Method and system for intelligently adjusting charging and discharging of energy storage device based on multiple regions

Abstract

The invention relates to the technical field of intelligent charging and discharging adjustment, in particular to an intelligent charging and discharging adjustment method and system for an energy storage device based on multiple regions, and the method comprises the following steps: based on the data of energy users in each region, analyzing the power trend of a time period, recognizing and marking abnormal nodes, judging the deviation of a load trend in combination with monitoring and early warning, and optimizing an adjustment sequence; and dynamically balancing capacity distribution, and generating a charging and discharging instruction. According to the invention, real-time data fusion is carried out based on periodic factors and holiday correction, so that abnormal fluctuation and energy consumption trend change in each time period have dynamic identification capability, and advanced identification and linkage of load abnormal omen are realized through trend monitoring and early warning feature construction for load differences among nodes. The energy support distribution is ranked according to the energy storage state and the output capacity, fine matching of task objects and instruction schemes is promoted, the participation sequence of the energy storage units and an energy release window are dynamically optimized, and regional energy adjustment action collaboration is achieved.

Description

Technical Field

[0001] The present invention relates to the technical field of intelligent charging and discharging regulation, and in particular to a method and system for intelligent charging and discharging regulation of a multi-region energy storage device. Background Art

[0002] The field of intelligent charge and discharge regulation mainly involves the efficient management and optimized scheduling of energy flows during the charging and discharging process of energy storage devices, including the formulation of energy storage device operation strategies, the design of energy distribution plans, and the regulation of charging and discharging time and power. Overall, the charging and discharging behavior of energy storage systems is scientifically managed through intelligent means to improve energy storage efficiency, ensure the safe and stable operation of the power system, and adapt to changing electricity demand. Among them, the traditional intelligent charge and discharge regulation method for energy storage devices refers to the use of centralized calculations or distributed rules to determine the charging and discharging sequence, time period, and specific charging and discharging power of energy storage devices in each region for multiple energy storage units distributed in different geographical regions or power network nodes. It usually uses historical load data statistical analysis and fixed threshold methods to judge the load status, and then preset the charging and discharging priority and working period allocation of energy storage units in different regions.

[0003] Existing technologies are limited to historical load data and fixed judgment rules. They are not sensitive enough to abnormal fluctuations and sudden behaviors in time series data, and are unable to track inter-regional load trends and dynamic changes in a timely manner. During special periods such as holidays, the actual load may not match the preset pattern. When load offset occurs, the early warning and adjustment measures rely on preset thresholds, resulting in delayed response actions. It is difficult to cope with rapid changes in energy storage status and energy flow between nodes, affecting the timeliness and accuracy of the dispatching system's inter-regional energy balance and emergency management. Summary of the Invention

[0004] The purpose of the present invention is to solve the shortcomings of the prior art and to propose a method and system for intelligently regulating the charging and discharging of an energy storage device based on multiple regions.

[0005] In order to achieve the above object, the present invention adopts the following technical solution: a method for intelligently regulating the charge and discharge of an energy storage device based on multiple regions, comprising the following steps: S1: Based on the energy users in each region, analyze the power changes during daily periods, compare the load optimization items, cycle factors, and holiday correction parameters, determine the stability of the cycle factors, screen out abnormal fluctuation periods, mark them as adjustment nodes, and obtain the time series abnormality mark set; S2: Based on the time series anomaly mark set, analyze the difference between the load forecast and the real-time collected data of the adjustment node, determine the difference trend between consecutive time periods, and establish trend anomaly warning features in combination with the warning information of the load monitoring equipment; S3: Based on the abnormal trend warning characteristics, analyze the remaining power and output power of the energy storage units, calculate their load support capabilities, and select the energy storage units with the best capabilities after sorting to obtain a dynamic load support sequence; S4: Based on the dynamic load support sequence, analyze the capacity utilization ratio and change rate of the energy storage unit, determine the amplitude difference, identify the energy storage unit with a low change rate as the charging target and the energy storage unit with a high change rate as the discharging target, and obtain a capacity balancing instruction list.

[0006] The present invention has the following improvements: the timing anomaly mark set includes a time period label, an abnormal fluctuation level, and an adjustment reference index; the trend anomaly warning feature includes an offset trend factor, an abnormality type classification, and a warning level parameter; the dynamic load support sequence includes a support capacity ranking, an energy allocation identifier, and a node response attribute; and the capacity balance instruction list includes a charging object number, a discharging object number, and a capacity balance level.

[0007] The present invention is improved in that the steps of obtaining the time series anomaly marker set are specifically as follows: S111: Based on the energy users in each region, analyze the power change trend of the users in each time period of the day, compare the relationship between the maximum load item in each time period and the cycle factor and holiday correction parameter, determine the synchronization and difference in the change process, and obtain the correlation trend feature sequence; S112: Based on the correlation trend feature sequence, determine whether the change of the periodic factor in the continuous time period is stable, calculate the fluctuation of the periodic factor in each continuous time period, screen the time period with abnormal fluctuation, optimize the time period division, and obtain the periodic continuous change pattern; S113: Based on the periodic continuous change pattern, screen the time nodes with abnormal fluctuation amplitudes, compare the distribution and change characteristics of each node in the sequence, calculate the abnormal fluctuation aggregation index, combine the index with the node trajectory, and obtain a time series abnormality label set.

[0008] The present invention is improved in that the steps of obtaining the abnormal trend warning feature are specifically as follows: S211: Based on the time series anomaly marker set, analyze and adjust the changes in the load forecast data and the real-time collected data of the node, calculate the forecast offset development trend between consecutive time periods, compare the change rate of each node in the continuous time with the change of the regional load change limit, screen the trend change of each node, and obtain the trend change response sequence; S212: Based on the trend change response sequence, determine the trend growth of each node, analyze the relationship between the node trend change and the upper and lower bounds of the regional load change allowable range, filter out nodes whose trend changes exceed the allowable range, optimize node classification, and obtain a trend deviation node set; S213: Based on the trend deviation node set, compare the trend change of each node with the alarm status of the monitoring equipment, screen the nodes with key trend warning signal strength, judge the trend direction, diffusion speed and device response of the node, and establish trend abnormality warning characteristics.

