Virtual power plant adjustable load hierarchical access method based on dynamic priority scheduling

Abstract

This invention relates to the field of power system management technology, specifically to a method for hierarchical access of adjustable loads in a virtual power plant based on dynamic priority scheduling. This invention systematically analyzes the historical power variation and regulation characteristics of power-consuming terminals, constructs a comprehensive relative fit, accurately selects power-capable power-consuming terminals, and achieves refined screening and hierarchical layering of adjustable load resources. It overcomes the limitations of traditional methods that rely on manual experience or single indicators for screening, ensuring that the group of dispatchable power-consuming terminals simultaneously possesses sufficient regulation potential, rapid response speed, and high execution stability, thereby improving the accuracy and reliability of virtual power plant response scheduling.

Description

Technical Field

[0001] This invention relates to the field of power system management technology, specifically to a method for hierarchical access of adjustable loads in virtual power plants based on dynamic priority scheduling. Background Technology

[0002] Virtual power plants, as a technology that aggregates distributed adjustable resources to participate in the coordinated operation of the power grid, are an important carrier for integrating distributed power sources, energy storage systems, and adjustable loads. During power grid operation, it is necessary to maintain a real-time power balance between generation and consumption. To cope with the time-varying nature of load demand and maintain system frequency stability, virtual power plants must precisely regulate and control the large number of adjustable load resources they aggregate.

[0003] However, existing technologies for scheduling and managing adjustable load resources either rely on a single indicator or on subjective judgment based on human experience, making it impossible to accurately identify truly adjustable and high-performance power terminals from a large pool of users. This can easily lead to slow response and large execution deviations in the dispatched power terminal group, making it difficult to reliably complete dispatch instructions, and may even introduce disturbances into the power grid.

[0004] In addition, there are significant differences in power grid dispatching needs, but traditional dispatching strategies are often relatively fixed, making it difficult to dynamically adjust the priority order of power terminal calls according to the real-time needs of dispatching tasks, resulting in insufficient, excessive, or ineffective dispatching. Summary of the Invention

[0005] The purpose of this invention is to overcome the shortcomings of the prior art. Through multi-dimensional evaluation and dynamic clustering and hierarchical structure, it achieves accurate screening of power-accessible power terminals and refined hierarchical structure of adjustable load resource layers, ensuring that dispatching instructions correspond and match with adjustable load resource layers, and improving the accuracy and reliability of virtual power plant response dispatch.

[0006] The technical solution adopted by the present invention to solve its technical problem is: a method for hierarchical access of adjustable load in a virtual power plant based on dynamic priority scheduling, which includes the following steps: acquiring load data of all power terminals in the virtual power plant area; determining the historical power change regulation characteristics of each power terminal based on the load data, wherein the historical power change regulation characteristics include power fluctuation amplitude, power regulation response time, and regulation stability parameters used to characterize power deviation.

[0007] The power fluctuation amplitude, power regulation response time, and regulation stability parameters of all power terminals are sorted from smallest to largest, and the corresponding median values ​​are determined.

[0008] Based on each power terminal: the power fluctuation amplitude is compared with the corresponding median value to obtain the fluctuation amplitude adaptation coefficient; the median value of the power regulation response time is compared with the power regulation response time to obtain the response time adaptation coefficient; the median value of the regulation stability parameter is compared with the regulation stability parameter to obtain the stability adaptation coefficient; the weighted average of the fluctuation amplitude adaptation coefficient, response time adaptation coefficient, and stability adaptation coefficient is calculated to obtain the comprehensive relative adaptation degree.

[0009] Sort all power terminals by their overall relative suitability from highest to lowest, select power terminals sequentially starting from the beginning of the sorting sequence, and simultaneously accumulate the power fluctuation amplitude; when the accumulated power fluctuation amplitude value is greater than or equal to the amount of power to be reduced for the first time, the selection is terminated, and the selected power terminals are designated as power-accessible power terminals.

[0010] Based on the historical power change adjustment characteristics of the power-accessible power terminals, an adjustable load feature set is constructed; cluster analysis of the adjustable load feature set is performed to divide the power-accessible power terminals into multiple adjustable load resource layers.

[0011] Receive scheduling instructions containing the amount of power consumption to be reduced within the target time period, and construct scheduling priorities for each adjustable load resource layer based on the scheduling instructions and adjustable load feature sets.

[0012] Based on scheduling priority, all power-accessible terminals within each adjustable load resource layer are sorted and selected from high to low until the total adjustable power meets the scheduling instructions, and corresponding load control instructions are generated.

[0013] The load control command is sent to the corresponding power-accessible terminal for execution, the execution effect is monitored, and the operating status of each power-accessible terminal is updated.

[0014] Compared with the prior art, the present invention has the following beneficial effects: (1) The present invention systematically analyzes the historical power change adjustment characteristics of the power-accessible power terminals, constructs a comprehensive relative fit, accurately screens the power-accessible power terminals, and realizes the fine-grained layering of the adjustable load resource layer. It overcomes the one-sidedness of the traditional method of relying on manual experience or a single indicator for screening, and ensures that the group of power-accessible power terminals to be dispatched simultaneously has sufficient adjustment potential, fast response speed and high execution stability, thereby improving the accuracy and reliability of virtual power plant response dispatch.

[0015] (2) Based on the specific requirements of each scheduling instruction, this invention dynamically calculates the priority of each adjustable load resource layer and each power-capable call terminal. It always prioritizes calling the power-capable call terminal within the adjustable load resource layer that best suits the current scheduling instruction. Simultaneously, a closed-loop feedback mechanism is introduced to update the corresponding power fluctuation amplitude, power adjustment response time, and adjustment stability parameters based on the actual execution effect of the power-capable call terminals, thus achieving self-learning and continuous optimization of the scheduling strategy. Attached Figure Description

[0016] To more clearly illustrate the technical solutions of the embodiments of the present invention, the accompanying drawings used in the description of the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0017] Figure 1 This is a flowchart of the hierarchical access method of the present invention.

