction demand characteristics of each dispatchable resource based on the current regulation direction, the available power of each dispatchable resource in the current regulation direction, the maximum adjustable power of each dispatchable resource in the current regulation direction, and the matching degree between each dispatchable resource and the grid dispatch command includes: determining the ratio between the available power of each dispatchable resource in the current regulation direction and the maximum adjustable power in the current regulation direction to determine the available regulation capacity ratio of each dispatchable resource; and determining the regulation direction demand characteristics of each dispatchable resource based on the matching degree between each dispatchable resource and the grid dispatch command and the available regulation capacity ratio of each dispatchable resource.

[0011] Optionally, the determination of the state deviation degree of each schedulable resource based on the current measured power, the all-day planned power sequence, and the short-term planned power sequence of each schedulable resource includes: interpolating the all-day planned power sequence and the short-term planned power sequence to obtain the planned power at the current moment in the day-ahead phase and the planned power at the current moment in the intraday rolling correction phase; determining the average of the planned power at the current moment in the day-ahead phase and the planned power at the current moment in the intraday rolling correction phase as the target planned power; and determining the state deviation degree of each schedulable resource based on the difference between the current measured power and the target planned power.

[0012] Optionally, the above-mentioned determination of the dynamic adjustment weight of each schedulable resource based on the matching degree and coordination degree index of each schedulable resource includes: obtaining the basic weight of each schedulable resource; determining the weight adjustment factor of each schedulable resource based on the matching degree and coordination degree index of each schedulable resource; and determining the dynamic adjustment weight of each schedulable resource based on the basic weight and weight adjustment factor of each schedulable resource.

[0013] Optionally, the weight adjustment factor of the first schedulable resource is determined based on the matching degree and coordination degree index of the first schedulable resource, including: normalizing the matching degree and coordination degree index of the first schedulable resource to obtain normalized matching degree and normalized coordination degree, wherein the first schedulable resource is any one of all schedulable resources in the virtual power plant; and determining the sum of the normalized matching degree and normalized coordination degree as the weight adjustment factor of the first schedulable resource.

[0014] Optionally, the above-mentioned generation and issuance of control instructions for each schedulable resource based on the power of the instruction to be allocated and the dynamic adjustment weight of each schedulable resource includes: determining the normalized adjustment weight of each schedulable resource as the ratio between the dynamic adjustment weight of each schedulable resource and the total dynamic adjustment weight, wherein the total dynamic adjustment weight is the sum of the dynamic adjustment weights of all schedulable resources; allocating the power of the instruction to be allocated based on the normalized adjustment weight of each schedulable resource to obtain the target power allocation value of each schedulable resource; and generating control instructions for each schedulable resource based on the target power allocation value of each schedulable resource.

[0015] Optionally, the above-mentioned allocation of the command power to be allocated based on the normalized adjustment weight of each schedulable resource to obtain the target power allocation value of each schedulable resource includes: determining the product of the normalized adjustment weight and the command power to be allocated as the initial power allocation value; determining the initial power allocation value of each schedulable resource as the target power allocation value when the initial power allocation value of each schedulable resource is less than its respective maximum adjustable power; determining the maximum adjustable power as the target power allocation value of the at least one schedulable resource when the initial power allocation value of at least one schedulable resource exceeds its corresponding maximum adjustable power, and determining the remaining power to be allocated; redistributing the remaining power to be allocated based on the dynamic adjustment weight of other schedulable resources until the total power allocation value of all schedulable resources is within their respective maximum adjustable power, and determining the final total power allocation value allocated to each schedulable resource as the target power allocation value of each schedulable resource, wherein the other schedulable resources are the schedulable resources other than the at least one schedulable resource among all schedulable resources in the virtual power plant.

[0016] This application has the following beneficial effects:

[0017] By determining the matching degree between each schedulable resource and the grid dispatch command, and the coordination index between each resource and the multi-stage plan data, and further determining the dynamic adjustment weights, schedulable resources with higher matching degrees and higher coordination indices can occupy a higher proportion during resource dispatch. Resource allocation based on dynamic adjustment weights can improve the rationality of resource allocation, thereby enabling the control commands generated by the virtual power plant to take into account both the real-time status of schedulable resources and the objectives of the multi-stage plan, improving the overall control effect and the operational reliability of the virtual power plant's multi-stage optimized control. Attached Figure Description

[0018] To more clearly illustrate the technical solutions and advantages in the embodiments of this application or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of this application. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0019] Figure 1 A flowchart of a multi-stage optimization control method for a virtual power plant considering supply load, provided as an embodiment of this application;

[0020] Figure 2 A flowchart of another method for multi-stage optimization control of a virtual power plant considering supply load, provided as an embodiment of this application;

[0021] Figure 3 This is a flowchart illustrating another method for multi-stage optimization control of a virtual power plant considering supply load, provided as an embodiment of this application. Detailed Implementation

[0022] To further illustrate the technical means and effects adopted by this application to achieve the intended purpose of the invention, the following, in conjunction with the accompanying drawings and preferred embodiments, details the specific implementation, structure, features, and effects of a multi-stage optimization control method for a virtual power plant considering supply load proposed in this application. In the following description, different "one embodiment" or "another embodiment" do not necessarily refer to the same embodiment. Furthermore, specific features, structures, or characteristics in one or more embodiments can be combined in any suitable form.

[0023] Unless otherwise defined, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this application pertains.

[0024] A virtual power plant is a power management system that aggregates and coordinates distributed power sources, energy storage systems, and dispatchable resources through advanced information and communication technologies and software systems. It is not a physical power plant, but rather an integrated, adjustable "energy pool" that combines multiple dispersed energy resources using the Internet of Things (IoT) and AI algorithms to participate in the electricity market and grid operation. The core objective of virtual power plant optimization control is to respond to grid dispatch commands or market signals by precisely regulating internal resources, enabling it to exhibit stable, controllable, and adjustable "power plant" characteristics externally. Due to uncertainties such as load demand and market electricity prices, the optimization control problem is typically decomposed into multiple time periods, with each stage addressing decision-making issues at different time scales.

