Regional power dispatching optimization method and system driven by dynamic power consumption data

Abstract

The invention discloses a regional power dispatching optimization method and system driven by dynamic power consumption data, and relates to the technical field of power dispatching. The method comprises the following steps: collecting historical power consumption data of users, extracting dynamic characteristics and carrying out suitability analysis among the users to form a plurality of user pairs; configuring transient energy storage blocks controlled by an energy storage center in a distributed manner for each user pair; receiving a scheduling request, and identifying a requesting user side and a corresponding block thereof; and monitoring the power energy storage state data set of each energy storage block, and controlling and requesting the transient energy storage block corresponding to the user side to carry out power dispatching on the transient energy storage block according to the power energy storage state data set. The technical problems that scheduling response is lagged, local load fluctuation is large and the energy storage utilization rate is low due to the fact that scheduling resources are difficult to match dynamically according to user-side power consumption behaviors in traditional power scheduling are solved, and accurate user adaptation and distributed transient energy storage cooperative scheduling is achieved based on user-side dynamic power characteristics. And the real-time performance, the stability and the energy storage utilization efficiency of power dispatching are improved.

Description

Technical Field

[0001] This invention relates to the field of power dispatching technology, specifically to a regional power dispatching optimization method and system driven by dynamic power consumption data. Background Technology

[0002] With the rapid penetration of distributed renewable energy (photovoltaics, wind power), electric vehicles, and various flexible loads into distribution networks, the load patterns of regional power grids are exhibiting greater randomness and time-varying characteristics. Traditional dispatching models centered on the generation and grid sides mainly rely on day-ahead planning and static load forecasting, which are unable to reflect changes in fine-grained electricity consumption behavior on the user side in a timely manner, leading to increased operational risks such as widening peak-valley differences, local feeder overruns, and voltage fluctuations.

[0003] Existing demand response and energy storage participation scheduling methods are mostly based on group average curves or rule triggering, focusing on "single-point local" control or "unified downlink" command allocation. On the one hand, they ignore the dynamic characteristics of user-side loads such as ramp-up, sudden changes and peak-hour distribution on a short time scale. On the other hand, energy storage is mostly controlled by independent units, lacking cross-user and cross-node coordination and matching mechanisms, making it difficult to achieve flexible compensation and transient support for local unbalanced loads at the second / minute level.

[0004] In terms of user-side data utilization, mainstream practices tend to focus on daily / weekly statistics and medium- to long-term trends, lacking methods for feature extraction, correlation measurement, and suitability assessment of second- to minute-level data, making it difficult to identify complementary relationships between users. This leads to regional-level energy storage configurations generally relying on experience or coarse-grained indicators, resulting in capacity redundancy or under-configuration, affecting scheduling efficiency and effectiveness. Summary of the Invention

[0005] This application provides a regional power dispatch optimization method and system driven by dynamic power consumption data, which solves the technical problems of traditional power dispatching, which is difficult to dynamically match dispatch resources according to user-side electricity consumption behavior, resulting in delayed dispatch response, large local load fluctuations and low energy storage utilization.

[0006] A first aspect of this application provides a regional power dispatch optimization method driven by dynamic power consumption data, the method comprising: Historical electricity consumption datasets from the user side are collected, and the user-side power dynamic characteristics of the historical electricity consumption datasets are extracted. Based on these user-side power dynamic characteristics, pairwise adaptability analysis is performed on each user side to obtain multiple user-side pairs. Multiple transient power storage blocks are configured according to these multiple user-side pairs, and these multiple transient power storage blocks are distributedly controlled by a power storage center. Power dispatch requests are received, and the requesting user side and its corresponding transient power storage block are identified. The power storage status monitoring dataset of the multiple transient power storage blocks is obtained, and the transient power storage block corresponding to the requesting user side is controlled to perform power dispatch to the requesting user side according to the power storage status monitoring dataset.

[0007] A second aspect of this application provides a regional power dispatch optimization system driven by dynamic power consumption data, the system comprising: The module includes the following components: Adaptability Analysis Module: Collects historical electricity consumption data from the user side, extracts the user-side power dynamic characteristics from the historical electricity consumption data, and performs pairwise adaptability analysis on each user side based on these characteristics to obtain multiple user-side pairs; User-side Configuration Module: Configures multiple transient power storage blocks corresponding to the multiple user-side pairs, with the multiple transient power storage blocks being distributed and controlled by the power storage center; Dispatch Request Identification Module: Receives power dispatch requests and identifies the requesting user side and the corresponding transient power storage block; Power Dispatch Module: Monitors the power storage status monitoring data of the multiple transient power storage blocks and controls the transient power storage block corresponding to the requesting user side to perform power dispatch to the requesting user side according to the power storage status monitoring data.

[0008] One or more technical solutions provided in this application have at least the following technical effects or advantages: First, historical electricity consumption data is collected from each user, and dynamic characteristics reflecting short-term fluctuations are extracted. Then, based on these characteristics, pairwise matching analysis is performed on all user sides to select user combinations with good matching in terms of load complementarity, etc. Corresponding transient energy storage blocks are then configured for each combination, and distributed unified management is achieved by the energy storage center. Subsequently, when a user issues a power dispatch request, the system identifies the user and its corresponding energy storage block, monitors the state of charge and availability of all energy storage blocks in real time, and, based on the monitoring data, dispatches and controls the energy storage block corresponding to the requesting user to output an appropriate amount of power to meet demand and maintain the safety and coordination of overall dispatch. Attached Figure Description

[0009] To more clearly illustrate the technical solutions in the embodiments of the present invention, the accompanying drawings used in the description of the embodiments will be briefly introduced below. Obviously, the accompanying 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.