[0018] Figure 2 This is a flowchart illustrating the process of dividing the adjustable load resource layer in this invention.

[0019] Figure 3 This is a flowchart illustrating the sorting of all power-enabled mobile terminals within each adjustable load resource layer according to the present invention.

[0020] Figure 4 This is a flowchart illustrating the process of selecting power from high to low until the total adjustable power satisfies the scheduling command. Detailed Implementation

[0021] Various exemplary embodiments of the present invention will now be described in detail with reference to the accompanying drawings. It should be noted that, unless otherwise specifically stated, the relative arrangement, numerical expressions, and values ​​of the components and steps set forth in these embodiments do not limit the scope of the invention. Furthermore, it should be understood that, for ease of description, the dimensions of the various parts shown in the drawings are not drawn to actual scale.

[0022] The following description of at least one exemplary embodiment is merely illustrative and is in no way intended to limit the invention or its application or use. Techniques, methods, and apparatus known to those skilled in the art may not be discussed in detail, but where appropriate, such techniques, methods, and apparatus should be considered part of the specification.

[0023] In all examples shown and discussed herein, any specific values ​​should be interpreted as merely exemplary and not as limitations. Therefore, other examples of exemplary embodiments may have different values.

[0024] Please see Figure 1As shown, the method for hierarchical access of adjustable loads in a virtual power plant based on dynamic priority scheduling includes the following steps: acquiring load data of all power terminals within the virtual power plant area; determining the historical power change regulation characteristics of each power terminal based on the load data, and selecting power-adjustable power terminals.

[0025] Not all electrical terminals are adjustable. The power changes of some loads are passive fluctuations or rigid demands. For example, the refrigerator compressor starts and stops according to the internal temperature, which the user cannot actively control; the power of the server needs to be stable at a fixed value, and adjustment will cause equipment failure.

[0026] Historical power variation regulation characteristics are core attributes directly related to regulation capability extracted by analyzing long-term power data from electricity terminals. These historical power characteristics allow for precise differentiation:

[0027] Non-power-accessible terminals with historically stable power or fluctuations triggered only by the device's own logic; and power-accessible terminals whose historical power can be actively reduced under specific conditions with controllable fluctuation range. This provides a benchmark for subsequent hierarchical access, ensuring the feasibility and reliability of dynamic priority scheduling.

[0028] In the implementation of the above scheme, the historical power change regulation characteristics of each power terminal are determined based on load data, including: dividing the load data according to a preset time window, calculating the difference between the maximum and minimum power values ​​within each time window, and taking the arithmetic mean of all differences as the power fluctuation amplitude.

[0029] When the historical power of a power terminal is stable over a long period with minimal fluctuations, it indicates that its power adjustment space is limited; when the historical power of a power terminal exhibits a stable and repeatable fluctuation range, it indicates that it has the potential to intervene in power changes through external scheduling means.

[0030] The adjustability of a power terminal is reflected in its power being able to change as needed within a specific range, and this change process is inevitably accompanied by fluctuations in the power value. By extracting the power fluctuation amplitude, the range boundary of the power change of the power terminal within the operating cycle can be directly reflected, providing a quantitative standard for distinguishing between power-adjustable power terminals and non-power-adjustable power terminals, and avoiding screening errors caused by subjective judgment based on the type of power terminal.

[0031] The preset time window can be determined according to the load type of the power terminal, aiming to unify the time scale. In this invention, for example: for temperature-controlled loads such as air conditioning systems and electric water heaters, the preset time window can be 30 minutes; for equipment with frequent start-stop operations such as industrial motors and water pumps, the preset time window can be 10 minutes.

[0032] Calculate the time difference between the issuance of the load control command and the first change in power, and take the arithmetic mean of all time differences as the power regulation response time.

[0033] The period from the start to the end of the load control command is taken as the target time period. The deviation ratio between the required power reduction and the actual power reduction within the target time period is calculated, and the arithmetic mean of all deviation ratios is taken as the regulation and stability parameter.

[0034] The power fluctuation amplitude, power regulation response time, and regulation stability parameters are collected as historical power change regulation characteristics.

[0035] Power fluctuation amplitude characterizes the range of adjustable power at the user terminal. The larger the fluctuation amplitude, the more power the user terminal can increase or decrease when receiving load dispatch instructions, and the more sufficient the adjustment capacity it can contribute. Conversely, the smaller the fluctuation amplitude, the limited the adjustment capacity it contributes. If used in dispatch scenarios with large-scale power reduction, it will lead to insufficient actual power reduction, failing to meet the total demand of the power grid for load regulation.

[0036] Power regulation response time characterizes the speed at which a power terminal responds to load control commands. A shorter response time means the power terminal can initiate power regulation faster, making it suitable for time-sensitive dispatching scenarios and meeting the timeliness requirements of virtual power plant dispatching. Using slow-response power terminals in emergency frequency regulation scenarios will lead to delayed execution of load control commands and grid frequency instability; conversely, using fast-response power terminals in low-timeliness scenarios will result in resource waste.

[0037] Adjusting stability parameters characterizes the reliability and accuracy of power terminals in executing load control commands. Better stability parameters, i.e., a smaller deviation percentage, mean a higher consistency between the actual adjustment effect of the power terminal and the load control command requirements, and a more stable adjustment process. Conversely, poor stability parameters can lead to excessive or insufficient actual power reduction, interfering with grid stability, or indicate potential equipment failures or control logic defects in the power terminal, which could increase dispatching risks if connected to the system.

[0038] By combining power fluctuation amplitude, power regulation response time, and regulation stability parameters, the effective execution of load regulation commands is ensured, which facilitates the subsequent realization of precise hierarchical access and ensures the stable operation of the power grid.