[0025] Typical phase divisions include: day-ahead phase, intraday rolling optimization phase, and real-time control phase.

[0026] Day-ahead planning, intraday rolling adjustments, and real-time control often employ different models or rules, lacking coordination. For example, day-ahead planning might keep energy storage in a low-power state for economic reasons, but if a sudden intraday need arises for extended frequency support, the storage might be unable to meet the demand due to insufficient power. Inconsistent optimization objectives across different stages lead to suboptimal resource configuration over the entire timeline, impacting the overall reliability and robustness of the virtual power plant's response.

[0027] The following description, in conjunction with the accompanying drawings, details a specific scheme for a multi-stage optimization control method for a virtual power plant that considers supply load, as provided in this application.

[0028] Please see Figure 1 The diagram illustrates a flowchart of a multi-stage optimization control method for a virtual power plant that takes into account the supply load, according to an embodiment of this application.

[0029] like Figure 1 As shown, this multi-stage optimization control method for a virtual power plant that considers the supply load includes S101-S105.

[0030] S101. Obtain power grid dispatch instructions, real-time status data of each dispatchable resource in the virtual power plant, and multi-stage planning data.

[0031] The dispatchable resources include supply-side resources and load-side resources; the multi-stage planning data includes day-ahead planning data and intraday rolling correction planning data; and the grid dispatch instructions include the power to be allocated.

[0032] It should be understood that the grid dispatch command is an automatic generation control (AGC) command issued by the superior grid dispatching agency and received in real time by the virtual power plant through a high-speed communication link. This grid dispatch command is a continuously changing digital signal, which is usually updated at a fixed frequency (e.g., once every 4 seconds) to specify a command power. The sign of the command power indicates the direction of demand (positive values ​​usually indicate that power output needs to be increased, and negative values ​​indicate that power output needs to be reduced), and the absolute value indicates the magnitude of demand.

[0033] For example, assuming the command power is +20W, then the command power represents the need to increase the power output by 20W.

[0034] It is understandable that grid dispatch instructions are usually issued periodically (e.g., once every 5 minutes). A grid dispatch instruction usually includes multiple instruction powers. The virtual power plant needs to allocate each instruction power in sequence to determine the share allocated to each dispatchable resource. In this embodiment, the instruction power that is about to be allocated is determined as the instruction power to be allocated.

[0035] Optionally, supply-side resources include energy storage systems, distributed generators, etc., while load-side resources include air conditioners, electric vehicle charging stations, etc.

[0036] Optionally, the real-time status data for each schedulable resource includes the available power in the current adjustment direction, the maximum adjustable power in the current adjustment direction, and the current measured power.

[0037] It should be understood that available power refers to the maximum power that the schedulable resources can actually output safely and stably in the current adjustment direction at the current moment, while maximum adjustable power refers to the theoretical limit of adjustment capability that the schedulable resources can achieve in the current adjustment direction.

[0038] In one alternative implementation, real-time status data of each schedulable resource can be collected in real time by sensors, smart meters, or resource controllers deployed locally on each schedulable resource, and recorded synchronously with a unified timestamp.

[0039] For energy storage systems, real-time status data also includes current state of charge, maximum charge / discharge power, charge / discharge efficiency, discharge cutoff state of charge, charging cutoff state of charge, total capacity, and response time. For distributed generators, real-time status data also includes maximum output, minimum stable output, ramp rate, remaining fuel, rated fuel consumption rate, and maximum continuous operating time. For load-side resources, real-time status data also includes adjustable upper limit, adjustable lower limit, and response time.

[0040] It should be understood that the discharge cutoff state of charge is used to characterize the minimum safe line for battery discharge, the charge cutoff state of charge is used to characterize the maximum safe line for battery charging, the response time is the delay from receiving the command to starting to execute the adjustment, and is used to calculate the response speed, the adjustable power upper limit and adjustable power lower limit define the power range that the load can safely adjust under the current operating state (e.g., the maximum power reduction of an air conditioner), and the ramp rate is the active power that the generator can increase or decrease per unit time.

[0041] For an energy storage system, when it needs to discharge, the available power is the maximum continuous discharge power converted from the available capacity between the current state of charge and the discharge cutoff state of charge; when it needs to charge, the available power is the maximum continuous charging power converted from the available capacity between the current state of charge and the charging cutoff state of charge; the maximum adjustable power is the rated charging and discharging power of the energy storage system.

[0042] For distributed generators, when it is necessary to increase output, the available power is the difference between the maximum output and the current output; when it is necessary to decrease output, the available power is the difference between the current output and the minimum stable output; the maximum adjustable power is taken as the rated maximum output of the generator.

[0043] For load-side resources, when an increase in output is required, the available power is the difference between the adjustable upper limit and the current power; when a decrease in output is required, the available power is the difference between the current power and the adjustable lower limit; the maximum adjustable power is the difference between the upper and lower limits of the load's adjustable range.

[0044] It should be understood that the virtual power plant stores multi-stage planning data. The day-ahead planning data refers to the full-day planning data formulated the previous day for the current day. This planning data includes the full-day planned power sequence, which includes the planned power of each schedulable resource for the entire day with a first time interval resolution. The intraday rolling correction planning data refers to the planning data generated by rolling correction of the plan for the next few hours on the same day, which is closer to the real-time situation. This planning data includes the short-term planned power sequence, which includes the planned power of each schedulable resource for the next few hours with a second time interval resolution.

[0045] For example, the first duration is longer than the second duration. The first duration is usually 15 minutes, and the second duration is 5 minutes. The plan data for the intraday rolling correction phase can be the plan data for the next 4 hours, which is predicted 15 minutes in advance.

[0046] S102. Based on the power grid dispatch instructions and the real-time status data of each dispatchable resource, determine the matching degree between each dispatchable resource and the power grid dispatch instructions.