[0010] Figure 1 A schematic diagram of the process for a regional power dispatch optimization method driven by dynamic power consumption data, provided in an embodiment of this application.

[0011] Figure 2 A schematic diagram of the structure of a regional power dispatch optimization system driven by dynamic power consumption data provided in this application embodiment.

[0012] Figure labeling: Adaptability analysis module 11, user-side configuration module 12, dispatch request identification module 13, power dispatch module 14. Detailed Implementation

[0013] To further illustrate the technical means and effects of the present invention in achieving its intended purpose, the following detailed description of the specific implementation methods, structure, features and effects of the present invention, in conjunction with the accompanying drawings and preferred embodiments, is provided below.

[0014] Example 1, as Figure 1 As shown, this application provides a regional power dispatch optimization method driven by dynamic power consumption data, the method including: Historical electricity consumption datasets from the user side are collected, and the user-side dynamic electricity features of the historical electricity consumption datasets are extracted. Based on the user-side dynamic electricity features, pairwise adaptability analysis is performed on the user sides to obtain multiple user-side pairs.

[0015] In this embodiment, historical power consumption datasets are first collected from each user side within the target area. These datasets include active power, reactive power, voltage, current, and related load measurements recorded according to a preset sampling period. Subsequently, based on the collected time-series data, power dynamic characteristics of each user side are extracted using methods such as moving window calculation, differential analysis, and correlation analysis. These dynamic characteristics characterize the load variation trend, fluctuation intensity, and periodicity of the user side over a short time scale. For example, the instantaneous power change rate is calculated within a fixed-length sliding time window to obtain the short-term ramp rate, and the average short-term slope is obtained from the average power increment within the window. After acquiring the dynamic characteristics of all user sides, a compatibility analysis is performed on any two user sides. Specifically, the system calculates compatibility indicators such as load complementarity index, peak-hour shifting index, instantaneous demand complementarity index, and correlation confidence level based on the power dynamic characteristics of the two user sides. These indicators quantify the degree of matching between the two user sides in terms of load change patterns, peak-valley occurrence times, and fluctuation modes. Subsequently, based on the calculated adaptability index, a greedy adaptation algorithm is used to analyze all user sides, select user side combinations that meet the threshold conditions, and finally form multiple user side pairs for subsequent energy storage block configuration and scheduling matching.

[0016] Furthermore, based on the dynamic characteristics of the user-side power, pairwise adaptability analysis is performed on each user-side to obtain multiple user-side pairs. The method includes: The user-side power dynamic characteristics include short-term ramp rate, average short-term slope, autocorrelation coefficient, peak-hour location distribution vector, and abrupt change frequency based on a preset moving window. Adaptability analysis is performed on each user side according to these user-side power dynamic characteristics to obtain a set of adaptability indicators, including load complementarity, peak-hour shifting degree, instantaneous demand complementarity, and correlation confidence. An adaptation group is initialized, and a greedy adaptation algorithm is used to analyze all user sides within the initialized adaptation group based on the set of adaptability indicators, resulting in multiple user-side pairs.

[0017] Preferably, the system first extracts short-term dynamic features from the load power sequence of each user based on the collected historical electricity consumption data of the user side, according to a preset moving time window. Within each window, the instantaneous change in the power sequence and the average power increment within the window are calculated to obtain the short-term ramp rate and average short-term slope; the autocorrelation coefficient is calculated using the autocorrelation function of the time series within the window to reflect the periodicity and stability of load changes; the time of power peak occurrence is statistically analyzed and encoded into a peak-time distribution vector to describe the intraday location characteristics of user load peaks; power abrupt events are identified and their frequencies are statistically analyzed by detecting abrupt change points within the window to obtain abrupt change frequency characteristic. After calculating these features, the system summarizes them to form user-side power dynamic features, which are used to characterize the short-term electricity consumption behavior of each user.

[0018] Subsequently, based on these user-side power dynamic characteristics, adaptability analysis was performed on all user sides in pairs. Specifically, the system uses a negative Pearson correlation coefficient as its load complementarity index based on load data from two user sides. This index reflects the difference in the direction and magnitude of power changes between the two users at the same time. Generally, if the load changes of the two users have similar trends (i.e., both are high or low simultaneously), the correlation coefficient is positive, and the complementarity score becomes negative, indicating poor complementarity. If the trends are opposite, the correlation coefficient is negative, and the complementarity score is positive, indicating good complementarity. By comparing the peak-hour distribution vectors of the two user sides, the system identifies the top N time periods with the highest probability for each user as typical peak-hour sets. For example, the system selects the first four hours with a probability greater than a threshold and calculates the Jaccard similarity coefficient between the two typical peak-hour sets (i.e., the intersection size divided by the union size). Subtracting this Jaccard similarity coefficient from 1 yields the peak-hour shifting degree, which measures whether there is a usable peak-valley complementarity relationship between the two users. Generally, when the peak hours of the two users do not overlap at all, the Jaccard coefficient is 0, and the shifting degree is 1; when they completely overlap, the Jaccard coefficient is 1, and the shifting degree is 0. Statistical analysis is then performed. The set of time points where the short-term ramp rate of user-side A exceeds a positive threshold is used. For each time point in the set, user-side B is checked, and the number of times user-side B is in a complementary state at that time point is counted. For example, if the short-term ramp rate of user-side B is negative at a certain time point, it is counted as one effective complementarity. If the load is lower than the preset proportion of its daily average, it is counted as one effective complementarity. The number of times it is in a complementary state is divided by the number of elements in the time point set to calculate the instantaneous demand complementarity between user-side A and user-side B. Similarly, the instantaneous demand complementarity between user-side B and user-side A is calculated. By averaging the two instantaneous demand complementarities, the required instantaneous demand complementarity between user-side A and user-side B is obtained, which is used to evaluate the degree of complementarity between the two during short-term fluctuations. When calculating the correlation coefficient of load complementarity, a significance test, such as a t-test, is also performed simultaneously to obtain a p-value. Then, 1 is subtracted from this p-value to obtain the correlation confidence score. Generally, the smaller the p-value, the more significant the correlation, and the closer the confidence score is to 1. By summarizing the above-calculated indicators, a set of adaptability indicators is formed, which is used to characterize the degree of matching between any two user sides.