[0009] The present invention is improved in that the steps of obtaining the dynamic load support sequence are specifically as follows: S311: Based on the abnormal trend warning characteristics, analyze the remaining power and sustainable output status of the regional energy storage unit, calculate its corresponding unit time support capacity and energy release limit range, adjust the parameter proportions and normalize them, and obtain the unit support capacity parameter; S312: Based on the unit support capacity parameters, compare the support differences of each unit at the abnormal node, calculate the regulation capacity strength of each energy storage unit, perform a sequence sorting operation, and obtain an energy storage unit sorting sequence; S313: Based on the energy storage unit sorting sequence, filter the units with priority, determine the matching order between them and each node, adjust the node mapping structure and rearrange the numbering correspondence, determine the dynamic linkage order of the regulation resources, and obtain the dynamic load support sequence.

[0010] The present invention is improved in that the steps of obtaining the capacity balancing instruction list are specifically as follows: S411: Based on the dynamic load support sequence, analyze the current capacity utilization ratio and change rate of the energy storage unit, compare the change range of each parameter within the same cycle, and determine the difference between the two by comparing their growth or decline trends to obtain a capacity parameter comparison feature; S412: Based on the capacity parameter comparison characteristics, classify the energy storage units according to their change rates, optimize the classification results, classify the units with low change rates as charging targets, and classify the units with high change rates as discharging targets, and obtain charge and discharge group identifiers; S413: Analyze the current capacity utilization ratio, change rate, output power, and regional load response of each energy storage unit based on the charge and discharge group identifier, determine the charge and discharge task category and allocation relationship of each energy storage unit in the current cycle, and obtain a capacity balancing instruction list.

[0011] The present invention is improved in that the steps further include: S5: Based on the capacity balancing instruction list, adjust the task category and regional allocation of each energy storage unit, optimize the charging and discharging capacity of the energy storage units in the adjustment sequence, determine the energy release order and instruction time period, issue adjustment commands, and obtain the partition linkage execution results; The partition linkage execution result includes a coordinated adjustment identifier, a regional energy allocation ratio, and an execution feedback signal.

[0012] The present invention is improved in that the steps of obtaining the partition linkage execution result are specifically as follows: S511: Based on the capacity balancing instruction list, analyze the charging object number, discharging object number, and capacity balancing level, optimize the task category allocation of each energy storage unit, compare the area number and node distribution corresponding to each unit, determine the adaptability of the task category and area division, and obtain the energy storage task partition mapping list; S512: Calculate the matching relationship between the task category of each energy storage unit and the energy storage capacity of the region in which it is located based on the energy storage task partition mapping list, optimize the adjustment order, adjust the correspondence between the region number and the task category, and obtain the energy storage task scheduling sorting identifier; S513: Call the energy storage task scheduling sorting identifier, analyze the adjustment command issued by the regional energy storage scheduling communication terminal, compare the energy storage unit feedback signal with the task category action, screen the task completion status and organize the communication results to obtain the partition linkage execution result.

[0013] A multi-region energy storage device charging and discharging intelligent regulation system, the system comprising: The time series feature extraction module analyzes the power change trends in each time period of the day based on energy users in each region. It compares the correlation characteristics of the maximum load item in each time period with the cycle factor and holiday correction parameters, optimizes the adaptability of the cycle factor and correction parameters at different times, determines the stability of the continuous change of the cycle factor, screens out abnormal fluctuation periods and marks them as adjustment nodes, and obtains a time series abnormality tag set. The load anomaly identification module analyzes the difference between the load forecast data and the real-time collected data of each adjustment node based on the time series anomaly mark set, calculates the difference trend between consecutive time periods, compares the degree of matching between the trend change and the regional load variation range, identifies the nodes whose trend change is greater than the range, and establishes trend anomaly warning features in combination with the warning prompts of the regional node load monitoring equipment; The energy storage optimization sorting module optimizes the remaining power and maximum output power of the energy storage units in the associated area based on the abnormal trend warning characteristics, calculates the load support capacity index of each energy storage unit per unit time, compares the order of the capacity index of each energy storage unit in the node, selects the energy storage unit with the best capacity index, and obtains the dynamic load support sequence; The intelligent capacity dispatch module analyzes the current capacity utilization ratio and change rate of the energy storage unit based on the dynamic load support sequence, determines the difference between the two parameters, optimizes the dynamic balance between the capacity utilization ratio and the change rate, identifies the energy storage unit with a low change rate as the charging target and the energy storage unit with a high change rate as the discharging target, and obtains a capacity balance instruction list; Based on the capacity balancing instruction list, the collaborative scheduling execution module adjusts the task category and regional allocation of each energy storage unit, optimizes the charging and discharging capabilities of the energy storage units within the adjustment sequence, determines the energy release order and instruction execution period between the energy storage units, and synchronously issues adjustment commands based on the regional energy storage scheduling communication terminal to obtain the partition linkage execution results.

[0014] Compared with the prior art, the advantages and positive effects of the present invention are: In the present invention, through the deep extraction of multi-dimensional data of regional energy consumption behavior time series, real-time data fusion is carried out based on periodic factors and holiday corrections, so that abnormal fluctuations and energy consumption trend changes in each time period can be dynamically identified. Trend monitoring and early warning features are constructed for load differences between nodes to achieve early identification and linkage of load anomaly signs. Energy support allocation is sorted according to energy storage status and output capacity, promoting the fine matching of task objects and instruction plans, dynamically optimizing the participation sequence of energy storage units and the energy release window, realizing the coordination of regional energy regulation actions, making the inter-regional task instruction allocation more adaptive, promoting the continuous optimization of energy flow paths and resource allocation relationships, and improving the flexibility and accuracy of multi-region power network regulation. BRIEF DESCRIPTION OF THE DRAWINGS

[0015] Figure 1 It is a flow chart of the main steps of the present invention; Figure 2 This is a flow chart for obtaining a time series anomaly marker set in the present invention; Figure 3 This is a flow chart for obtaining the abnormal trend warning feature in the present invention; Figure 4 This is a flow chart for obtaining a dynamic load support sequence in the present invention; Figure 5 This is a flow chart for obtaining a capacity balancing instruction list in the present invention; Figure 6 This is a flowchart for obtaining partition linkage execution results in the present invention. DETAILED DESCRIPTION

[0016] In order to make the purpose, technical solutions and advantages of the present invention more clearly understood, the present invention will be further described in detail below with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain the present invention and are not intended to limit the present invention.