[0047] It is understandable that if the real-time status data of a schedulable resource can meet the requirements of the power grid dispatching command, then the higher the degree of matching between the schedulable resource and the power grid dispatching command.

[0048] S103. Based on the matching degree between each schedulable resource and the power grid dispatching command, the real-time status data of each schedulable resource, and the multi-stage planning data, determine the coordination index of each schedulable resource.

[0049] The coordination index characterizes the degree of consistency between the real-time status data of schedulable resources and the multi-stage planning data. A higher coordination index indicates a stronger regulatory advantage of the schedulable resources in the current adjustment direction. The closer the actual operating state of the schedulable resources is to the expected state of the multi-stage plan, the better.

[0050] S104. Based on the matching degree and coordination degree index of each schedulable resource, determine the dynamic adjustment weight of each schedulable resource.

[0051] It should be understood that the dynamic adjustment weight of a schedulable resource represents the proportion of resource scheduling for that schedulable resource among all schedulable resources.

[0052] In one implementation of this application, the basic weight of each schedulable resource can be obtained first; the weight adjustment factor of each schedulable resource can be determined based on the matching degree and coordination degree index of each schedulable resource; and the dynamic adjustment weight of each schedulable resource can be determined based on the basic weight and weight adjustment factor of each schedulable resource.

[0053] It should be understood that the base weight reflects the proportion of a schedulable resource that can be scheduled under ideal conditions.

[0054] Optionally, the base weight of each schedulable resource can be obtained from the multi-stage planning data, which is a weight predicted based on the forecast data.

[0055] Optionally, in the absence of a base weight in the multi-stage planning data, the base weight of each schedulable resource can be set to be the same, that is, the base weight of each schedulable resource is the reciprocal of the number of all schedulable resources.

[0056] It should be understood that the weight adjustment factor is used to quantify the magnitude of the correction of the basic weight by real-time dynamic factors, so that the final dynamic adjustment weight can reflect the real-time advantage of schedulable resources in the current control cycle.

[0057] In one alternative implementation, taking the first schedulable resource as an example, the matching degree and coordination degree index of the first schedulable resource can be normalized to obtain the normalized matching degree and the normalized coordination degree; the sum of the normalized matching degree and the normalized coordination degree is determined as the weight adjustment factor of the first schedulable resource.

[0058] The first schedulable resource is any one of all schedulable resources within the virtual power plant.

[0059] Optionally, the matching degree and the coordination degree index can be normalized based on the maximum-minimum normalization method, and both the matching degree and the coordination degree index can be mapped to the interval [0, 1]. Based on this, the value range of the weight adjustment factor is [0, 2].

[0060] It is understandable that the larger the weight adjustment factor, the greater the adjustment to the basic weight.

[0061] Optionally, the product of the weight adjustment factor and the base weight can be used to determine the dynamically adjusted weight.

[0062] S105. Based on the power of the instruction to be allocated and the dynamic adjustment weight of each schedulable resource, generate and issue control instructions for each schedulable resource.

[0063] It should be understood that a control command includes a schedulable resource and a target power allocation value, and the control command is used to instruct the schedulable resource to reduce or increase its power output based on the target power allocation value.

[0064] It is understandable that both the power to be allocated and the target power allocation value are directional values, with positive values ​​representing increased power output and negative values ​​representing decreased power output.

[0065] In one optional implementation, the ratio between the dynamic adjustment weight of each schedulable resource and the total dynamic adjustment weight is determined as the normalized adjustment weight of each schedulable resource, and the total dynamic adjustment weight is the sum of the dynamic adjustment weights of all schedulable resources; the power of the instruction to be allocated is allocated based on the normalized adjustment weight of each schedulable resource to obtain the target power allocation value of each schedulable resource; and the control instruction for each schedulable resource is generated based on the target power allocation value of each schedulable resource.

[0066] It is understandable that the sum of the dynamic adjustment weights of all the schedulable resources identified above may be greater than 1. Directly multiplying the dynamic adjustment weight by the power of the instruction to be allocated may cause the sum of the target power allocation values ​​of all schedulable resources to exceed the power of the instruction to be allocated. Therefore, it is necessary to normalize the dynamic adjustment weight of each schedulable resource so that the sum of the normalized adjustment weights of all schedulable resources is 1.

[0067] In this embodiment, the ratio between the dynamic adjustment weight of each schedulable resource and the total dynamic adjustment weight is determined as the normalized adjustment weight of each schedulable resource, so that the sum of the normalized adjustment weights of all schedulable resources is 1.

[0068] In one optional implementation, the product of the normalized adjustment weight and the power of the instruction to be allocated is determined as the initial power allocation value. If the initial power allocation value of each schedulable resource is less than its respective maximum adjustable power, the initial power allocation value of each schedulable resource is determined as the target power allocation value. If the initial power allocation value of at least one schedulable resource exceeds its corresponding maximum adjustable power, the maximum adjustable power is determined as the target power allocation value of that at least one schedulable resource, and the remaining power to be allocated is determined. The remaining power to be allocated is redistributed based on the dynamic adjustment weights of other schedulable resources until the total power allocation value of all schedulable resources is within their respective maximum adjustable power, and the final total power allocation value allocated to each schedulable resource is determined as the target power allocation value for each schedulable resource.

[0069] Among these, the other schedulable resources are schedulable resources other than at least one schedulable resource among all schedulable resources in the virtual power plant.

[0070] It should be understood that the maximum adjustable power is the physical limit of the schedulable resource in the current adjustment direction. When the initial power allocation value of each schedulable resource is less than its own maximum adjustable power, it means that each schedulable resource can provide the initial power allocation value while maintaining stable operation. At this time, the initial power allocation value of each schedulable resource can be determined as the target power allocation value, and control commands can be generated.