[0019] Next, the adaptation team is initialized, and all user sides are included in the initial set of users to be adapted. The system analyzes the user sides in the initial set of users to be adapted using a greedy adaptation algorithm based on a set of adaptability indicators. Specifically, it calculates the average adaptability indicator between each user side and other user sides, sorts all user sides from highest to lowest according to the average indicator, selects the top-ranked user sides as seed user sides, and then selects a user side from the seed user side set. It iterates through the candidate user side pairs with the highest scores in the adaptability indicator set, and sequentially selects the user side with the highest adaptability score from the candidate pairs, forming a user side pair. After completing one pairing, the paired user sides are removed from the set of users to be adapted, and the above greedy search process is repeated for the remaining user sides. After traversing all seed user sides, the system ultimately generates multiple user side pairs for subsequent energy storage block configuration and power dispatch optimization.

[0020] Furthermore, based on the set of adaptability indicators, a greedy adaptation algorithm is performed on all user sides in the initialization adaptation group to obtain multiple user side pairs. The method includes: Based on the set of adaptability metrics, calculate the average adaptability metric for each user side and other user sides; sort all user sides in descending order based on the average adaptability metric to obtain a set of seed user sides and a set of non-seed user sides; select a first user side from the set of seed user sides, and perform greedy adaptability metric optimization based on the set of non-seed user sides until the set of seed user sides is traversed to obtain the user sides that are suitable for the set of seed user sides.

[0021] Optionally, after obtaining the set of compatibility metrics for all user sides, the average compatibility metric for each user side with all other user sides is first calculated. Specifically, for any user, all corresponding compatibility metrics are iterated, including load complementarity, peak-hour shifting degree, instantaneous demand complementarity, and correlation confidence. The metric values ​​obtained by combining this user with all other user sides are then weighted and summed to obtain the overall average compatibility metric for that user side, used to quantify its comprehensive compatibility potential globally. After obtaining the average compatibility metric for all user sides, all user sides are sorted in descending order based on this metric. After sorting, a subset of user sides with high overall compatibility potential are grouped into a seed user side set (those exceeding a preset threshold), and the remaining user sides are grouped into a non-seed user side set. The seed user side set is used to initiate user side pairing, while the non-seed user side set serves as candidate objects, providing priority matching partners for the seed user sides. Subsequently, the first user side selected from the seed user side set in order of sorting is chosen as the primary user side for the current pairing. Starting with the primary user side, the system traverses the set of non-seed user sides and calculates the current optimal pairing based on the compatibility index between the primary user side and each candidate user side in the compatibility index set. During the traversal, the system employs a greedy strategy, selecting the candidate user side with the highest compatibility index score with the primary user side at each step and forming a user side pair. After pairing, both paired user sides are simultaneously removed from their respective sets to prevent reuse. The system then continues to select the next user side from the seed user side set as the new primary user side and repeats the greedy optimization process. After traversing and completing the pairing of all seed user sides, a set of user side pairs based on the optimal compatibility principle is finally obtained, laying the foundation for subsequent transient power storage block configuration.

[0022] Furthermore, after obtaining the suitable user sides from the seed user side set, the adapted user sides are removed from the non-seed user side set to obtain the remaining non-seed user side set; it is determined whether the number of user sides in the remaining non-seed user side set is greater than or equal to 2; if the number of user sides in the remaining non-seed user side set is greater than or equal to 2, a greedy adaptation index is used to optimize the remaining non-seed user side set until the remaining non-seed user side set is traversed to obtain the suitable user sides in the remaining non-seed user side set.

[0023] Optionally, after adapting the seed user set, the system obtains a subset of user pairs consisting of the seed user and its optimal matching user. Subsequently, the system updates the non-seed user set. Specifically, it first identifies all user pairs selected as matching objects in the aforementioned pairing process within the non-seed user set and removes these adapted user pairs from the non-seed user set, thus obtaining the remaining non-seed user set. This ensures that subsequent matching only targets user pairs that have not yet participated in pairing, avoiding the reuse of the same user pair. Next, it checks if the number of user pairs in the remaining non-seed user set is greater than or equal to 2. If the number of user pairs in this set is less than 2, it indicates that the remaining users cannot form valid user pairs, and the adaptation process for the remaining set ends in this case. If the number of user pairs in the remaining non-seed user set is greater than or equal to 2, a new round of greedy optimization based on the adaptability index is initiated. When adapting the remaining non-seed user side set, a greedy strategy similar to that used in the seed set pairing process is adopted. Specifically, the system iterates through each user side in the remaining non-seed user side set, treating it as the current primary user side, and calculates its compatibility score with other candidate user sides within the set based on a set of compatibility metrics. The system selects the user side with the highest compatibility score from the candidate objects as the pairing object with the primary user side, generating the corresponding user side pair. After pairing, the two user sides that participated in the pairing are removed from the remaining non-seed user side set to ensure that these user sides are not reused in subsequent pairings. The system continues to execute the greedy adaptation process according to the above steps, iterating and pairing the remaining non-seed user side set one by one until all user sides in the set have been matched or the number of users in the set is insufficient to form a new pairing. Ultimately, multiple user side pairs are obtained from the remaining non-seed user side set, thus completing the adaptation structure of the entire user side set and providing a complete data foundation for subsequent energy storage block configuration and scheduling optimization.