[0017] In the description of the present invention, it should be understood that the terms "length", "width", "up", "down", "front", "back", "left", "right", "vertical", "horizontal", "top", "bottom", "inside", "outside", etc., indicating directions or positional relationships, are based on the directions or positional relationships shown in the accompanying drawings and are only for the convenience of describing the present invention and simplifying the description. They do not indicate or imply that the devices or elements referred to must have a specific direction, be constructed and operate in a specific direction, and therefore should not be understood as limiting the present invention. In addition, in the description of the present invention, the meaning of "plurality" is two or more, unless otherwise clearly and specifically defined.

[0018] Example:

[0019] See also Figure 1 The present invention provides a technical solution: a multi-region energy storage device charging and discharging intelligent regulation method, comprising the following steps: S1: Based on the energy users in each region, analyze the power change trend in each time period of the day, compare the correlation characteristics of the maximum load item in each time period with the cycle factor and holiday correction parameters, optimize the adaptability of the cycle factor and correction parameters in the time difference, determine the stability of the continuous change of the cycle factor, screen the abnormal fluctuation period and mark it as an adjustment node, and obtain the time series abnormality mark set; S2: Based on the time series anomaly tag set, analyze the difference between the load forecast data and the real-time collected data of each adjustment node, calculate the difference trend between consecutive time periods, compare the degree of match between the trend change and the regional load variation range, identify the nodes with trend changes greater than the range, and combine the early warning prompts of the regional node load monitoring equipment to establish the trend anomaly warning feature; S3: Based on the trend anomaly warning characteristics, the remaining power and maximum output power of the energy storage units in the associated area are optimized, the load support capacity index of each energy storage unit per unit time is calculated, the order of the capacity index of each energy storage unit in the node is compared, and the energy storage unit with the best capacity index is selected to form the current adjustment sequence and obtain the dynamic load support sequence; S4: Based on the dynamic load support sequence, analyze the current capacity utilization ratio and change rate of the energy storage unit, determine the difference in the amplitude of the two parameters, optimize the dynamic balance of capacity utilization ratio and change rate, identify energy storage units with low change rate as charging targets and energy storage units with high change rate as discharging targets, and obtain a capacity balance instruction list; S5: Based on the capacity balancing instruction list, adjust the task category and regional allocation of each energy storage unit, optimize the charging and discharging capacity of the energy storage units within the adjustment sequence, determine the energy release order and instruction execution period between each energy storage unit, and synchronously issue adjustment commands based on the regional energy storage scheduling communication terminal to obtain the partition linkage execution results.

[0020] The time series anomaly tag set includes time period label, abnormal fluctuation level, and adjustment reference index. The trend anomaly warning characteristics include offset trend factor, abnormal type classification, and warning level parameter. The dynamic load support sequence includes support capacity ranking, energy allocation identification, and node response attributes. The capacity balance instruction list includes charging object number, discharging object number, and capacity balance level. The partition linkage execution results include collaborative adjustment identification, regional energy allocation ratio, and execution feedback signal.

[0021] In S1, the maximum load item refers to the item with the largest value in the user's electricity (or energy) load data within a certain time period, reflecting the peak energy consumption during that period and is usually used for load characteristic analysis. The periodic factor refers to the parameter that affects the change of energy load within a period (such as day, week, or month), and common factors include working days / non-working days, time period (morning or evening), and temperature. The holiday correction parameter is a parameter used to correct energy consumption anomalies caused by special times such as statutory holidays, and is usually set based on historical holiday data and social habits. The adaptability of the difference time refers to the degree of match and adjustment flexibility between the periodic factor and holiday correction parameter and actual energy consumption data in different time periods (or different types of time periods). The stability of continuous change refers to whether the change trend of the periodic factor is stable (i.e., with small fluctuations) and does not experience drastic jumps over multiple consecutive time periods. The period of abnormal fluctuation refers to the period when parameters such as the periodic factor suddenly change or deviate from the normal pattern, which usually indicates abnormal energy consumption pattern or load. The adjustment node refers to the time point / data node that needs to be focused on and subsequently analyzed when the period is determined to be abnormally fluctuating.

[0022] In S2, the difference trend refers to the development trend of the deviation between the load forecast data and the real-time collected data, which changes continuously over time, reflecting the accuracy of the forecast and its changing state over time; the regional load variation range refers to the amplitude range of the energy load in a specific area that is allowed to fluctuate under normal circumstances, which can be determined by historical data statistics and used to determine whether abnormal deviations occur; nodes larger than this range refer to time nodes where the difference trend exceeds the normal range of regional load variation mentioned above and are identified as abnormal nodes; load monitoring equipment refers to instruments and equipment installed at nodes in each area that can collect and record energy consumption data in real time and have alarm / early warning functions, such as smart meters and sensors.

[0023] In S3, an energy storage unit refers to an energy storage system module that can charge and discharge independently and receive scheduling instructions, such as an energy storage device, a group of batteries, or a regional energy storage system; supported load refers to the power that an energy storage unit can release to the outside, which is used to provide supplementary power to the power grid or energy users and buffer load fluctuations; the capacity index is a parameter that comprehensively evaluates the ability of an energy storage unit to support the load over a period of time, generally considering remaining power, output power, response speed, etc.; sorting in the node refers to sorting the capacity indicators of multiple energy storage units according to their quality, which is used for subsequent scheduling priority allocation; the adjustment sequence refers to the arrangement queue of energy storage units that can participate in this round of scheduling after being sorted by capacity indicators.