[0039] In implementing the above scheme, screening power-adjustable terminals includes: sorting the power fluctuation amplitude, power adjustment response time, and adjustment stability parameters of all power terminals from smallest to largest, and determining the corresponding median value.

[0040] Based on each power terminal: the power fluctuation amplitude is compared with the corresponding median value to obtain the fluctuation amplitude adaptation coefficient; the median value of the power regulation response time is compared with the power regulation response time to obtain the response time adaptation coefficient; the median value of the regulation stability parameter is compared with the regulation stability parameter to obtain the stability adaptation coefficient.

[0041] Power fluctuation amplitude is a positive indicator. A larger value indicates greater power adjustment potential and better performance. Adjustment stability parameter is a negative indicator. A smaller value indicates smaller deviation, higher stability, and better performance. Power adjustment response time is also a negative indicator. A smaller value indicates faster response speed and better performance. By using different ratio methods, both positive and negative indicators are positively correlated, eliminating significant dimensional differences and avoiding ambiguity in subsequent priority ranking.

[0042] The weighted average of the fluctuation amplitude adaptation coefficient, response time adaptation coefficient, and stability adaptation coefficient is calculated to obtain the comprehensive relative adaptation degree.

[0043] It should be noted that when configuring corresponding weights for the fluctuation amplitude adaptation coefficient, response time adaptation coefficient, and stability adaptation coefficient, the weights can be determined according to the actual power grid dispatching needs.

[0044] In this invention, for example: if the current core demand of the power grid is to rapidly reduce high-power loads, the weights of the fluctuation amplitude adaptation coefficient and the response time adaptation coefficient are 0.5 and 0.4, respectively, and the weight of the stability adaptation coefficient is 0.1; if the power grid has extremely high requirements for regulation accuracy, such as in sensitive industrial load scenarios, the weight of the stability adaptation coefficient is 0.6, and the weights of the response time adaptation coefficient and the fluctuation amplitude adaptation coefficient are 0.2 and 0.2, respectively; if the power grid demand is conventional flexible regulation without special emphasis, the weights of the fluctuation amplitude adaptation coefficient, the response time adaptation coefficient, and the stability adaptation coefficient can be 0.4, 0.3, and 0.3, respectively.

[0045] All power terminals are sorted by their overall relative suitability from highest to lowest. Power terminals are selected sequentially starting from the beginning of the sorting sequence, and the power fluctuation amplitude is accumulated synchronously.

[0046] The above-described sorting and selection method essentially involves dynamically stratifying all power-consuming terminals within the virtual power plant area according to their overall performance. Starting selection from the beginning of the sequence prioritizes power-consuming terminals with the best overall capabilities. This ensures that the scheduling task is accomplished with the fewest number of terminals, the lowest control complexity, and the highest response reliability, while still meeting the power reduction requirements within the target time period. This avoids problems such as low resource utilization efficiency, redundant control commands, and unstable responses caused by random or disordered selection.

[0047] When the cumulative value of power fluctuation amplitude is greater than or equal to the amount of power that needs to be reduced for the first time, the selection process will terminate, and the selected power terminals will be used as power-accessible power terminals.

[0048] The core requirement of the load dispatch instructions received by the virtual power plant is to complete a specific total power reduction within the target time period, that is, to reduce the amount of electricity consumption. Taking this instruction target as the ultimate guide ensures the purposefulness and practicality of the screening process.

[0049] The overall relative fit ranking sequence is a queue with performance ranked from best to worst. The power terminals at the front of the queue are high-quality resources; as selection progresses, the performance of subsequent power terminals decreases progressively, leading to a decline in the reliability gain and an increase in control complexity and uncertainty. Selection is terminated immediately when the accumulated power fluctuation amplitude first exceeds or equals the amount of power to be reduced, meaning that within the current scheduling cycle, only the necessary and optimal subset of power terminals is used.

[0050] Blindly selecting all power terminals or selecting power terminals whose cumulative value far exceeds the amount of power reduction required will lead to a large number of poorly adapted power terminals being connected to the dispatch system, causing over-dispatch and wasting resources, while also introducing uncertainty.

[0051] Based on the historical power change adjustment characteristics of the power-adjustable terminal, an adjustable load feature set is constructed.

[0052] In the implementation of the above scheme, the process of constructing the adjustable load feature set is as follows: assign unique identification information to each power adjustable power terminal; associate its unique identification information with the corresponding power fluctuation amplitude, power regulation response time, regulation stability parameters and comprehensive relative adaptability to form a feature vector; and collect all feature vectors to construct the adjustable load feature set.

[0053] The virtual power plant area contains a large number of power-consuming terminals, which vary in model, type, and installation location, necessitating unique management and precise control. Assigning a unique identifier to each power-consuming terminal is a prerequisite for achieving refined scheduling.

[0054] In the absence of unique identification information, load dispatch instructions may be issued to non-target power users, leading to dispatch chaos.

[0055] Different dispatching scenarios place different demands on the core regulation capabilities of power-consuming terminals. For example, rapid power reduction dispatching scenarios emphasize power regulation potential, while emergency frequency regulation dispatching scenarios emphasize response speed and regulation accuracy. By comprehensively considering relative adaptability, dynamic matching between the performance of power-consuming terminals and dispatching scenarios can be achieved.

[0056] By constructing a set of adjustable load features that includes unique identifiers, power fluctuation amplitude, power regulation response time, regulation stability parameters, and overall relative adaptability, the problem of biased decision-making caused by relying on a single indicator for screening and scheduling is avoided. For example, a power terminal with a large regulation capacity but a response time exceeding 5 minutes should be excluded in an emergency frequency regulation scenario. However, if the power regulation response time is missing, it may be mistakenly selected as a power-adjustable power terminal, leading to scheduling failure.