[0024] Furthermore, the method for optimizing the fit index of the remaining non-seed user set includes: Obtain the initial adapted user side for each user side in the remaining non-seed user side set; identify the adapted user side including each user side in the remaining non-seed user side set; readjust the pairing relationship by comparing the adaptability index of the initial adapted user side and the adapted user side, and output the updated user side pair.

[0025] Optionally, after completing the partial pairing in the previous stage, the remaining non-seed user side set is used as a new set to be adapted. To improve the accuracy and stability of subsequent pairing, the system first obtains the initial adapted user side for each user side in this set. That is, based on the set of adaptability indicators, the system obtains the adaptability score of each user side with all other user sides in the set to be adapted, and selects the user side with the highest score as the initial adapted object for that user side, forming an initial adaptation relationship. After completing the construction of the initial adaptation relationship, the system identifies the user sides in the remaining non-seed user side set that have entered the pairing process, forming an adapted user side set. This adapted user side set is used to identify the user sides that have participated in pairing in previous steps. Subsequently, the current pairing relationship is evaluated and corrected by comparing the adaptability indicators corresponding to the initial adapted user sides and the adapted user sides. Specifically, when an initial matching target for a user side already belongs to the set of already matched user sides, or when the compatibility index of the current initial pairing combination is lower than a preset threshold, a pairing adjustment mechanism will be activated. This involves calculating the compatibility index between the current user side and the remaining candidate user sides, re-sorting them according to their compatibility scores from highest to lowest, and selecting the candidate user side with the highest compatibility as the new matching target. Simultaneously, if a new matching target conflicts with other user sides (i.e., multiple user sides compete for the same optimal target), the compatibility indices of related pairings will be compared. The pairing relationship will be retained for the combination with higher compatibility, while a second-best option that still meets the threshold condition will be reselected for the other user side. Through this comparison and dynamic adjustment process, conflicting relationships, duplicated targets, and inefficient pairing combinations in the initial pairings are gradually eliminated, ensuring that each user side obtains a counterpart user side that best matches its dynamic characteristics. Finally, after completing all conflict handling and pairing optimization, an updated set of user side pairs is output, providing a stable, reliable, and optimized user adaptation structure for subsequent energy storage block configurations.

[0026] Furthermore, based on the set of adaptability indicators, a greedy adaptation algorithm is performed on all user sides in the initialization adaptation group to obtain multiple user side pairs. The method includes: Randomly select two pairs from the multiple user-side pairs and calculate the initial overall fitness index; swap the two randomly selected pairs and recalculate the overall fitness index. If the recalculated overall fitness index is greater than the initial overall fitness index, readjust the pairing relationship until the updated user-side pairs are output after K rounds of iteration.

[0027] Optionally, after completing the greedy matching, to further improve the overall pairing quality, a random swap mechanism is introduced to iteratively optimize the user-side pairs. Specifically, firstly, two pairs are randomly selected from the multiple user-side pairs already obtained. The system calculates the suitability scores of these two pairs based on the suitability index set, and sums them to obtain an initial total suitability index, which is used to measure the overall matching quality of the current pairing structure. Subsequently, the two pairs are swapped, and the suitability scores of the two new pairs after the swap are recalculated based on the suitability index set, and summed to obtain the total suitability index after the swap. Then, the total suitability index before and after the swap is compared. If the total suitability index after the swap is greater than the initial total suitability index, it indicates that the swap of the two pairs can improve the overall matching degree. At this time, the swapped pairing replaces the original pairing, thus completing one structural optimization. If the total suitability index after the swap does not improve, the swap is abandoned, and the original pairing structure remains unchanged. The system repeats the above iterative optimization process of random selection—swap attempt—suitability comparison, with the number of iterations being a preset value K. In each iteration, different pairings are randomly selected to prevent getting trapped in local optima while ensuring that more potential pairing combinations are explored. After all K iterations are completed, the updated user-side pairing results are output. These results outperform the pairing structure generated by the initial greedy algorithm in terms of overall suitability, providing a better user combination basis for subsequent energy storage block configurations.

[0028] Multiple transient energy storage blocks are configured according to the multiple user-side configurations, and the multiple transient energy storage blocks are distributedly controlled by the energy storage center.

[0029] In one embodiment, after completing the adaptation analysis for multiple user pairs, a load superposition curve is constructed for each user pair based on its corresponding historical load curve. This analyzes the power variation patterns of the combined load during typical periods and identifies power gap indicators on the superposition curve. These indicators are used to determine key configuration parameters such as required energy storage capacity, discharge power level, and response time. This allows for the allocation of corresponding transient energy storage blocks to each user pair under the unified management of the power energy storage center. Each transient energy storage block consists of a set of controllable energy storage units, possessing independent state-of-load monitoring, power regulation, and communication control interfaces. Based on the load complementarity characteristics of the user pairs and historical operating data, the system sets different operating modes for different blocks, thereby achieving dynamic balancing of local power fluctuations and ensuring the continuity and stability of power dispatch throughout the region.