[0071] It is understandable that if the initial power allocation value of at least one schedulable resource exceeds the corresponding maximum adjustable power, it means that the at least one schedulable resource cannot provide the initial power allocation value, or cannot maintain stable operation after providing the initial power allocation value. In this case, it is unreasonable to determine the initial power allocation value as the target power allocation value. Instead, the maximum adjustable power of each of the at least one schedulable resource can be determined as its respective target power allocation value.

[0072] Simultaneously, the remaining power to be allocated is determined, which is the sum of the differences between the initial power allocation value of each of the at least one schedulable resources and their respective maximum adjustable power.

[0073] Optionally, the remaining power to be allocated satisfies the following formula:

[0074]

[0075] in, Indicates the remaining power to be allocated. This indicates the number of schedulable resources whose initial power allocation exceeds the corresponding maximum adjustable power. Indicates the first The initial power allocation value for a schedulable resource. Indicates the first The maximum adjustable power of a schedulable resource.

[0076] It should be understood that the remaining unallocated power can be allocated to other schedulable resources.

[0077] Specifically, the redistribution method is as follows: the dynamic adjustment weights of other schedulable resources are normalized to obtain a secondary normalized adjustment weight for each schedulable resource, so that the sum of the secondary normalized adjustment weights of other schedulable resources is 1 (this normalization method is similar to the method for determining the normalized adjustment weights mentioned above, and will not be repeated here). Then, the product of the secondary normalized adjustment weight and the remaining power to be allocated is determined as the secondary power allocation value.

[0078] It should be understood that at this point, it is necessary to re-verify whether the sum of the initial power allocation value and the secondary power allocation value of each other schedulable resource is less than or equal to the maximum adjustable power. If it is less than or equal to the maximum adjustable power, the sum of the initial power allocation value and the secondary power allocation value is determined as the target power allocation value. If it is greater than the maximum adjustable power, it is necessary to determine the remaining power to be allocated and redistribute it again until the total power allocation value of all schedulable resources is within their respective maximum adjustable power.

[0079] It should be understood that the total power allocation value of a schedulable resource is the sum of the power values ​​allocated in each power allocation process.

[0080] Finally, control commands for each schedulable resource are generated based on the target power allocation value for each schedulable resource and sent to the corresponding schedulable resource.

[0081] The methods provided in S101-S105 above determine the matching degree between each schedulable resource and the power grid dispatching command, and the coordination index between each resource and the multi-stage planning data. By further determining the dynamic adjustment weight, the schedulable resources with higher matching degree and higher coordination index can occupy a higher proportion during resource dispatching. Based on the dynamic adjustment weight, resource allocation can improve the rationality of resource allocation. This enables the control commands generated by the virtual power plant to take into account both the real-time status of the schedulable resources and the objectives of the multi-stage plan, thereby improving the overall control effect and the operational reliability of the multi-stage optimized control of the virtual power plant.

[0082] Combination Figure 1 ,like Figure 2 As shown, in one implementation of this application embodiment, the above-mentioned S102 can be specifically implemented by S201-S203.

[0083] S201. Determine the current adjustment direction based on the power grid dispatch instructions, and extract the instruction feature vector from it.

[0084] The current adjustment direction is the power adjustment direction indicated by the power instruction to be allocated, and the instruction feature vector includes at least the power change amplitude, the direction change frequency, and the duration of continuous same direction.

[0085] In this embodiment of the application, the instruction feature vector is used to characterize the overall regulation requirements of the power grid on the virtual power plant.

[0086] Among them, the power change amplitude is used to characterize the degree of fluctuation in the commanded power in the power grid dispatch command.

[0087] Optionally, the absolute value of the power to be assigned can be determined as the power change amplitude.

[0088] It should be understood that the direction change frequency is the number of times the command changes direction per unit time. It is used to characterize the frequency of power fluctuations in the power dispatch command. The higher the direction change frequency, the more frequently the power dispatch command switches between "increasing power output" and "decreasing power output", reflecting that the power grid frequency or power is in a state of rapid fluctuation.

[0089] Optionally, the method for determining the direction change frequency is as follows: traverse all commanded powers of the power grid dispatching command, determine the direction of change of all commanded powers except the first commanded power relative to the previous commanded power, and obtain the direction change sequence. When a commanded power has the same direction as the previous commanded power, it is recorded as a positive change (represented by +1), and when it is different from the previous commanded power, it is recorded as a negative change (represented by -1). When two adjacent direction changes are different, it is determined that the direction has changed once. Count the number of direction changes, and determine the direction change frequency by the ratio between the total number of direction changes and the total number of commanded powers of the power grid dispatching command.

[0090] For example, assuming the direction change sequence is [+1, +1, -1, -1, +1], then there is one change between the second and third directions, and one change between the fourth and fifth directions, for a total of 2 direction changes.

[0091] It should be understood that the duration of continuous unidirectional adjustment is used to characterize the strength and stability of the trend of power grid dispatch commands remaining unchanged in the same direction. The higher the duration of continuous unidirectional adjustment, the longer the adjustment needs to be carried out in one direction.

[0092] Optionally, the method for determining the duration of continuous same-direction instruction power is as follows: starting from the instruction power to be assigned, determine the number of instruction powers that are in the same direction and continuous with the instruction power to be assigned (including the instruction power to be assigned), and multiply the time interval between instruction powers by this number to obtain the duration of continuous same-direction instruction power.

[0093] For example, assuming the time interval is 4 seconds, the grid command power is [+20W, +20W, +40W, -20W, -10W], and the command power to be allocated is the first +20W, then the number of command powers that are in the same direction and consecutive with the command power to be allocated is 3, and the duration of consecutive in the same direction is 12 seconds.

[0094] It should be understood that the power change amplitude, the frequency of the direction change, and the duration of continuous unidirectional changes together constitute the feature vector of this command.

[0095] S202. Based on the real-time status data of each schedulable resource, extract the capability feature vector of each schedulable resource in the current adjustment direction.

[0096] The capability feature vector includes at least the adjustable power range, response speed, and duration of the schedulable resource in the current adjustment direction.