[0057] Please see Figure 2 As shown, the adjustable load feature set clustering analysis divides the power-accessible power terminals into multiple adjustable load resource layers, including: standardizing the power fluctuation amplitude, power regulation response time, regulation stability parameters and comprehensive relative fit in each feature vector.

[0058] It should be noted that, for example, in this invention, Z-score standardization is used to process the power fluctuation amplitude, power adjustment response time, adjustment stability parameter and overall relative fit in the feature vector.

[0059] The following steps are executed iteratively until all power-reclaimable terminals are clustered: the distance between the feature vector of the current power-reclaimable terminal and the center feature vector of all existing clusters is calculated using Euclidean distance; the minimum distance and its corresponding cluster are extracted.

[0060] Among them, the central eigenvector is the benchmark characterizing the central tendency of the adjustment features of all power callable terminals in the cluster, and it is determined by the arithmetic mean of the eigenvectors of all power callable terminals in the cluster.

[0061] If the minimum distance is less than the preset distance threshold, the power-enabled mobile terminal is assigned to the corresponding cluster, and the center feature vector of the cluster is recalculated; if the minimum distance is greater than or equal to the preset distance threshold, a new cluster is created with the power-enabled mobile terminal as the core.

[0062] The smaller the spatial distance of the feature vectors, the smaller the differences in power fluctuation amplitude, power regulation response time, regulation stability parameters, and overall relative adaptability among the power-requisite terminals, indicating more similar regulation capabilities and a basis for grouping them into the same cluster. Conversely, a larger spatial distance indicates significant differences in the regulation capabilities of the power-requisite terminals, and forcibly grouping them into an existing cluster would lead to high feature dispersion within the cluster. Allowing the creation of new clusters ensures that all power-requisite terminals have a designated affiliation.

[0063] The adjustable load feature set consists of discrete power-adjustable terminals, and the spatial distance relationship between the feature vectors of each power-adjustable terminal has no predetermined pattern.

[0064] A single determination can only identify the association between adjacent power-accessible terminals and cannot form a complete cluster. Iterative execution can gradually incorporate adjacent and similar power-accessible terminals that are linked in a chain into the same cluster through successive comparisons. This avoids splitting power-accessible terminals that should be classified into one category into different clusters, ensuring that subsequent automated and accurate layering is achieved and guaranteeing the scientific nature of the layering.

[0065] It should be noted that the preset distance threshold can be determined based on historical data and is used to objectively divide clusters according to the spatial proximity of feature vectors. Its value range can be 0.8-1.4. In this invention, for example, for scheduling scenarios with extremely high requirements for response speed and stability, the preset distance threshold can be 0.85; for scheduling scenarios with high requirements for adjustment capacity, the preset distance threshold can be 1.25.

[0066] Calculate the average values ​​of all power fluctuation amplitudes, power regulation response times, and regulation stability parameters within each cluster, and use these values ​​as characterization values ​​for regulation capability, response speed, and stability of each cluster.

[0067] Clusters whose regulation capacity, response speed, and stability values ​​fall within a certain range of their respective values ​​are merged into the same adjustable load resource layer, and each adjustable load resource layer is statistically obtained.

[0068] Within a cluster, numerous power-accessible terminals exhibit dispersed power fluctuations, power regulation response times, and regulation stability parameters. Directly utilizing this dispersed data is insufficient to guide scheduling. Therefore, a macroscopic representation of the cluster's collective capability is achieved through regulation capacity, response speed, and stability metrics.

[0069] Requiring the regulation capacity, response speed, and stability values ​​to fall within a certain range not only avoids the mixing of clusters with vastly different capabilities within the same adjustable load resource layer, but also ensures that clusters within the same adjustable load resource layer meet the core requirements of a specific scheduling scenario in terms of regulation capacity, response speed, and regulation accuracy.

[0070] It should be noted that the corresponding ranges for the regulation capability characterization value, response speed characterization value, and stability characterization value can be determined based on the actual regulation capability distribution of the power-adjustable terminals. Dividing the data into ranges rather than single values ​​is to avoid excessive fragmentation of the adjustable load resource layer due to minor differences in a single value, thereby ensuring the consistency and substitutability of the regulation capability of clusters within the same adjustable load resource layer.

[0071] In this invention, for example: the actual distribution data of the regulation capacity characterization value, response speed characterization value, and stability characterization value of all clusters in the adjustable load characteristic set can be statistically analyzed first; if the statistics show that the regulation capacity characterization value is mainly concentrated in the 15kW-80kW range, and forms obvious data clusters in the three sub-ranges of 15kW-35kW, 35kW-60kW, and 60kW-80kW; the response speed characterization value is concentrated in the 5s-45s range, and is densely distributed in the three sub-ranges of 5s-15s, 15s-30s, and 30s-45s; the stability characterization value is concentrated in the 1.5%-9% range, and shows clustering characteristics in the three sub-ranges of 1.5%-3.5%, 3.5%-6.5%, and 6.5%-9%. Based on this actual distribution, the corresponding intervals for the three characterization values ​​can be defined as follows: the basic interval for regulation capacity is 15kW-35kW, the medium-optimal interval is 35kW-60kW, and the high-optimal interval is 60kW-80kW; the basic interval for response speed is 30s-45s, the medium-optimal interval is 15s-30s, and the high-optimal interval is 5s-15s; and the basic interval for stability is 6.5%-9%, the medium-optimal interval is 3.5%-6.5%, and the high-optimal interval is 1.5%-3.5%.

[0072] Receive a scheduling instruction containing the amount of power consumption to be reduced within a target time period. Based on the scheduling instruction and the adjustable load feature set, construct the scheduling priority of each adjustable load resource layer, including: extracting the time length of the target time period and the amount of power consumption to be reduced from the scheduling instruction.

[0073] Based on each adjustable load resource layer: calculate the reciprocal of the ratio of the response speed characterization value to the time length as the response time matching degree; calculate the ratio of the regulation capacity characterization value to the amount of power to be reduced as the regulation capacity matching degree; take the reciprocal of the stability characterization value as the regulation accuracy matching degree.