[0030] Furthermore, based on the configuration of multiple transient energy storage blocks corresponding to the multiple user-side configurations, the method includes: A load superposition curve is constructed based on the historical load curves of each user side among the multiple user sides. A power shortage index is identified based on the load superposition curve. The configuration parameters of the transient power storage block are obtained using the power shortage index as a constraint. The configuration parameters include energy storage capacity, maximum discharge power, and safety margin.

[0031] Preferably, after obtaining multiple user pairs, the system first acquires the historical load curves for each user pair. These historical load curves are sequential data of active power changes over time, recorded based on a preset sampling period. The system performs time alignment processing on the load curves of two users within the same user pair and constructs a load superposition curve for that user pair by adding them hourly. These load superposition curves reflect the overall load change trend of the two users under combined operation, and are used to identify the time periods and magnitudes that may generate power shortages. After obtaining the load superposition curves, the system searches for high-load periods on the load superposition curves and calculates the peak power shortage based on the difference between the maximum peak load and the regional power grid or local power supply capacity. Then, by calculating the local slope and instantaneous ramp rate of the load superposition curves, it identifies high-fluctuation periods and assesses their corresponding short-term power shortage demand. In addition, by statistically analyzing the duration of high-load intervals, it obtains a load imbalance duration index, which is used to assess the time scale requiring energy storage support. By summarizing these indicators, the system generates a set of power shortage indicators, including peak shortage power, peak shortage duration, instantaneous demand shortage, and fluctuation intensity indicators. Subsequently, using the power shortage index as a configuration constraint, the configuration parameters of the corresponding transient power storage block are determined. Specifically, the system calculates the maximum discharge power of the storage block based on the peak shortage power to ensure it can meet the instantaneous support requirements of the maximum generation shortage. Then, it calculates the required energy storage capacity based on the product of the load imbalance duration and the shortage power, ensuring the storage block can stably output power during the shortage period. Furthermore, a safety margin is set based on historical load volatility and regional operational stability requirements. This safety margin is used to reserve an additional redundancy ratio for energy storage capacity and discharge power to cope with factors such as sudden load surges or energy storage degradation. Finally, the system outputs a complete set of configuration parameters, including energy storage capacity, maximum discharge power, and safety margin, as the basic configuration scheme for the transient power storage block, used for subsequent scheduling and distributed control execution.

[0032] Receive power dispatch requests and identify the requesting user side and the corresponding transient power storage block.

[0033] In one embodiment, the system continuously monitors power dispatch requests from the power dispatch center or the user side during operation. These requests can be issued by the regional power grid dispatch platform, user-side smart terminals, or distributed energy management units, and include information such as the requesting user's identification information, target power demand, power dispatch period, priority, and real-time operating status. Upon receiving a power dispatch request, the system first parses the dispatch instruction content through the communication interface module and converts it into a structured dispatch request data packet. Then, based on the requesting user's identifier, it retrieves the corresponding user-side information from the user characteristic database, including the user's historical load curve, power dynamic feature vector, and the user-side pair information to which it belongs. The system quickly locates the transient power storage block number and its control unit address corresponding to the requesting user by querying the pre-established user-side pair-energy storage block mapping table. If the user belongs to multiple user-side pairs or has a cross-block dispatch relationship, the system selects the optimal transient power storage block as the primary dispatch unit based on the current power demand, block state of charge, and allocation priority, and optionally designates one or more backup blocks to form a redundancy support mechanism. After identifying the corresponding energy storage block, the system further verifies its current availability, including real-time state of charge (SOC), dischargeable power, temperature decay factor, and operating mode (grid-connected / off-grid). If the energy storage block's status meets the scheduling execution conditions, the system marks the block as scheduling-ready and sends a response confirmation message to the power energy storage center to complete the scheduling preparation phase. If insufficient block capacity or maintenance status is detected, the system automatically switches to the backup energy storage block associated with the requesting user, ensuring the continuity and stability of the scheduling process. Through the above identification and matching mechanism, the system can quickly determine the target user and its corresponding energy storage support unit after receiving a power dispatch request, realizing automatic mapping and coordinated management between dispatch requests and distributed energy storage resources, providing accurate and real-time target objects for subsequent power output control.

[0034] The power storage status monitoring dataset of the multiple transient power storage blocks is obtained, and the power storage status monitoring dataset is used to control the transient power storage block corresponding to the requesting user to perform power dispatch to the requesting user.

[0035] In one embodiment, during power dispatching, multiple transient energy storage blocks are continuously monitored in real time to construct an energy storage status monitoring dataset. This dataset is periodically generated by the energy storage center through a distributed communication network and includes parameters such as the real-time state of charge, temperature decay factor, capacity decay factor, available capacity, and maximum discharge power of each energy storage block. The system dynamically updates and archives the collected data to ensure that the operating status information of each energy storage block is up-to-date during dispatching decisions. When a power dispatching request is detected, the system first extracts the operating status of the energy storage block corresponding to the requesting user from the energy storage status monitoring dataset and determines whether the block has immediate discharge capability based on its state of charge and available capacity. If the energy storage block is in a dischargeable range and its power output capability meets the dispatching request, the system calculates the corresponding discharge control command based on the power dispatching objective, including the discharge power setpoint, output duration, and control mode. Then, the discharge control command is sent to the local control unit of the target energy storage block via a distributed control bus. Upon receiving a command, the energy storage block activates its power conversion devices, such as DC / AC converters or inverter modules, to supply power to the requesting user according to the set discharge power curve. Simultaneously, the system performs closed-loop monitoring of the energy storage block's output status through real-time feedback signals of current, voltage, and SOC, ensuring that the output power and duration meet preset control targets. When the energy storage block's state of charge approaches a safe threshold or its temperature parameters exceed a set range, the system automatically reduces its output power or switches to a backup energy storage block to maintain the safety and stability of the dispatch process. Throughout the dispatch execution process, the power storage center continuously and dynamically corrects and records the monitoring dataset, creating a traceable operation log to provide a basis for subsequent energy management optimization and energy storage parameter adjustments. Through this monitoring and control process, precise power control and adaptive dispatch of transient power storage blocks are achieved, enabling the energy storage system to quickly respond to fluctuations in user-side power demand and ensuring local power supply-demand balance and operational stability of the power grid.