[0097] It should be understood that the adjustable power range is the maximum power adjustment capability that the schedulable resource can provide instantaneously in the current adjustment direction, which can be determined from real-time status data based on the current adjustment direction and the resource type of the schedulable resource.

[0098] Optionally, when the current adjustment direction is positive (i.e., power output needs to be increased and energy storage discharge is required), for energy storage systems, the maximum discharge power can be determined as the adjustable power range; for distributed generators, the difference between the rated maximum output and the current output can be determined as the adjustable power range; for load-side resources, the difference between the adjustable upper limit and the current power can be determined as the adjustable power range.

[0099] When the current adjustment direction is negative (i.e., output needs to be reduced for energy storage charging), for energy storage systems, the maximum charging power can be defined as the adjustable power range; for distributed generators, the difference between the current output and the minimum stable output can be defined as the adjustable power range. For load-side resources, the difference between the current power and the adjustable lower limit can be defined as the adjustable power range.

[0100] It should be understood that response speed is the reciprocal of the time required for a schedulable resource to start executing a regulation task from receiving a control command. It is used to characterize the rapid response capability of a schedulable resource in the current regulation direction. The larger the value, the faster the response speed and the stronger the response capability.

[0101] Alternatively, for energy storage systems and load-side resources, the response speed is typically the reciprocal of the response time. For distributed generators, the response speed can be determined by the ratio of the maximum ramp rate to the rated power.

[0102] It should be understood that the duration of operation is the longest time a schedulable resource can operate continuously in the current regulation direction with an adjustable power range, and is used to characterize the "endurance" or persistent regulation capability of a schedulable resource.

[0103] Optionally, for an energy storage system, if the current adjustment direction is positive, the percentage difference between the current state of charge and the discharge cutoff state of charge can be determined first, and the product of this percentage difference and the total capacity can be determined as the current available discharge capacity. The ratio between the current available discharge capacity and the adjustable power range can be determined as the duration. When the current adjustment direction is negative, the percentage difference between the charging cutoff state of charge and the current state of charge can be determined first, and the product of this percentage difference and the total capacity can be determined as the current available charging capacity. The ratio between the current available charging capacity and the adjustable power range can be determined as the duration.

[0104] For distributed generators, the duration can be determined based on the remaining fuel quantity or the maximum continuous operating time: when the generator is running continuously in the current adjustment direction with an adjustable power range, the ratio between the remaining fuel quantity and the rated fuel consumption rate corresponding to the adjustable power range is determined as the duration; if there is a maximum continuous operating time limit for the generator, the maximum continuous operating time is determined as the duration.

[0105] For load-side resources, the duration is constrained by user comfort or equipment operation limitations: taking air conditioners as an example, the time required for the room temperature to change from the current value to the upper or lower limit of the allowable temperature under the adjustable power range can be determined as the duration; taking electric vehicle charging piles as an example, the time for them to continue operating under the adjustable power range until the user's charging needs are met or the time constraint for leaving the user can be determined as the duration.

[0106] S203. Based on the instruction feature vector and the capability feature vector of each schedulable resource, determine the matching degree between each schedulable resource and the power grid dispatch instruction.

[0107] Among them, the matching degree characterizes the degree of matching between the adjustment capability of a schedulable resource and the demand of power grid dispatch instructions. The higher the matching degree, the greater the degree of matching.

[0108] Optionally, each feature in the instruction feature vector can be compared with the corresponding feature in the capability feature vector to obtain the matching degree.

[0109] In one alternative implementation, the difference between the command feature vector and the capability feature vector of each schedulable resource can be determined based on the difference between the power change amplitude and the adjustable power range, the difference between the direction change frequency and the response speed, and the difference between the continuous same-direction duration and the allowable duration. Based on the real-time status data of each schedulable resource, the changing trend of the adjustable power range of each schedulable resource in the current adjustment direction can be determined. Based on the difference between the command feature vector and the capability feature vector of each schedulable resource, and the changing trend of the adjustable power range of each schedulable resource in the current adjustment direction, the matching degree between each schedulable resource and the grid dispatch command can be determined.

[0110] It should be understood that comparing the power change amplitude with the adjustable power range can reflect whether the schedulable resources can cover the power of the instruction to be allocated in terms of power capacity. Subtracting the adjustable power range from the power change amplitude will result in a negative difference if the adjustable power range is greater than the power change amplitude, indicating that the resource capacity is sufficient and the matching degree should be high. If the adjustable power range is less than the power change amplitude, the difference will be positive, indicating that the resource capacity is insufficient and the matching degree should be low.

[0111] Comparing the frequency of direction changes with the response speed reflects whether the resource's response to command changes matches the speed at which they are changing. Subtracting the response speed from the frequency of direction changes results in a negative difference if the response speed is greater than the frequency of direction changes, indicating that the resource can keep up with the pace of command changes and the matching degree should be high. If the response speed is less than the frequency of direction changes, the difference is positive, indicating that the resource response is too slow and the matching degree should be low.

[0112] Comparing the duration of continuous unidirectional movement with the allowable duration reflects whether the resource's continuous adjustment capability meets the persistence requirements of the instruction. Subtracting the allowable duration from the continuous unidirectional movement duration results in a negative difference if the allowable duration is greater than the continuous unidirectional movement duration, indicating that the resource can support the continuous execution of the instruction and the matching degree should be high; if the allowable duration is less than the continuous unidirectional movement duration, the difference is positive, indicating that the resource's endurance is insufficient and the matching degree should be low.

[0113] The three differences mentioned above reflect the degree of difference between the regulation capacity of dispatchable resources and the demand of power grid dispatch instructions from different dimensions.

[0114] Optionally, based on the importance of these three differences in actual control, preset weights can be adaptively configured for each difference. For example, in scenarios requiring rapid response, the weight of response speed can be appropriately increased. Unless otherwise specified, each of the three weights can be set to 1 / 3, and then the three differences can be summed using a weighted average to obtain the difference value. This difference value may be negative; a negative value indicates that the adjustment capability exceeds the requirement, while a positive value indicates that the adjustment capability is insufficient.