[0074] The overall priority score is obtained by integrating the matching degree of response timeliness, the matching degree of adjustment capability, and the matching degree of adjustment accuracy.

[0075] In this invention, the response time matching degree, adjustment capability matching degree, and adjustment accuracy matching degree can be processed first by the Min-Max standardization algorithm. Then, the standardized response time matching degree, adjustment capability matching degree, and adjustment accuracy matching degree can be multiplied and accumulated with their corresponding weights by the weighted summation algorithm to obtain the comprehensive priority score of each adjustable load resource layer. The higher the comprehensive priority score, the stronger the comprehensive adaptability of the adjustable load resource layer to the current scheduling task.

[0076] Each weight can be determined based on the core requirements of the scheduling scenario. For example, in a scheduling scenario where the core requirement is rapid response and precise adjustment, the weight for response time matching can be set to 0.5, the weight for adjustment accuracy matching to 0.3, and the weight for adjustment capacity matching to 0.2. Similarly, in a scheduling scenario where the core requirement is to meet large-scale power consumption reduction, the weight for adjustment capacity matching can be set to 0.5, the weight for response time matching to 0.3, and the weight for adjustment accuracy matching to 0.2.

[0077] All adjustable load resource layers are sorted from high to low according to their scheduling priority scores to form scheduling priorities.

[0078] Please see Figure 3 and Figure 4 As shown, based on the scheduling priority, all power-accessible terminals within each adjustable load resource layer are sorted and selected from high to low until the total adjustable power meets the scheduling instructions, and corresponding load control instructions are generated.

[0079] Different adjustable load resource layers have varying adjustment capabilities, response speeds, and adjustment precision, and different scheduling tasks have different priorities for their requirements. For example, when a scheduling task requires rapid response and precise adjustment within seconds or minutes, the priority for adjustable load resource layers is: response speed > adjustment precision > adjustment capability. Adjustable load resource layers with a response speed characteristic value of less than or equal to 8 seconds and a stability characteristic value of less than or equal to 2% should be prioritized. Even if their adjustment capability is slightly lower, rapid and precise adjustment can prevent further frequency instability.

[0080] If priorities are not established, the adaptability of all adjustable load resource layers must be evaluated one by one during scheduling. This is especially problematic when there are a large number of adjustable load resource layers, which can easily lead to decision delays or omissions.

[0081] By prioritizing, the adjustable load resource layer that best suits the current scheduling task can be placed at the top, ensuring that the adjustable load resource layer that can best meet the scheduling task is called first, and avoiding scheduling task failure due to prioritizing non-core adjustable load resource layers.

[0082] It should also be noted that the dispatching instructions clearly define the core quantitative objectives of this regulation. The selection process must be terminated primarily when the total adjustable power reaches the amount of power reduction required by the dispatching instructions. This is to avoid insufficient power reduction due to insufficient selection, or redundant power reduction due to excessive selection, and to ensure that the regulation results accurately match the requirements of the dispatching instructions.

[0083] In the implementation of the above scheme, according to the scheduling priority, all power-accessible call terminals in each adjustable load resource layer are sorted, including: calculating the maximum and minimum values ​​of the power fluctuation amplitude, power regulation response time, regulation stability parameters and comprehensive relative adaptability of all power-accessible call terminals in each adjustable load resource layer.

[0084] Based on the maximum and minimum values, the feature vectors of each power-adjustable callable terminal within the current adjustable load resource layer are standardized. The Min-Max standardization algorithm can be used in this invention.

[0085] The ratio of the standardized power regulation response time to the duration is subtracted from the setpoint to determine the individual response time matching degree; the ratio of the standardized power fluctuation amplitude to the amount of power to be reduced is determined to determine the individual regulation capability matching degree; and the reciprocal of the standardized regulation stability parameter is determined to determine the individual regulation accuracy matching degree.

[0086] The aforementioned setpoint is a dimensionless preset reference benchmark used to represent the upper limit threshold of the ratio of the standardized power regulation response time to the time length. When this ratio equals the upper limit threshold, the individual response time matching degree is 0, corresponding to the critical state where the power regulation response time is exactly equal to the target time period length. If the ratio exceeds the upper limit threshold, the matching degree is negative, indicating that the power callable terminal is responding too slowly; if the ratio is below the upper limit threshold, the matching degree is positive, and the larger the value, the faster the power callable terminal's response speed is compared to the scheduling task requirements, and the higher the matching degree. In this invention, the setpoint is set to 1.

[0087] Different scheduling tasks have rigid differences in their response speed requirements; for example, emergency frequency regulation requires a response time in seconds, while regular throttling allows for a response time in minutes.

[0088] Ignoring the matching degree of individual response time may cause the power available terminal to miss the best control window due to slow response after the dispatch command is issued, resulting in further frequency instability and directly affecting the timeliness of the control task.

[0089] The dispatching instructions rely on the total amount of actual power reduction by the power-accessible terminals. If the power-accessible terminals have insufficient adjustment capabilities, even the fastest response will not be able to meet the demand.

[0090] The greater the individual adjustment capability matching degree, the more sufficient the actual power reduction of the power-accessible terminal is.

[0091] Ignoring the matching degree of individual adjustment capabilities may result in an excessive number of selected power-accessible terminals, increasing the complexity of issuing control instructions; or the actual reduction in power consumption by a single power-accessible terminal may be insufficient, directly causing a gap in the amount of power consumption that needs to be reduced.

[0092] Poor power regulation accuracy of the power dispatch terminal can lead to secondary fluctuations in grid power and may cause frequent start-ups and shutdowns of user-side equipment.

[0093] The greater the individual adjustment accuracy matching degree, the smaller the power available power terminal adjustment deviation and the higher the adjustment accuracy.