[0036] Furthermore, the method of controlling the transient energy storage block corresponding to the requesting user side to perform power dispatch to the requesting user side according to the energy storage status monitoring dataset includes: The power storage status monitoring dataset includes real-time state of charge, capacity decay factor, and temperature decay factor; the power dispatch target of the requesting user is read, and the real-time state of charge, capacity decay factor, and temperature decay factor are analyzed based on the power dispatch target to obtain the real-time power dispatch data output by the transient power storage block to the requesting user.

[0037] Preferably, upon receiving a power dispatch request, the system retrieves real-time state parameters related to the target transient power storage block from the power storage state monitoring dataset. These parameters include real-time state of charge, capacity decay factor, and temperature decay factor. The real-time state of charge reflects the current available energy ratio of the storage block; the capacity decay factor characterizes the effective capacity decay of the storage unit during long-term operation; and the temperature decay factor describes the degree of reduction in the output capacity of the storage unit under high or low temperature conditions. Subsequently, the system reads the power dispatch target from the requesting user, which may include information such as target discharge power, dispatch duration, and voltage or frequency support requirements. The system comprehensively analyzes the dispatch target and state parameters; that is, it packages these state parameters and the dispatch target as input data and transmits them to the discharge analysis model. This discharge analysis model employs a multilayer perceptron structure and has undergone iterative training using historical state data and historical discharge data, including forward propagation, loss calculation (e.g., mean square error), backpropagation, and parameter optimization (e.g., Adam optimizer). Based on the received dispatch target and state parameters, it can generate discharge control commands suitable for the current situation. Subsequently, the obtained discharge control command is sent as the final real-time power dispatch data to the corresponding transient power storage block, and its local control unit performs the power output operation to realize dynamic power dispatch to the requesting user side.

[0038] Furthermore, after obtaining the real-time power dispatch data output by the transient power storage block to the requesting user, the method further includes: Record characteristic abnormal events, wherein the characteristic abnormal event is an event in which the real-time power dispatch data exceeds the upper limit of the energy storage capacity of the transient power storage block; optimize the configuration parameters of the transient power storage block based on the characteristic abnormal events.

[0039] Optionally, during power dispatch, the real-time output capacity of transient power storage blocks is continuously monitored and compared with the generated real-time power dispatch data. When the real-time power dispatch data at a certain moment exceeds the current energy storage capacity limit of the energy storage block—for example, when the discharge power exceeds the maximum allowable discharge power, the output energy demand exceeds the block's available energy range, or the actual available capacity is insufficient due to temperature and attenuation factors—it is considered a characteristic abnormal event. At this time, the system records the abnormal event in the abnormal event database, storing the occurrence time, corresponding user side, power demand value, and block status parameters to form a traceable abnormal behavior dataset. After the abnormal event is recorded, the system triggers a configuration parameter optimization process. Specifically, the system first performs statistical analysis on historical abnormal events to identify the patterns of abnormal occurrence, including high-frequency occurrence time periods, corresponding user side types, demand fluctuation magnitudes, and specific manifestations of insufficient energy storage block capacity. These statistical results, along with the current configuration parameters of the energy storage block, are input into the abnormal handling model, which is constructed in a similar manner to the aforementioned method. By analyzing the anomaly handling model, the root causes of anomalies can be identified, such as insufficient energy storage capacity configuration or peak demand fluctuations exceeding safety margins. Corresponding optimization suggestions are then generated. For example, when anomalies are caused by short-term high-power requests, the maximum discharge power configuration of the energy storage block can be increased; when anomalies are concentrated in long-term high-load requests, the energy storage capacity can be appropriately increased. Finally, the system outputs optimized energy storage block configuration parameters and applies them to reconfiguration or expansion adjustments of energy storage blocks in subsequent configuration or maintenance cycles, thereby reducing the probability of future anomalies and improving overall scheduling reliability and stability.

[0040] In summary, the embodiments of this application have at least the following technical effects: First, historical electricity consumption datasets from the user side are collected, and the user-side dynamic electricity characteristics of these datasets are extracted. Based on these dynamic characteristics, pairwise adaptability analysis is performed on each user side to obtain multiple user-side pairs. Next, multiple transient energy storage blocks are configured according to these user-side pairs, and these transient energy storage blocks are distributedly controlled by the power storage center. Then, power dispatch requests are received, and the requesting user side and its corresponding transient energy storage block are identified. Finally, the power storage status monitoring dataset of the multiple transient energy storage blocks is obtained, and the transient energy storage block corresponding to the requesting user side is controlled to dispatch power to the requesting user side according to the power storage status monitoring dataset. This solves the technical problems of traditional power dispatch, which struggles to dynamically match dispatch resources based on user-side electricity consumption behavior, leading to delayed dispatch response, large local load fluctuations, and low energy storage utilization. It achieves the technical effect of accurate user adaptation and distributed transient energy storage collaborative dispatch based on dynamic electricity characteristics of the user side, improving the real-time performance, stability, and energy storage utilization efficiency of power dispatch.