[0024] In S4, the utilization ratio refers to the ratio between the current available capacity of the energy storage unit and its rated maximum capacity, reflecting the real-time availability of the energy storage resource; the change rate refers to the speed at which the utilization ratio changes over time, which is used to judge the charging and discharging activity of the energy storage unit; the parameter amplitude difference refers to the absolute difference between the utilization ratio and the change rate, which is used to analyze the fluctuation characteristics of the energy storage unit status; dynamic balance refers to coordinating the division of labor of charging and discharging of each energy storage unit during the scheduling process, so that the overall capacity status maintains a reasonable distribution and avoids overload or idleness of some energy storage units.

[0025] In S5, task category refers to the functional tasks undertaken by the energy storage unit according to the dispatch instructions, such as charging, discharging, and standby; regional allocation refers to the allocation of specific service areas for different energy storage units and the determination of their support objects (such as a regional power grid, load nodes, etc.); energy release sequence refers to when multiple energy storage units need to release energy in sequence, the order of their actions is determined according to priority to avoid energy fluctuations caused by simultaneous actions; instruction execution period refers to the time window when the dispatch instructions are specifically effective and executed, ensuring that the operation sequence is synchronized with system requirements; dispatch communication terminal refers to the terminal equipment used for real-time communication and instruction issuance with each energy storage unit, including the background dispatch host, remote communication module, etc.

[0026] See also Figure 2 ,The specific steps for obtaining the time series anomaly tag set are: S111: Based on the energy users in each region, analyze the power change trend of the users in each time period of the day, compare the relationship between the maximum load item in each time period and the cycle factor and holiday correction parameter, determine the synchronization and difference in the change process, and obtain the correlation trend feature sequence; First, collect the historical power data of all electricity users in each area, divide each day into multiple time periods of fixed length, such as one time period per hour, and then count the power changes of each user in each time period in turn, and statistically average the historical power in the same time period of each day to obtain the typical power curve of each user in the time period. Then, identify the maximum value of all power data in the time period as the maximum power item, and add the time label, date category (such as weekday or weekend) and holiday status mark to the data point. Then, according to the date category corresponding to the time label, extract the corresponding label from the periodic factor parameter library, such as "Monday morning" or "Friday afternoon", etc., compare the label with the maximum power item of the time period, and extract the corresponding correction factor from the holiday correction parameter table and associate it with the current time period. For example, during statutory holidays, the average load of users drops to about 70% of weekdays, then the time period is The correction factor is set to a negative value, indicating a deviation from the normal cycle energy consumption level. Then, the maximum power of each user in this period is combined with the cycle factor and the holiday correction factor to form a combined data item. Such combinations of all users in each area in this period are collected, and the numerical difference between the maximum power values ​​is compared item by item. If the difference exceeds a certain set threshold, such as the maximum power difference between different users exceeds 3 kilowatts, the period is marked as having a synchronization deviation. The change process of the maximum power under different cycle factor labels is then checked. For example, if the difference in the maximum power values ​​between "Wednesday morning" and "Friday morning" in the same time period also reaches a certain amplitude, it is determined that the influence of the cycle factor is volatile. Finally, the linkage trend between this maximum power item, the cycle factor, and the holiday correction item in all time periods is summarized. By observing the continuously changing data, the power change trend of each time period under different cycles is extracted to form a correlation trend feature sequence.

[0027] S112: Based on the correlation trend feature sequence, determine whether the change of the periodic factor in the continuous time period is stable, calculate the fluctuation of the periodic factor in each continuous time period, screen the time period with abnormal fluctuation, optimize the time period division, and obtain the periodic continuous change pattern; Divide the three consecutive time periods of each day into multiple small segments, calculate the maximum and minimum difference between the power trend values ​​in each small segment, and record the difference as the fluctuation amplitude of the segment. Further check whether the amplitude exceeds the preset fluctuation threshold. For example, if the set threshold is 1.5 kilowatts, when the difference between the maximum and minimum power trend values ​​in a certain segment reaches 2.2 kilowatts, the time period is identified as an abnormal fluctuation segment. Then, traverse all consecutive time periods and analyze the difference in the trend change rate before and after. If the change rate between two adjacent time periods is greater than the set percentage threshold, such as 25%, it is regarded as a manifestation of continuous and unstable change of the periodic factor, and the time period is further recorded as Unstable segments, sort out all time periods that are judged to be abnormal fluctuations or drastic changes, and form an abnormal time period set, and then compare them with the originally set time period division, and further split or reorganize those abnormal time periods. For example, if an abnormal fluctuation segment occurs within a certain 45 minutes in the middle of an originally set four-hour period, the four-hour period will be divided into three smaller time periods, so that the fluctuation characteristics within each divided unit remain relatively consistent. Finally, based on the trend changes of all time periods and the re-division results, draw a fluctuation stability map of the entire periodic factor in the continuous time period, identify the change trajectory and continuity of the periodic factor value in each time period, and complete the construction of the periodic continuity change model.

[0028] S113: Based on the periodic continuous change pattern, screen the time nodes with abnormal fluctuation amplitude, compare the distribution and change characteristics of each node in the sequence, and use the formula: ; Calculate the abnormal fluctuation aggregation index and combine it with the node trajectory to obtain the time series abnormality label set, where represents the abnormal fluctuation aggregation index of the jth period in the i-th region, is the maximum load item in the kth period of the i-th region, is the fluctuation of the period factor in the kth period, is the time span of continuous change of the periodic factor corresponding to the kth period, is the intra-period variation term of the holiday correction parameter in the kth period of the i-th region, and n is the total number of consecutive periods.

[0029] The abnormal fluctuation aggregation index is a comprehensive quantitative indicator that reflects the intensity of abnormal fluctuations in energy load during a specific period of time in each region after aggregating and analyzing multiple data such as the maximum load item, the fluctuation amount of the periodic factor, the periodic change span and the holiday correction parameters. It is used to measure the coupling fluctuations between the load changes and the periodic laws and special time correction parameters in a certain period of time in the region, and to highlight the time nodes when obvious anomalies or mutations occur in the load behavior.