[0094] Ignoring the matching degree of individual regulation precision may lead to chaos in the power grid after regulation, or cause user complaints, directly affecting the quality of regulation and system safety.

[0095] The individual response time matching degree, individual regulation capability matching degree, individual regulation accuracy matching degree, and standardized comprehensive relative fit degree of each power-capable callable terminal are weighted and integrated to obtain a comprehensive ranking score.

[0096] The weighted fusion mentioned above is implemented through a weighted summation algorithm. It should be noted that each weight can be configured according to the current scheduling instruction type. In this invention, for example: when the scheduling instruction is a routine peak load reduction, and the core requirement is to meet the large-scale, continuous power consumption reduction and the required power reduction amount, the weights can be configured as follows: individual regulation capability matching degree weight 0.4, individual response time matching degree weight 0.2, individual regulation accuracy matching degree weight 0.2, and standardized comprehensive relative fit degree weight 0.2.

[0097] All power-accessible terminals within each adjustable load resource layer are sorted from highest to lowest based on their comprehensive ranking scores. When the comprehensive ranking scores are the same, they are arranged from largest to smallest based on their power fluctuation amplitude. A priority ranking sequence of all power-accessible terminals within each adjustable load resource layer is then generated.

[0098] Prioritizing the use of power-requesting terminals with higher overall ranking scores means that within the same adjustable load resource layer, the global optimal principle of prioritizing the most suitable power-requesting terminal is still followed. This ensures that even within the same adjustable load resource layer, the system can further distinguish which power-requesting terminals are the best choices for completing the current task, thereby guaranteeing the best overall response performance of the entire virtual power plant.

[0099] With identical overall performance scores, the power-accessible terminals with larger power fluctuation ranges represent those that can contribute a greater actual reduction in power consumption per call. Arranging them from largest to smallest power fluctuation range is an efficiency-first strategy, allowing for faster fulfillment of the power consumption reduction task required by the adjustable load resource layer with fewer calls and simpler control commands. It also reduces the potential negative impacts of frequently starting and stopping power-accessible terminals with varying adjustment capacities.

[0100] In the implementation of the above scheme, the process of selecting from high to low until the total adjustable power meets the scheduling instructions includes: screening the adjustable load resource layer with the highest priority, and then selecting each power adjustable power terminal in turn.

[0101] The power fluctuation amplitude of the selected power adjustable terminal is accumulated synchronously and used as the current adjustable power sum.

[0102] If the current total adjustable power does not reach the amount of power that needs to be reduced, and all power-accessible terminals within the current adjustable load resource layer have been selected, then the next adjustable load resource layer is selected from high to low according to the scheduling priority, and the selection and accumulation are repeated.

[0103] Reducing power consumption is a rigid objective that the power grid must achieve when issuing dispatch instructions. When the highest priority adjustable load resource layer cannot independently meet the objective, the system must be able to call upon other adjustable load resource layers to fill the gap; otherwise, dispatching will fail.

[0104] By prioritizing the highest-priority adjustable load resource layer and then utilizing the secondary adjustable load resource layers, a multi-adjustable load resource layer collaborative supplementation control mode is formed. This avoids blindly mixing and calling all adjustable load resource layers from the beginning, or skipping high-priority adjustable load resource layers and directly calling low-priority adjustable load resource layers, which would lead to high-priority adjustable load resource layers being idle and low-priority adjustable load resource layers being overused. This ensures the overall accuracy and reliability of the virtual power plant's response dispatch.

[0105] If the total adjustable power is greater than or equal to the amount of power that needs to be reduced, the selection process is terminated, and the selected adjustable power terminals are recorded to form a set of control terminals.

[0106] The goal of dispatching is to precisely meet the required reduction in power consumption, rather than mobilizing all adjustable load resources. If the current total adjustable power is far greater than the required reduction, the actual power reduction in the grid will exceed demand, potentially leading to new operational risks such as excessive frequency and voltage anomalies. Therefore, once the current total adjustable power reaches the target, the selection process terminates. This allows for precise control of the dispatching intensity, ensuring the grid's operation remains stable within a safe range and preventing excessive dispatching from harming grid security.

[0107] In the implementation of the above scheme, the process of generating corresponding load control instructions is as follows: Calculate the ratio of the power fluctuation amplitude of each power-accessible terminal in the control terminal set to the total current adjustable power, which is used as the power proportion coefficient. Multiply the power to be reduced by each power proportion coefficient to obtain the power reduction value for each power-accessible terminal.

[0108] The power fluctuation ranges of the various callable terminals in the control terminal set vary considerably. Simply distributing the required power reduction evenly is insufficient. Even distribution would result in: callable terminals with small power fluctuations exceeding their regulation capacity, leading to terminal shutdowns or malfunctions, thus impacting grid stability; and callable terminals with large power fluctuations being wasted, causing the total required power reduction value for all callable terminals to fall short of the target, necessitating the use of more callable terminals and increasing control costs and complexity.

[0109] Generate load control instructions containing unique identifiers for each power-accessible terminal and the corresponding power reduction values.

[0110] In actual power grids, the adjustable load resource layer contains a large number of power-accessible terminals, and the parameters of similar power-accessible terminals may be completely identical. Only power-accessible terminals with matching identification information will receive and execute load control commands, thus preventing mis-sending, mis-receiving, and mis-execution from the source and ensuring the accuracy of the controlled objects.

[0111] The load control command is sent to the corresponding power callable terminal for execution, the execution effect is monitored and the operating status of each power callable terminal is updated, including: sending the load control command to each power callable terminal in the control terminal set for execution.

[0112] Record the time from the issuance of the load control command to the first change in power as the actual response time.

[0113] Calculate the deviation rate between the total actual power reduction value and the required power reduction value within the target time period.

[0114] Calculate the deviation between the actual power reduction value and the power reduction value at each sampling time, and perform variance calculation on all deviation values ​​to obtain the power fluctuation variance.