[0115] It should be understood that the larger the difference value, the greater the overall deviation between the schedulable resource adjustment capacity and the demand of the power grid dispatch instructions, and the lower the degree of matching; the smaller the value, the smaller the overall deviation and the higher the degree of matching.

[0116] It should be understood that the trend is used to assess the dynamic evolution of the adjustability of schedulable resources, and to avoid inaccurate matching assessments due to instantaneous fluctuations or unidirectional decay of capabilities.

[0117] Optionally, the method for determining the trend is as follows: determine a past time window (e.g., the past 1 minute), determine the adjustable power range sequence within the past time window, perform linear fitting on the sequence, and determine the slope of the fitted line as the trend.

[0118] Understandably, if the slope is positive, it indicates that the adjustable power range is increasing, meaning that the regulation capability of the schedulable resources is improving; if the slope is negative, it indicates that the adjustable power range is decreasing, meaning that the regulation capability of the schedulable resources is decreasing; if the slope is zero or close to zero, it indicates that the adjustable power range remains stable, and the regulation capability has not changed significantly.

[0119] Optionally, to improve computational stability, the slope of the fitted line can be smoothed or normalized (e.g., maximum-minimum normalization) to keep its value range within a reasonable interval (e.g., [-0.5, 0.5]) to facilitate subsequent calculations.

[0120] Optionally, the matching degree satisfies the following formula:

[0121]

[0122] in, Indicates schedulable resources The degree of matching with power grid dispatch instructions, Represents the instruction feature vector. Indicates schedulable resources capability feature vector Representing instruction feature vectors and schedulable resources The difference between the capability feature vectors This represents a very small positive number (such as 0.001) and is used to prevent the denominator from being zero. Indicates schedulable resources The trend of the adjustable power range in the current adjustment direction. This represents a normalization function used to map input values ​​to a uniform dimensionless interval (such as [0, 1]). Common methods include max-min normalization.

[0123] In this formula, the instruction feature vector and schedulable resources The greater the difference between the capability feature vectors, the smaller the matching degree. This can be understood as basic matching degree; Reflecting the changing trend of resource regulation capacity, if , indicating schedulable resources If the regulatory capacity is enhanced, then This amplifies the basic matching degree. , indicating schedulable resources If the regulatory capacity decreases, then It has a decaying effect on the basic matching degree. ,Should This does not affect the basic matching degree.

[0124] The methods provided in S201-S203 above first clarify the current adjustment direction of the power grid dispatch and extract multi-dimensional command feature vectors. Then, for the current adjustment direction, they extract the capability feature vectors of each schedulable resource. Next, they quantify the static differences between the command feature vectors and capability feature vectors from multiple dimensions. At the same time, they combine the dynamic change trend of the adjustable power range to determine the matching degree. This can more realistically reflect the real-time adaptability of schedulable resources to power grid dispatch commands, further improving the accuracy and reliability of matching degree analysis and providing a more accurate reference for subsequent resource dispatch decisions.

[0125] Combination Figure 1 ,like Figure 3 As shown, in one implementation of this application embodiment, the above-mentioned S103 can be specifically implemented by S301-S303.

[0126] S301. Based on the current adjustment direction, the available power of each dispatchable resource in the current adjustment direction, the maximum adjustable power of each dispatchable resource in the current adjustment direction, and the matching degree between each dispatchable resource and the grid dispatch command, determine the adjustment direction demand characteristics of each dispatchable resource.

[0127] In one alternative implementation, the ratio between the available power of each dispatchable resource in the current adjustment direction and the maximum adjustable power in the current adjustment direction can be used to determine the available adjustment capacity ratio of each dispatchable resource; based on the matching degree between each dispatchable resource and the grid dispatch command and the available adjustment capacity ratio of each dispatchable resource, the adjustment direction demand characteristics of each dispatchable resource can be determined.

[0128] It should be understood that the available adjustability ratio is used to characterize the percentage of a schedulable resource's actual adjustable capacity relative to its maximum potential capacity in its current state.

[0129] The closer the available adjustment capacity ratio is to 1, the closer the available adjustment capacity of the schedulable resources is to its limit in the current adjustment direction. The closer it is to 0, the closer the available adjustment capacity of the schedulable resources is to its limit in the current adjustment direction, making it difficult to undertake more adjustment tasks.

[0130] Optionally, the sum of the matching degree of a schedulable resource and the proportion of its available adjustment capacity can be determined as the adjustment direction demand characteristic.

[0131] It should be understood that the demand characteristic of the adjustment direction represents the available adjustment capacity of a schedulable resource in the current adjustment direction.

[0132] The method described above for determining the demand characteristics of the adjustment direction combines the proportion of available adjustment capacity with the matching degree to jointly determine the demand characteristics of the adjustment direction. It can accurately reflect the comprehensive ability of schedulable resources to undertake tasks in the current adjustment direction, taking into account both the matching quality of resource and instruction characteristics and the actual available capacity of resources in the current adjustment direction, ensuring that resources with high matching degree and spare capacity can be identified first.

[0133] S302. Based on the current measured power, the planned power sequence for the whole day, and the planned power sequence for the short-term power of each schedulable resource, determine the degree of state deviation for each schedulable resource.

[0134] In one alternative implementation, the planned power sequence for the whole day and the planned power sequence for the short time can be interpolated to obtain the planned power at the current moment in the day-ahead phase and the planned power at the current moment in the intraday rolling correction phase. The average of the planned power at the current moment in the day-ahead phase and the planned power at the current moment in the intraday rolling correction phase is determined as the target planned power. Based on the difference between the current measured power and the target planned power of each schedulable resource, the degree of state deviation of each schedulable resource is determined.

[0135] It should be understood that the time resolution of the full-day planned power series and the short-time planned power series are different, and their start times may not be aligned. Therefore, it is necessary to perform interpolation processing on the full-day planned power series and the short-time planned power series separately to obtain the planned power value that accurately corresponds to the current time.