[0030] If the sum of the squares of the periodic factor fluctuations in three consecutive periods exceeds the set reference interval, the node is judged to have an abnormal fluctuation amplitude and is numbered j. The area number i and its combination are used as the unique identification mark. Then, further data aggregation calculation is performed on such nodes in each area, and the maximum load item of each area is called. , Cycle Factor Volatility , cycle span 、Holiday changes , calculate its abnormal fluctuation aggregation index; In a certain area i=1, select a set of consecutive time periods k=1,2,3, and the data is as follows: Period 1: , after normalization, it is 0.83; , after normalization, it is 0.75; , after normalization, it is 0.60; , after normalization, it is 0.50; Period 2: , after normalization, it is 0.88; , after normalization, it is 0.45; , after normalization, it is 0.60; , after normalization, it is 0.75; Period 3: , after normalization, it is 0.95; , after normalization, it is 0.35; , after normalization, it is 0.60; , after normalization, it is 0.25; After substituting the normalized data, the numerator is calculated: ; Denominator calculation: Calculation results: ; at this time , indicating that strong abnormal fluctuation aggregation occurred in the process of cyclical continuity changes in the first area during this period. If 0.50 is defined as the upper-middle fluctuation level threshold in the set fluctuation response identification interval, then the node will be identified as an abnormal node and included in the subsequent regulation strategy identification scope. The trajectory number, location index, and indicator record of the node will be constructed, and finally written into the time series anomaly mark set as the node information source for subsequent energy storage regulation judgment. This formula introduces the fusion effect of the load and rhythm change factors by multiplying the maximum load term with the square root of the cycle factor fluctuation. Combined with the superposition factors of the cycle span and holiday parameters, it quantitatively expresses the abnormal intensity level of the node within the overall scheduling cycle, improving the granularity of node status identification during the scheduling process.

[0031] See also Figure 3 ,The specific steps for obtaining the trend anomaly warning features are: S211: Based on the time series anomaly marker set, analyze and adjust the changes in the load forecast data and real-time collected data of the node, calculate the forecast offset development trend between consecutive time periods, compare the change rate of each node in the continuous time with the change of the regional load change limit, screen the trend change of each node, and obtain the trend change response sequence; The location of each marked node during a specific time period of the day is determined, and the load forecast data for that node during that time period is extracted. This forecast data is derived from the historical average load curve data of the previous cycle. This data is collected multiple times daily to form a dynamically updated forecast curve. Real-time load data collected for that node during the corresponding time period is also extracted. This data comes from smart meter collection modules deployed in the area, recording each load value at the minute level. The forecast data and real-time data are arranged in chronological order, and the difference between the forecast value and the actual value at each adjacent time point is compared. The trend of the overall forecast deviation is determined by calculating the continuous increase and decrease direction of the difference sequence point by point. For example, between 6:00 and 7:00 in the morning, the forecast data is 4.8, 5.1, 5.3, 5.6, and 6.0 kilowatts, while the real-time data is 5.0, 5.5, 5.9, 6.3, and 6.8 kilowatts, corresponding to the offset values ​​of 0.2, 0.4, 0.6, 0.7, and 0. 8 kW, the offset value shows an increasing trend in a continuous time period, and the growth amplitude of this development trend is recorded as the trend change rate. Then, the upper and lower bound values ​​are extracted according to the load change limit set for the area. For example, the normal fluctuation range of the historical load in this area during this period is set to ±0.5 kW, and fluctuations exceeding this range are considered abnormal. The relationship between the trend change rate and the upper and lower bounds of the change limit is compared. If the offset value growth rate exceeds 0.5 kW every fifteen minutes, it is marked as a node with a trend change that is too fast. The trend change rates of all nodes are horizontally sorted to screen out nodes with obvious growth rate changes. For example, the boundary interval is set to three levels, with less than 0.3 kW / 15 minutes as a stable period, between 0.3 and 0.5 as an acceptable period, and above 0.5 as a drastic period. All nodes are classified accordingly, and parameters such as their time period, change value, and growth direction are recorded. Finally, the screened nodes are arranged in order to form a trend change response sequence.

[0032] S212: Based on the trend change response sequence, determine the trend growth of each node, analyze the relationship between the node trend change and the upper and lower bounds of the regional load change allowable range, filter out nodes whose trend changes exceed the allowable range, optimize node classification, and obtain a set of trend deviation nodes; Extract the trend change direction and change rate value of each node to determine whether the node is in a growth trend within the time period. The judgment is made by comparing whether the load value at the current time point and the previous time point continues to rise. For example, if the load of a node increases from 5.2 kilowatts to 7.1 kilowatts between 8:00 and 9:00 in the morning, it is determined to be a trend growth node. Then compare the trend change value of the node with the upper and lower bounds of the regional load allowable change to make an associated judgment. The regional load allowable fluctuation range is formulated according to the historical seasonal load changes. For example, the historical maximum fluctuation range from 8:00 to 9:00 in the morning during the summer peak period is 1. The degree is 1.5 kilowatts. If the node change value reaches 2.2 kilowatts, it exceeds the upper limit and is determined to be a trend deviation node. Then all nodes that exceed the upper and lower limits are classified, and multi-dimensional labels are assigned according to the trend growth amplitude, duration, and area. For example, the growth amplitude level of the trend deviation node is set to 1.5-2.0 kilowatts as a first-level deviation, 2.0-3.0 kilowatts as a second-level deviation, and more than 3.0 kilowatts as a third-level deviation. Then, the deviation node set is sorted according to the classification label, and the node number, occurrence time period, deviation direction, deviation amplitude and its area information are summarized to form a trend deviation node set.

[0033] S213: Based on the trend deviation node set, compare the trend change of each node with the alarm status of the monitoring device, using the formula: ; Screen the nodes with key trend warning signal strength, determine the trend direction, diffusion speed and device response of the nodes, and establish trend abnormality warning features, among which, Indicates the trend warning signal strength of node z, Indicates the trend change amplitude of node z, represents the total load offset of node z, represents the prediction difference variance of node z, Indicates the upper limit of load change in node z area, Indicates the lower bound of load variation in node z area, represents the alarm level factor of node z and monitoring device j, represents the trend coupling amplitude between node z and monitoring device j, Indicates the number of monitoring devices associated with node z, Represents the total number of monitoring devices associated with node z, and the sum is accumulated for each j.