[0115] Based on the deviation rate, actual response time, and power fluctuation variance, update the power fluctuation amplitude, power regulation response time, and regulation stability parameters of the power-adjustable call terminal corresponding to the adjustable load feature set.

[0116] The response speed of a power-restricted terminal may vary due to factors such as equipment aging, communication delays, and adjustments to control strategies. Updating historical values ​​with the actual response time measured in this experiment allows the power regulation response time to increasingly approximate the current true response capability of the power-restricted terminal, thus enabling more accurate predictions during future scheduling. In this invention, a moving average method can be used to update historical values ​​with the actual response time measured in this experiment.

[0117] When a power-capable dispatch terminal consistently exhibits a high deviation rate, it indicates that either it is unable to reliably execute load control commands, or its power fluctuations are artificially inflated. Updating and adjusting the stability parameters can reduce the reliability of this power-capable dispatch terminal in future scheduling, preventing it from further impacting the overall performance.

[0118] Power fluctuation variance measures the volatility and stability of a power-adjustable terminal throughout its operation. A large value indicates poor stability of the power-adjustable terminal, requiring a reduction in adjustment and stabilization parameters. It may also indicate that the terminal may not be able to stably maintain its maximum adjustable power, necessitating a reassessment and potentially a reduction in its power fluctuation amplitude.

[0119] The above formulas are all dimensionless calculations. The formulas are derived from software simulations based on a large amount of collected data to obtain the most recent real-world results. The preset parameters in the formulas are set by those skilled in the art according to the actual situation.

[0120] The above embodiments can be implemented, in whole or in part, by software, hardware, firmware, or any other combination thereof. When implemented using software, the above embodiments can be implemented, in whole or in part, in the form of a computer program product.

[0121] Those skilled in the art will recognize that the modules and algorithm steps of the various examples described in conjunction with the embodiments disclosed herein can be implemented in electronic hardware, or a combination of computer software and electronic hardware. Whether these functions are implemented in hardware or software depends on the specific application and design constraints of the technical solution. Those skilled in the art can use different methods to implement the described functions for each specific application, but such implementations should not be considered beyond the scope of this invention.

[0122] In addition, the functional modules in the various embodiments of the present invention can be integrated into one processing module, or each module can exist physically separately, or two or more modules can be integrated into one module.

[0123] The above description is merely a specific embodiment of the present invention, but the scope of protection of the present invention is not limited thereto. Any variations or substitutions that can be easily conceived by those skilled in the art within the technical scope disclosed in the present invention should be included within the scope of protection of the present invention. Therefore, the scope of protection of the present invention should be determined by the scope of the claims.

[0124] Finally, the above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the protection scope of the present invention.

Claims

1. A method for hierarchical access of adjustable loads in a virtual power plant based on dynamic priority scheduling, characterized in that, Includes the following steps: Obtain load data for all power terminals within the virtual power plant area; determine the historical power change regulation characteristics of each power terminal based on the load data, wherein the historical power change regulation characteristics include power fluctuation amplitude, power regulation response time, and regulation stability parameters used to characterize power deviation; The power fluctuation amplitude, power regulation response time, and regulation stability parameters of all power terminals are sorted from smallest to largest, and the corresponding median values ​​are determined. Based on each power terminal: the power fluctuation amplitude is compared with the corresponding median value to obtain the fluctuation amplitude adaptation coefficient; the median value of the power regulation response time is compared with the power regulation response time to obtain the response time adaptation coefficient; the median value of the regulation stability parameter is compared with the regulation stability parameter to obtain the stability adaptation coefficient. Calculate the weighted average of the fluctuation amplitude adaptation coefficient, response time adaptation coefficient, and stability adaptation coefficient to obtain the comprehensive relative adaptation degree; The overall relative suitability of all power terminals is sorted from high to low. Power terminals are selected sequentially starting from the beginning of the sorting sequence, and the power fluctuation amplitude is accumulated simultaneously. When the cumulative value of power fluctuation amplitude is greater than or equal to the amount of power that needs to be reduced for the first time, the selection will be terminated and the selected power terminal will be used as the power available power terminal. Based on the historical power change adjustment characteristics of the power-accessible power terminals, an adjustable load feature set is constructed; cluster analysis is performed on the adjustable load feature set to divide the power-accessible power terminals into multiple adjustable load resource layers; Receive scheduling instructions containing the amount of power consumption to be reduced within the target time period, and construct scheduling priorities for each adjustable load resource layer based on the scheduling instructions and adjustable load feature sets; Based on scheduling priority, all power-accessible terminals within each adjustable load resource layer are sorted and selected from high to low until the total adjustable power meets the scheduling instructions, and corresponding load control instructions are generated. The load control command is sent to the corresponding power-accessible terminal for execution, the execution effect is monitored, and the operating status of each power-accessible terminal is updated.

2. The method for hierarchical access of adjustable loads in a virtual power plant based on dynamic priority scheduling according to claim 1, characterized in that, The process of obtaining the power fluctuation amplitude, power regulation response time, and regulation stability parameters is as follows: The load data is divided into preset time windows, and the difference between the maximum and minimum power values ​​within each time window is calculated. The arithmetic mean of all differences is taken as the power fluctuation amplitude. Calculate the time difference between the moment the load control command is issued and the moment the power first changes, and take the arithmetic mean of all time differences as the power regulation response time; The period from the start to the end of the load control command is taken as the target time period. The deviation ratio between the required power reduction and the actual power reduction within the target time period is calculated, and the arithmetic mean of all deviation ratios is taken as the regulation and stability parameter.

3. The method for hierarchical access of adjustable loads in a virtual power plant based on dynamic priority scheduling according to claim 1, characterized in that, The process of constructing an adjustable load feature set is as follows: Assign a unique identifier to each power-adjustable terminal; associate its unique identifier with the corresponding power fluctuation amplitude, power regulation response time, regulation stability parameters and comprehensive relative adaptability to form a feature vector; and collect all feature vectors to construct an adjustable load feature set.