[0136] Alternatively, a linear interpolation method can be used to calculate the planned power corresponding to the current time based on the distance ratio between the current time and the two adjacent times.

[0137] It should be noted that the linear interpolation method is existing technology and will not be elaborated here.

[0138] It should be understood that the target planned power comprehensively considers the day-ahead plan based on long-term economic optimization and the intraday rolling adjustment plan based on ultra-short-term forecasts, so that the target planned power can reflect both the long-term optimization intention and the short-term adjustment needs.

[0139] Understandably, the greater the difference between the current measured power and the target planned power, the greater the degree of state deviation.

[0140] Optionally, the degree of state deviation satisfies the following formula:

[0141]

[0142] in, Indicates schedulable resources The degree of state deviation, Indicates schedulable resources The current measured power, Indicates schedulable resources Target planned power, Indicates schedulable resources Rated power, This indicates taking the absolute value.

[0143] In this formula, the absolute value of the deviation The rated power characterizes the deviation between the current measured power and the target planned power. Normalization.

[0144] It should be understood that, under abnormal circumstances, It may be greater than 1, indicating schedulable resources. The actual situation deviates significantly from the planned expectations, exceeding the rated capacity. In this case, the schedulable resource can be... Marked as unavailable, it will not be included in subsequent calculations.

[0145] It should be noted that dispatchable resources refer to entities within a virtual power plant that can participate in power regulation. There are no dispatchable resources with a rated power of 0. Under extreme abnormal conditions, if the rated power of a dispatchable resource is 0, it means that the dispatchable resource has no regulation capability, and therefore the dispatchable resource will not be analyzed or controlled.

[0146] The above-mentioned method for determining the degree of state deviation integrates the planned power of multi-stage plans at different time scales at the current moment, making the target planned power more in line with the current scheduling needs. At the same time, it quantifies the state deviation by using the difference between the current measured power and the target planned power, which can accurately reflect the degree of deviation between the real-time operating status of schedulable resources and the multi-stage plan.

[0147] S303. Based on the adjustment direction demand characteristics and state deviation degree of each schedulable resource, determine the coordination index of each schedulable resource.

[0148] It should be understood that the greater the demand for adjustment direction, the greater the coordination index; conversely, the greater the degree of state deviation, the smaller the coordination index.

[0149] Optionally, the coordination index satisfies the following formula:

[0150]

[0151] in, Indicates schedulable resources The coordination index, Indicates schedulable resources The degree of state deviation, Indicates schedulable resources The adjustment direction demand characteristics.

[0152] In this formula, This represents the state consistency factor; the larger the state consistency factor, the greater the coordination index.

[0153] The methods provided in S301-S303 above combine matching degree, real-time status data, and multi-stage plan data to first determine the adjustment direction and demand characteristics of resources, then analyze the degree of state deviation of resources, and finally integrate the two to obtain a coordination degree index. This index can reflect both the adaptability of schedulable resources to the direction of the current scheduling instructions and the degree of fit between the real-time status of resources and the multi-stage plan. It realizes a multi-dimensional comprehensive evaluation of resource status, effectively solves the problem of neglecting the consistency between resource status and plan in traditional control, and can provide a comprehensive status basis for subsequent resource weight adjustment.

[0154] It should be noted that the order of the embodiments described above is merely for descriptive purposes and does not represent the superiority or inferiority of the embodiments. The processes depicted in the accompanying drawings do not necessarily require a specific or sequential order to achieve the desired result. In some embodiments, multitasking and parallel processing are also possible or may be advantageous.

[0155] The various embodiments in this specification are described in a progressive manner. The same or similar parts between the various embodiments can be referred to each other. Each embodiment focuses on describing the differences from other embodiments.

Claims

1. A multi-stage optimization control method for a virtual power plant considering supply load, characterized in that, include: The system acquires grid dispatch instructions, real-time status data of each dispatchable resource within the virtual power plant, and multi-stage planning data. The dispatchable resources include supply-side resources and load-side resources. The multi-stage planning data includes day-ahead planning data and intraday rolling correction planning data. The grid dispatch instructions include the power to be allocated. Based on the power grid dispatch instructions and the real-time status data of each schedulable resource, the matching degree between each schedulable resource and the power grid dispatch instructions is determined. Based on the matching degree between each schedulable resource and the power grid dispatching command, the real-time status data of each schedulable resource, and the multi-stage planning data, the coordination index of each schedulable resource is determined. The coordination index is used to characterize the degree of consistency between the real-time status data of the schedulable resource and the multi-stage planning data. Based on the matching degree and coordination degree index of each schedulable resource, determine the dynamic adjustment weight of each schedulable resource; Based on the power of the instruction to be allocated and the dynamic adjustment weight of each schedulable resource, control instructions for each schedulable resource are generated and issued.

2. The multi-stage optimization control method for a virtual power plant considering supply load according to claim 1, characterized in that, The determination of the matching degree between each schedulable resource and the power grid dispatch command based on the power grid dispatch command and the real-time status data of each schedulable resource includes: The current adjustment direction is determined based on the power grid dispatch command, and the command feature vector is extracted from it. The current adjustment direction is the power adjustment direction indicated by the power grid dispatch command at the current moment. The command feature vector includes at least the power change amplitude, the direction change frequency, and the duration of continuous same direction. Based on the real-time status data of each schedulable resource, the capability feature vector of each schedulable resource in the current adjustment direction is extracted. The capability feature vector includes at least the adjustable power range, response speed and duration of the schedulable resource in the current adjustment direction. Based on the instruction feature vector and the capability feature vector of each schedulable resource, the matching degree between each schedulable resource and the power grid dispatch instruction is determined.