[0034] Trend warning signal strength refers to a quantitative indicator used to measure the risk of abnormal load trends at nodes in multiple regions of energy storage devices during intelligent charging and discharging regulation. This indicator is calculated by integrating multiple factors, including node trend changes, load offsets, prediction errors, and the alarm status of monitoring equipment. It reflects the degree of deviation between the node's load change trend and multiple data sources, including actual collected data, forecast data, and device status, within a specific time period, and its ability to respond to abnormalities. A larger value indicates a greater deviation from the normal range for the region, and a higher correlation with signals such as monitoring equipment alarms, indicating a higher likelihood of abnormal load or warnings. This indicator facilitates the system's rapid identification and prioritized response to risky nodes, enabling proactive early warning and regulation of the zoned energy storage system.

[0035] Taking node Z01 as an example, its original data is: , alarm level factor The array is [1.2, 0.9, 1.1], the trend coupling amplitude The array is [1.0, 0.8, 1.2]. After normalizing each parameter according to its category, we get: , the normalized alarm factor group is [0.80, 0.66, 0.72], and the normalized coupling amplitude is [0.71, 0.59, 0.83].

[0036] Substitute the above normalized values ​​into the formula for calculation: Part I: Molecules ; The first part of the denominator: ; Results of the first part: ; The second part of the summation: The overall value of the second part: ; Combined calculations yield the final trend warning signal strength: ; This result indicates that the multi-parameter composite trend change characteristics of node Z01 in the current cycle have reached a high alarm response level, and its trend change value has exceeded the set warning response limit. The priority of the energy storage resource coordinated response of this node must be given priority in subsequent adjustment links. This value is used as the core input basis for constructing the trend anomaly warning feature. By simultaneously introducing trend amplitude, error factor and multi-device coupling data, the formula incorporates temporal trend evolution and spatial response into a unified measurement framework, improving the stability and difference identification accuracy of the distributed trend alarm mechanism.

[0037] See also Figure 4 , the steps for obtaining the dynamic load support sequence are as follows: S311: Based on the trend anomaly warning characteristics, analyze the remaining power and sustainable output status of the regional energy storage units, calculate their corresponding unit time support capacity and energy release limit range, adjust the parameter proportions and normalize them, and obtain the unit support capacity parameters; Extract the numbers of all regional energy storage units that are currently in the abnormal warning state, and obtain their latest remaining power data one by one. The remaining power is the ratio of the current power reading to the rated maximum power of the energy storage unit multiplied by the rated power. For example, the current power of the energy storage unit numbered A101 is 42 kWh and the rated capacity is 100 kWh, so its remaining power is 42 kWh. Then obtain the average output power and the maximum continuous discharge time of the unit in the past hour, calculate the amount of power that can be released per unit time and judge whether it is in a sustainable output state based on this. For example, if the output power is 7 kW and the discharge duration is 3 hours, the total sustainable output power is 21 kWh, and the corresponding unit hour support capacity is 7 kW. Repeat this operation for all energy storage units and record their respective unit time support capacities. Then, combine the rated output power upper limit and minimum release power of each unit to establish The limiting range of the energy release of this unit is 3 to 10 kilowatts if the rated power of number A101 is 10 kilowatts and the minimum release power is 3 kilowatts. The limiting range is 3 to 10 kilowatts. Then, the remaining electric energy value, unit time support capacity value, and upper and lower limits of the limiting range of each unit are multiplied by the adjustment weight factor. The setting of this factor is based on the importance level of the area to which the energy storage unit belongs and the frequency of its previous scheduling participation. For example, if the importance level is level three and the number of participations in the last five times is 3, the corresponding weight is set to 0.85. The three-dimensional data are multiplied by this factor and then proportionally normalized so that the final three parameters can be compared within the same range. The parameter values ​​of all energy storage units are mapped to between 0 and 1 using linear normalization. Finally, the normalized support capacity value of each energy storage unit is recorded separately, and this value is used as the basis for subsequent scheduling priority sorting to generate the unit support capacity parameter.

[0038] S312: Based on the unit support capacity parameters, compare the support differences of each unit at the abnormal node using the formula: ; Calculate the regulation capability strength of each energy storage unit, perform sequence sorting operations, and obtain the energy storage unit sorting sequence, where: Indicates the regulation capability strength of the energy storage unit numbered o, Indicates the standard support power per unit time of the energy storage unit numbered o. Indicates the current remaining available electrical energy of the energy storage unit numbered o. Indicates the matching degree of support capacity between the energy storage unit numbered o and the target node, Indicates the fluctuation interference amount generated by the energy storage unit numbered o in the node load regulation, Indicates the response time of the energy storage unit numbered o to the current node task; Regulation capacity intensity refers to the core indicator that quantitatively measures the ability of energy storage units to make immediate contributions to regional energy regulation tasks after integrating all relevant status parameters in a specific time period and node environment. It is the direct decision-making basis for the energy storage scheduling system to intelligently screen, allocate and sort energy storage resources.

[0039] The original parameters of the energy storage unit numbered E01 are: standard support power 85.5kW, remaining available power 180.0kWh, matching degree 0.92, fluctuation interference 5.6, and response time 2.0. Its normalized parameters are: , , , , , substitute into the formula and calculate as follows: Additive calculation: ; Multiply by the degree of match: ; Subtract the interference: ; Denominator processing: ; Adjustment ability strength: ; Similarly, for the energy storage unit numbered E02, the original parameters are: standard support power 92.0kW, remaining available power 150.0kWh, matching degree 0.87, fluctuation interference 7.2, and response time 2.5. Its normalized parameters are: , , , , , the calculation steps are as follows: Additive calculation: ; Multiply by the degree of match: ; Subtract the interference: ; Denominator processing: ; Final adjustment capability strength: ; For the energy storage unit numbered E03, the original parameters are: standard support power 88.3kW, remaining available power 210.0kWh, matching degree 0.89, fluctuation interference 4.9, and response time 1.8. Its normalized parameters are: , , , , , calculated as follows: Additive calculation: ; Multiply by the degree of match: ; Subtract the interference: ; Denominator processing: ; Final adjustment capability strength: ; The results show that E03 has the highest regulation capability, followed by E01, and finally E02. The ranking result is: E03>E01>E02. This result is the energy storage unit ranking sequence generated in this step, which can provide a basis for issuing priority instructions for subsequent node support linkage. The result shows that the larger the value of the regulation capability strength, the higher the support response capability of the energy storage unit to abnormal nodes in the current cycle.