4. The method for hierarchical access of adjustable loads in a virtual power plant based on dynamic priority scheduling according to claim 1, characterized in that, The process of dividing a power-adjustable power terminal into multiple adjustable load resource layers is as follows: Standardize the power fluctuation amplitude, power regulation response time, regulation stability parameters and overall relative fit in each feature vector; Iteratively execute the following steps until all power-enabled electrical terminals have completed clustering: Calculate the distance between the feature vector of the current power-enabled terminal and the center feature vector of all existing clusters; extract the minimum distance and its corresponding cluster; If the minimum distance is less than the preset distance threshold, the power-enabled mobile terminal will be assigned to the corresponding cluster, and the center feature vector of the cluster will be recalculated. If the minimum distance value is greater than or equal to the preset distance threshold, a new cluster is created with the power-enabled mobile terminal as the core. Calculate the average values ​​of all power fluctuation amplitudes, power regulation response times, and regulation stability parameters within each cluster, and use them as the regulation capability characterization value, response speed characterization value, and stability characterization value for each cluster. Clusters whose regulation capacity, response speed, and stability values ​​fall within a certain range of their respective values ​​are merged into the same adjustable load resource layer, and each adjustable load resource layer is statistically obtained.

5. The method for hierarchical access of adjustable loads in a virtual power plant based on dynamic priority scheduling according to claim 4, characterized in that, The process of constructing the scheduling priority for each adjustable load resource layer is as follows: Extract the duration of the target time period and the amount of power consumption to be reduced from the dispatch instructions; Based on each adjustable load resource layer: calculate the reciprocal of the ratio of the response speed characterization value to the time length, as the response time matching degree; The ratio of the regulation capacity characterization value to the amount of power to be reduced is used as the regulation capacity matching degree; the reciprocal of the stability characterization value is used as the regulation accuracy matching degree. The comprehensive priority score is obtained by integrating the matching degree of response timeliness, the matching degree of adjustment capability, and the matching degree of adjustment accuracy. All adjustable load resource layers are sorted from high to low according to their scheduling priority scores to form scheduling priorities.

6. The method for hierarchical access of adjustable loads in a virtual power plant based on dynamic priority scheduling according to claim 1, characterized in that, Based on scheduling priority, all power-available call terminals within each adjustable load resource layer are sorted, including: Calculate the maximum and minimum values ​​of power fluctuation amplitude, power regulation response time, regulation stability parameters, and overall relative adaptability of all power-accessible terminals within each adjustable load resource layer; Based on the maximum and minimum values, the feature vectors of each power-adjustable call terminal in the current adjustable load resource layer are standardized. The ratio of the standardized power regulation response time to the duration is subtracted from the setpoint to obtain the individual response time matching degree; the ratio of the standardized power fluctuation amplitude to the amount of power to be reduced is obtained as the individual regulation capability matching degree; and the reciprocal of the standardized regulation stability parameter is obtained as the individual regulation accuracy matching degree. The individual response time matching degree, individual regulation capability matching degree, individual regulation accuracy matching degree and standardized comprehensive relative fit degree of each power-capable mobile terminal are weighted and integrated to obtain the comprehensive ranking score; All power-accessible terminals within each adjustable load resource layer are sorted from highest to lowest based on their comprehensive ranking scores. When the comprehensive ranking scores are the same, they are arranged from largest to smallest based on their power fluctuation amplitude. A priority ranking sequence of all power-accessible terminals within each adjustable load resource layer is then generated.

7. The method for hierarchical access of adjustable loads in a virtual power plant based on dynamic priority scheduling according to claim 1, characterized in that, The process of selecting from high to low power until the total adjustable power satisfies the scheduling command includes: Select the adjustable load resource layer with the highest priority, and then select each power adjustable terminal in turn. The power fluctuation amplitude of the selected power adjustable terminal is accumulated synchronously and used as the current adjustable power sum. If the current total adjustable power does not reach the amount of power that needs to be reduced, and all power-accessible terminals in the current adjustable load resource layer have been selected, then the next adjustable load resource layer is selected from high to low according to the scheduling priority, and the selection and accumulation are repeated. If the total adjustable power is greater than or equal to the amount of power that needs to be reduced, the selection process will terminate, and the selected power adjustable terminals will be recorded to form a set of control terminals.

8. The method for hierarchical access of adjustable loads in a virtual power plant based on dynamic priority scheduling according to claim 1, characterized in that, The process of generating the corresponding load control instructions is as follows: The ratio of the power fluctuation amplitude of each power adjustable terminal in the control terminal set to the total current adjustable power is calculated and used as the power proportion coefficient. The required power reduction is calculated by multiplying the power consumption reduction amount by the power ratio coefficient of each power consumption terminal. Generate load control instructions containing unique identifiers for each power-accessible terminal and the corresponding power reduction values.

9. The method for hierarchical access of adjustable loads in a virtual power plant based on dynamic priority scheduling according to claim 1, characterized in that, The load control command is issued to the corresponding power-accessible terminals for execution, the execution effect is monitored, and the operating status of each power-accessible terminal is updated, including: The load control command is sent to each power-capable electric terminal in the control terminal set for execution. Record the time from the issuance of the load control command to the first change in power as the actual response time; Calculate the deviation rate between the sum of actual power reduction values ​​and the required power reduction values ​​within the target time period; Calculate the deviation between the actual power reduction value and the power reduction value at each sampling time, and perform variance calculation on all deviation values ​​to obtain the power fluctuation variance; Based on the deviation rate, actual response time, and power fluctuation variance, update the power fluctuation amplitude, power regulation response time, and regulation stability parameters of the power-adjustable call terminal corresponding to the adjustable load feature set.

Images