[0041] Example 2, based on the same inventive concept as the dynamic power consumption data-driven regional power dispatch optimization method in the foregoing examples, such as... Figure 2 As shown, this application provides a regional power dispatch optimization system driven by dynamic power consumption data. The system includes: Adaptability Analysis Module 11: Collects historical electricity consumption datasets from the user side, extracts the user-side power dynamic characteristics from the historical electricity consumption datasets, and performs pairwise adaptability analysis on the user sides based on the user-side power dynamic characteristics to obtain multiple user-side pairs; User-side Configuration Module 12: Configures multiple transient power storage blocks corresponding to the multiple user-side pairs, and the multiple transient power storage blocks are distributedly controlled by the power storage center; Dispatch Request Identification Module 13: Receives power dispatch requests, identifies the requesting user side and the corresponding transient power storage block of the power dispatch request; Power Dispatch Module 14: Monitors the power storage status monitoring dataset of the multiple transient power storage blocks, and controls the transient power storage block corresponding to the requesting user side to perform power dispatch to the requesting user side according to the power storage status monitoring dataset.

[0042] Furthermore, the adaptability analysis module 11 is used to perform the following method: The user-side power dynamic characteristics include short-term ramp rate, average short-term slope, autocorrelation coefficient, peak-hour location distribution vector, and abrupt change frequency based on a preset moving window. Adaptability analysis is performed on each user side according to these user-side power dynamic characteristics to obtain a set of adaptability indicators, including load complementarity, peak-hour shifting degree, instantaneous demand complementarity, and correlation confidence. An adaptation group is initialized, and a greedy adaptation algorithm is used to analyze all user sides within the initialized adaptation group based on the set of adaptability indicators, resulting in multiple user-side pairs.

[0043] Furthermore, the adaptability analysis module 11 is used to perform the following method: Based on the set of adaptability metrics, calculate the average adaptability metric for each user side and other user sides; sort all user sides in descending order based on the average adaptability metric to obtain a set of seed user sides and a set of non-seed user sides; select a first user side from the set of seed user sides, and perform greedy adaptability metric optimization based on the set of non-seed user sides until the set of seed user sides is traversed to obtain the user sides that are suitable for the set of seed user sides.

[0044] Furthermore, the adaptability analysis module 11 is used to perform the following method: After obtaining the suitable user sides from the seed user side set, the adapted user sides are removed from the non-seed user side set to obtain the remaining non-seed user side set; it is determined whether the number of user sides in the remaining non-seed user side set is greater than or equal to 2; if the number of user sides in the remaining non-seed user side set is greater than or equal to 2, a greedy adaptation index is used to optimize the remaining non-seed user side set until the remaining non-seed user side set is traversed to obtain the suitable user sides in the remaining non-seed user side set.

[0045] Furthermore, the adaptability analysis module 11 is used to perform the following method: Obtain the initial adapted user side for each user side in the remaining non-seed user side set; identify the adapted user side including each user side in the remaining non-seed user side set; readjust the pairing relationship by comparing the adaptability index of the initial adapted user side and the adapted user side, and output the updated user side pair.

[0046] Furthermore, the adaptability analysis module 11 is used to perform the following method: Randomly select two pairs from the multiple user-side pairs and calculate the initial overall fitness index; swap the two randomly selected pairs and recalculate the overall fitness index. If the recalculated overall fitness index is greater than the initial overall fitness index, readjust the pairing relationship until the updated user-side pairs are output after K rounds of iteration.

[0047] Furthermore, the user-side configuration module 12 is used to perform the following methods: A load superposition curve is constructed based on the historical load curves of each user side among the multiple user sides. A power shortage index is identified based on the load superposition curve. The configuration parameters of the transient power storage block are obtained using the power shortage index as a constraint. The configuration parameters include energy storage capacity, maximum discharge power, and safety margin.

[0048] Furthermore, the power dispatch module 14 is used to perform the following methods: The power storage status monitoring dataset includes real-time state of charge, capacity decay factor, and temperature decay factor; the power dispatch target of the requesting user is read, and the real-time state of charge, capacity decay factor, and temperature decay factor are analyzed based on the power dispatch target to obtain the real-time power dispatch data output by the transient power storage block to the requesting user.

[0049] Furthermore, the power dispatch module 14 is used to perform the following methods: Record characteristic abnormal events, wherein the characteristic abnormal event is an event in which the real-time power dispatch data exceeds the upper limit of the energy storage capacity of the transient power storage block; optimize the configuration parameters of the transient power storage block based on the characteristic abnormal events.

[0050] The above description is merely a preferred embodiment of the present invention and is not intended to limit the present invention in any way. Although the present invention has been disclosed above with reference to preferred embodiments, it is not intended to limit the present invention. Any person skilled in the art can make some modifications or alterations to the above-disclosed technical content to create equivalent embodiments without departing from the scope of the present invention. Any modifications, equivalent changes, and alterations made to the above embodiments based on the technical essence of the present invention without departing from the scope of the present invention shall still fall within the scope of the present invention.

Claims

1. A regional power dispatch optimization method driven by dynamic power consumption data, characterized in that, The method includes: Collect historical electricity consumption datasets from the user side, extract the user-side power dynamic features from the historical electricity consumption datasets, and perform pairwise adaptability analysis on the user-side based on the user-side power dynamic features to obtain multiple user-side pairs. According to the configuration of the multiple user sides, there are multiple transient power storage blocks, which are distributedly controlled by the power storage center. Receive power dispatch requests and identify the requesting user side of the power dispatch request and the corresponding transient power storage block; The power storage status monitoring dataset of the multiple transient power storage blocks is obtained, and the power storage status monitoring dataset is used to control the transient power storage block corresponding to the requesting user to perform power dispatch to the requesting user.