[0040] S313: Based on the energy storage unit sorting sequence, prioritized units are screened, their matching order with each node is determined, the node mapping structure is adjusted, and the numbering correspondence is rearranged to determine the dynamic linkage order of the regulation resources, thereby obtaining a dynamic load support sequence. Select the energy storage units with the highest ranking, extract their numbers as the current adjustment resource candidate set, read the load response parameters of each abnormal node one by one in the order of scheduling time, the parameters include the load gap value of the node in a certain period, the response delay and the area number, compare the energy storage unit with the best ranking currently with each node one by one, and compare whether its supporting capacity parameter is greater than the node gap value, whether it is located in the area to which the node belongs or can be scheduled to the area and the scheduling response time does not exceed the maximum tolerance delay time of the node. If all conditions are met, establish a preliminary matching relationship between the energy storage unit and the node, and then record the matching relationship as a record of the node mapping relationship, and adjust the remaining power and expected adjustment of the current energy storage unit after each successful match. The number of times it matches is inserted at the end of the next round of scheduling sorting queue. For example, if energy storage unit A101 matches node N005, then "node N005-energy storage unit A101" is recorded, and A101 is moved from the current sequence to the temporary task list. The above process is repeated until all nodes are matched or there is no energy storage unit that meets the matching conditions. Then the final mapping relationship of all nodes is counted, and the energy storage unit number corresponding to each node is renumbered to generate a mapping table. For example, the original number of node N005 corresponds to A101, and after re-arrangement, it is marked as number S01. Finally, according to the order of energy storage units matched to each node, the order in which the adjustment resources should be started on the timeline is arranged, and the time and sequence number of the energy storage unit scheduling actions are listed in sequence to obtain the dynamic load support sequence.

[0041] See also Figure 5 , the steps for obtaining the capacity balance instruction list are as follows: S411: Based on the dynamic load support sequence, analyze the current capacity utilization ratio and change rate of the energy storage unit, compare the change range of each parameter within the same cycle, and determine the difference between the two by comparing their growth or decline trends to obtain the capacity parameter comparison characteristics; Extract the numbers of all energy storage units that are currently in the adjustment task state, and obtain the current capacity utilization ratio of each energy storage unit, that is, convert the ratio of the current remaining power to the rated maximum capacity to obtain the utilization ratio in percentage. For example, if the current remaining power of a certain energy storage unit is 52 kWh and the rated maximum capacity is 100 kWh, the capacity utilization ratio is 52%. Then retrieve the capacity utilization ratio change record of the unit in the past two consecutive cycles, and calculate the change rate of the ratio in unit time. The change rate is calculated by dividing the capacity ratio difference between the two time points and the time interval. For example, if the previous cycle is 48% and the next cycle is 52%, and the time interval is 1 hour, the change rate is 4% per hour. Count the utilization ratio and change rate of all energy storage units in turn, and use these two parameters in the same cycle to calculate the change rate of the capacity utilization ratio. The absolute difference comparison is performed on the numbers. The comparison process subtracts the two parameters in each energy storage unit and records the difference. The larger the difference, the more obvious the growth or decrease trend. The above differences of all energy storage units are sorted from large to small to determine whether the trend difference is in the drastic change range. The threshold for the drastic range is set to 15%, the intermediate fluctuation range is 5%-15%, and the low-amplitude fluctuation range is less than 5%. Energy storage units with differences falling in different ranges are marked. When comparing the growth or decline direction, if the current utilization ratio is higher than the previous cycle and the change rate is positive, it is judged to be a growth trend, otherwise it is a downward trend. The above judgment results are sorted, and the four characteristic data of each energy storage unit, namely capacity ratio, change rate, difference and trend direction, are recorded. Finally, the data are combined to form the capacity parameter comparison feature.

[0042] S412: Based on the capacity parameter comparison characteristics, classify the energy storage units according to their change rates, optimize the classification results, classify the units with low change rates as charging targets, and classify the units with high change rates as discharging targets, and obtain charge and discharge group identifiers; The change rate values ​​of each energy storage unit are called and sorted by value. First, the division boundaries of the change rate are set into three intervals, among which the low rate is less than 2% per hour, the medium rate is 2% to 5% per hour, and the high rate is greater than 5% per hour. Then, all energy storage units are preliminarily classified according to this classification standard. Units with a change rate value less than 2% are classified into the low rate group, units with a change rate greater than 5% are classified into the high rate group, and the middle section is classified into the medium rate group. Then, the classification results are cross-checked. If a unit is in the low rate group but the capacity utilization ratio is higher than 80%, It is reclassified and marked as a standby unit instead of a charging unit. If a unit is in the high-rate group but the current capacity utilization ratio is less than 20%, it is marked as a discharge warning unit again. After optimizing the classification results, the initial screening set of charging and discharging objects is obtained. Subsequently, all units marked as abnormal or standby states are eliminated, and only energy storage units that meet the adjustment conditions are retained. The remaining low-rate units are classified into the charging object set, and the high-rate units are classified into the discharging object set. The current time period identifier, number identifier and classification identifier code are added to each member of the set to form a standardized charging and discharging grouping identifier.