3. The multi-stage optimization control method for a virtual power plant considering supply load according to claim 2, characterized in that, The determination of the matching degree between each schedulable resource and the power grid dispatch command based on the instruction feature vector and the capability feature vector of each schedulable resource includes: Based on the difference between the power change amplitude and the adjustable power range, the difference between the direction change frequency and the response speed, and the difference between the continuous unidirectional duration and the available duration, the difference between the instruction feature vector and the capability feature vector of each schedulable resource is determined. Based on the real-time status data of each schedulable resource, determine the trend of the change in the adjustable power range of each schedulable resource in the current adjustment direction; Based on the difference between the instruction feature vector and the capability feature vector of each schedulable resource, as well as the changing trend of the adjustable power range of each schedulable resource in the current adjustment direction, the matching degree between each schedulable resource and the grid dispatch instruction is determined.

4. The multi-stage optimization control method for a virtual power plant considering supply load according to claim 1, characterized in that, The real-time status data of each dispatchable resource includes the available power in the current adjustment direction, the maximum adjustable power in the current adjustment direction, and the current measured power. The day-ahead planning data includes the full-day planned power sequence, and the intraday rolling correction planning data includes the short-time planned power sequence. The coordination index of each dispatchable resource is determined based on the matching degree between each dispatchable resource and the grid dispatch command, the real-time status data of each dispatchable resource, and the multi-stage planning data, including: Based on the current adjustment direction, the available power of each dispatchable resource in the current adjustment direction, the maximum adjustable power of each dispatchable resource in the current adjustment direction, and the matching degree between each dispatchable resource and the grid dispatch command, the adjustment direction demand characteristics of each dispatchable resource are determined. Based on the current measured power, the full-day planned power sequence, and the short-term planned power sequence of each schedulable resource, determine the degree of state deviation for each schedulable resource; Based on the adjustment direction demand characteristics and state deviation degree of each schedulable resource, the coordination index of each schedulable resource is determined.

5. The multi-stage optimization control method for a virtual power plant considering supply load according to claim 4, characterized in that, The determination of the demand characteristics of each dispatchable resource in the current adjustment direction, based on the current adjustment direction, the available power of each dispatchable resource in the current adjustment direction, the maximum adjustable power of each dispatchable resource in the current adjustment direction, and the matching degree between each dispatchable resource and the grid dispatch command, includes: The ratio of the available power of each schedulable resource in the current adjustment direction to the maximum adjustable power in the current adjustment direction is determined as the proportion of the available adjustment capacity of each schedulable resource. Based on the matching degree between each dispatchable resource and the power grid dispatch command, and the proportion of the available adjustment capacity of each dispatchable resource, the adjustment direction demand characteristics of each dispatchable resource are determined.

6. The multi-stage optimization control method for a virtual power plant considering supply load according to claim 4, characterized in that, The determination of the state deviation degree of each schedulable resource based on the current measured power, the planned power sequence for the whole day, and the planned power sequence for the short-term time includes: Interpolate the full-day planned power sequence and the short-time planned power sequence to obtain the planned power at the current moment in the day-ahead phase and the planned power at the current moment in the intraday rolling correction phase. The average of the planned power at the current moment in the day-ahead phase and the planned power at the current moment in the intraday rolling correction phase is determined as the target planned power. The degree of state deviation for each schedulable resource is determined based on the difference between the current measured power and the target planned power.

7. The multi-stage optimization control method for a virtual power plant considering supply load according to claim 1, characterized in that, The determination of the dynamic adjustment weight for each schedulable resource based on its matching degree and coordination degree index includes: Obtain the base weight of each schedulable resource; Based on the matching degree and coordination degree index of each schedulable resource, determine the weight adjustment factor of each schedulable resource; Based on the base weight and weight adjustment factor of each schedulable resource, the dynamic adjustment weight of each schedulable resource is determined.

8. The multi-stage optimization control method for a virtual power plant considering supply load according to claim 7, characterized in that, Based on the matching degree and coordination degree index of the first schedulable resource, the weight adjustment factor of the first schedulable resource is determined, including: The matching degree and coordination degree index of the first schedulable resource are normalized to obtain the normalized matching degree and normalized coordination degree. The first schedulable resource is any one of all schedulable resources in the virtual power plant. The sum of the normalized matching degree and the normalized coordination degree is determined as the weight adjustment factor for the first schedulable resource.

9. The multi-stage optimization control method for a virtual power plant considering supply load according to claim 1, characterized in that, The process of generating and issuing control instructions for each schedulable resource based on the power of the instruction to be allocated and the dynamic adjustment weight of each schedulable resource includes: The ratio between the dynamic adjustment weight of each schedulable resource and the total dynamic adjustment weight is determined as the normalized adjustment weight of each schedulable resource. The total dynamic adjustment weight is the sum of the dynamic adjustment weights of all schedulable resources. The power of the instruction to be allocated is allocated based on the normalized adjustment weight of each schedulable resource to obtain the target power allocation value of each schedulable resource. Based on the target power allocation value of each schedulable resource, control instructions are generated for each schedulable resource.

10. The multi-stage optimization control method for a virtual power plant considering supply load according to claim 9, characterized in that, The process of allocating the power of the instruction to be allocated based on the normalized adjustment weight of each schedulable resource to obtain the target power allocation value for each schedulable resource includes: The product of the normalized adjustment weight and the power of the command to be allocated is determined as the initial power allocation value; If the initial power allocation value of each schedulable resource is less than its respective maximum adjustable power, the initial power allocation value of each schedulable resource is determined as the target power allocation value. If the initial power allocation value of at least one schedulable resource exceeds the corresponding maximum adjustable power, the maximum adjustable power is determined as the target power allocation value of the at least one schedulable resource, and the remaining power to be allocated is determined. The remaining power to be allocated is redistributed based on the dynamic adjustment weights of other schedulable resources until the total power allocation value of all schedulable resources is within their respective maximum adjustable power. The final total power allocation value allocated to each schedulable resource is determined as the target power allocation value for each schedulable resource. The other schedulable resources are schedulable resources other than the at least one schedulable resource among all schedulable resources in the virtual power plant.

Images