2. The regional power dispatch optimization method driven by dynamic power consumption data as described in claim 1, characterized in that, Based on the dynamic characteristics of the user-side power supply, pairwise adaptability analysis is performed on each user-side device to obtain multiple user-side pairs. The method includes: The user-side power dynamic characteristics include short-time ramp rate, average short-time slope, autocorrelation coefficient, peak position distribution vector, and abrupt change frequency based on a preset moving window; Based on the dynamic characteristics of the user-side power, a pairwise adaptability analysis is performed on each user-side to obtain a set of adaptability indicators, including load complementarity, peak-hour shifting degree, instantaneous demand complementarity, and correlation confidence. An initialization adaptation group is formed, and a greedy adaptation algorithm is used to analyze all user sides in the initialization adaptation group according to the set of adaptation indicators to obtain multiple user side pairs.

3. The regional power dispatch optimization method driven by dynamic power consumption data as described in claim 2, characterized in that, Based on the set of adaptability indicators, a greedy adaptation algorithm is performed on all user sides in the initialization adaptation group to obtain multiple user side pairs. The method includes: Based on the set of adaptability metrics, calculate the average adaptability metric for each user side and other user sides; Based on the average adaptability index, all user sides are sorted in descending order to obtain the seed user side set and the non-seed user side set. Select a first user side from the seed user side set, and perform greedy adaptation index optimization based on the non-seed user side set until the seed user side set is traversed to obtain the user side that is suitable for the seed user side set.

4. The regional power dispatch optimization method driven by dynamic power consumption data as described in claim 3, characterized in that, After obtaining the seed user side set and the matching user side, remove the matching user side from the non-seed user side set to obtain the remaining non-seed user side set. Determine whether the number of users in the remaining non-seed user set is greater than or equal to 2; If the number of users in the remaining non-seed user set is greater than or equal to 2, a greedy fit index is used to optimize the remaining non-seed user set until the remaining non-seed user set is traversed to obtain the users that are suitable for the remaining non-seed user set.

5. The regional power dispatch optimization method driven by dynamic power consumption data as described in claim 4, characterized in that, The method for optimizing the fit index of the remaining non-seed user set includes: Obtain the initial adapted user side for each user side in the remaining non-seed user side set; Identify the adapted user side, including each user side in the remaining non-seed user side set; The pairing relationship is readjusted by comparing the compatibility index of the initially adapted user side and the already adapted user side, and the updated user side pair is output.

6. The regional power dispatch optimization method driven by dynamic power consumption data as described in claim 2, characterized in that, Based on the set of adaptability indicators, a greedy adaptation algorithm is performed on all user sides in the initialization adaptation group to obtain multiple user side pairs. The method further includes: Randomly select two pairs from the plurality of user-side pairs and calculate the initial overall adaptability index; Swap two randomly selected pairs and recalculate the overall fitness index. If the recalculated overall fitness index is greater than the initial overall fitness index, readjust the pairing relationship until the updated user-side pair is output after K rounds of iteration.

7. The regional power dispatch optimization method driven by dynamic power consumption data as described in claim 1, characterized in that, The method includes configuring multiple transient energy storage blocks corresponding to the multiple user-side configurations: A load superposition curve is constructed based on the historical load curves of each user side among the multiple user sides, and power shortage indicators are identified based on the load superposition curves. Using the power shortage index as a constraint, the configuration parameters of the transient power storage block are obtained, including the energy storage capacity, maximum discharge power, and safety margin.

8. The regional power dispatch optimization method driven by dynamic power consumption data as described in claim 1, characterized in that, The method includes controlling the transient energy storage block corresponding to the requesting user side to perform power dispatch to the requesting user side according to the energy storage status monitoring dataset, and the method includes: The power storage status monitoring dataset includes real-time state of charge, capacity decay factor, and temperature decay factor. The power dispatch target of the requesting user is read, and the real-time state of charge, capacity decay factor and temperature decay factor are analyzed based on the power dispatch target to obtain the real-time power dispatch data output by the transient power storage block to the requesting user.

9. The regional power dispatch optimization method driven by dynamic power consumption data as described in claim 8, characterized in that, After obtaining the real-time power dispatch data output by the transient power storage block to the requesting user, the method further includes: Record characteristic abnormal events, wherein the characteristic abnormal event is an event in which the real-time power dispatch data exceeds the upper limit of the energy storage capacity of the transient power energy storage block; The configuration parameters of the transient power storage block are optimized based on the aforementioned abnormal events.

10. A regional power dispatch optimization system driven by dynamic power consumption data, characterized in that, The system is used to implement the dynamic power consumption data-driven regional power dispatch optimization method according to any one of claims 1-9, the system comprising: Adaptability Analysis Module: Collects historical electricity consumption datasets from the user side, extracts the user-side power dynamic characteristics from the historical electricity consumption datasets, and performs pairwise adaptability analysis on the user side based on the user-side power dynamic characteristics to obtain multiple user-side pairs; User-side configuration module: Based on the configuration of the multiple user-side configurations, there are multiple transient power storage blocks, which are distributedly controlled by the power storage center; Dispatch request identification module: Receives power dispatch requests and identifies the requesting user side of the power dispatch request and the corresponding transient power storage block; Power dispatch module: monitors and obtains the power storage status monitoring dataset of the multiple transient power storage blocks, and controls the transient power storage block corresponding to the requesting user side to perform power dispatch to the requesting user side according to the power storage status monitoring dataset.

Images