[0043] S413: Analyze the current capacity utilization ratio, change rate, output power, and regional load response of each energy storage unit based on the charge and discharge group identifier, determine the charge and discharge task category and allocation relationship of each energy storage unit in the current cycle, and obtain a capacity balancing instruction list; Extract the capacity utilization ratio, change rate, output power and regional load response records of each energy storage unit one by one, evaluate whether the utilization ratio is in the range where discharge can continue (such as more than 60%) or must enter the charging protection range (such as less than 20%), compare the change rate with the change value of the previous cycle to confirm whether its state is stable, retrieve the output power, and verify whether it is within the rated release range, and compare it with the load response records of the area where it is located to analyze whether the regional load is in a gap or redundant state. For example, if the current load in a certain area is 23 kilowatts, the current output capacity of the energy storage unit is 6 kilowatts, and the regional gap is 4 kilowatts, then the energy storage unit can bear To bear part of the load, a comprehensive judgment is made based on the above four types of data parameters. If the utilization ratio of the unit is higher than 60%, the change rate is greater than 5%, the output power is in the middle and upper range, and the regional response is a load gap, it is determined to be a discharge task unit. On the contrary, if the utilization ratio is lower than 30%, the change rate is less than 2%, the current output is zero, and the regional load is in a redundant state, it is classified as a charging task unit. After determining the task category of each energy storage unit, a task category instruction code is generated and attached to the allocation record of the energy storage unit. At the same time, its service area number and scheduling time period are recorded. Finally, the task category, allocation area, response object and scheduling sequence of all energy storage units are summarized to form a capacity balance instruction list.

[0044] See also Figure 6 , the specific steps for obtaining the partition linkage execution results are: S511: Based on the capacity balancing instruction list, analyze the charging object number, discharging object number, and capacity balancing level, optimize the task category assignment for each energy storage unit, compare the area number and node distribution corresponding to each unit, determine the compatibility of the task category and area division, and obtain the energy storage task partition mapping list; Read the list of charging object numbers and discharging object numbers marked in the list, extract the basic parameters of the energy storage unit corresponding to each number in turn, including its rated capacity, current utilization ratio, change rate and area number, and then extract the capacity balance level parameter. The level is divided according to the difference between the remaining value of the energy storage unit capacity and the target balance ratio. For example, if the balance target is set to 60%, if the current capacity of a unit is 78%, then its deviation is 18%, which can be divided into three levels of capacity deviation. After completing the above basic data extraction, classify the current task of each energy storage unit according to the task category mark to determine whether it is a unit to be adjusted, adjusted or standby task unit, and then match the task category with the area number one by one. The node situations are expanded and listed, and compared with the energy storage unit numbers to check whether the current task category is suitable for the current node location. For example, if the discharge task is assigned to an area with high node density and frequent voltage fluctuations, it is suitable; otherwise, it is unsuitable. For unsuitable configurations, they are marked as task adjustment units. Then, the relationship between all current energy storage units and their task categories and the node areas they serve is reorganized. The units within the same task category are tried to be re-matched in a regional concentrated distribution manner. For example, the three discharge task units located in areas A1, A2, and A3 are reintegrated into the node in area A1 and the configurations of other areas are released. Finally, the new correspondence between task category, energy storage unit number, area number, and node number is output to generate an energy storage task partition mapping list.

[0045] S512: Based on the energy storage task partition mapping list, the matching relationship between the task category of each energy storage unit and the energy storage capacity of the region in which it is located is calculated, the adjustment sequence is optimized, and the correspondence between the region number and the task category is adjusted to obtain the energy storage task scheduling sorting identifier; The task category of each energy storage unit is compared and matched with the energy storage capacity in the area where it is located. First, the regional energy storage capacity data is extracted. This capacity is based on the sum of the dispatchable capacities of all energy storage units in the area. At the same time, the maximum dispatch demand value corresponding to each area is extracted to calculate the task type that the area can undertake. For example, if the total energy storage capacity in area A1 is 240 kWh and the dispatch demand is 80 kWh, the allocation ratio is 3.0, which is a high-load area and allows the configuration of multiple discharge task units. Compare the task categories currently assigned to the area to check whether they exceed the regional dispatch capacity boundary. For example, area A1 already has four discharge task units with a total discharge capacity of 120 kW. If the dispatch demand exceeds 80 kilowatts, it is determined to be over-allocated and marked as an optimized adjustment area. The allocation of all areas is then sorted, and the adjustment order of the energy storage units is rearranged according to the priority of regional energy storage capacity. High-capacity areas are preferentially assigned to discharge task units, and low-capacity areas are preferentially configured with charging task units. At the same time, the correspondence between task categories and regional numbers is adjusted. For example, unit number D101 is adjusted from area A1 to area B2. The current task load of area B2 is less than 50%, making it more suitable for charging tasks. A new mapping list is formed for all updated regional numbers and task categories, with a timestamp and task number. After sorting, the energy storage task scheduling sorting identifier is output.

[0046] S513: Calling the energy storage task scheduling sorting identifier, analyzing the adjustment command issued by the regional energy storage scheduling communication terminal, comparing the energy storage unit feedback signal with the task category action, screening the task completion status and collating the communication results to obtain the partition linkage execution result; The task number of each energy storage unit is sequentially matched with the dispatch communication terminal in its area. The communication terminal's adjustment command records for the current cycle are read. For each command record, the issuance time, target unit number, adjustment command type (charge or discharge), and target parameter value are extracted. The actual execution signal fed back by the energy storage unit is then compared. This signal includes the execution start time, execution power, duration, and status flag. The issued command and feedback signals are compared item by item to ensure consistency. For example, if the command requires unit number C206 to discharge at 6 kW for 30 minutes, if the feedback from C206 shows a discharge power of 6 kW, a discharge duration of 28 minutes, and a status of "completed", the execution is considered successful. If the feedback shows only 4 kW or the status is "interrupted", the execution is recorded as a failure. The completion rate of each task type is then calculated, and the ratio between the number of successful matches and the total number of tasks is filtered out. A completion status table is generated by task category. All communication records and feedback results are organized to generate a multidimensional data list containing the area number, energy storage unit number, task category, execution status, timestamp, and feedback indicators, thus obtaining the results of the partitioned linkage execution.

[0047] A multi-region energy storage device charging and discharging intelligent regulation system, the system comprising: The time series feature extraction module analyzes the power change trends in each time period of the day based on energy users in each region. It compares the correlation characteristics of the maximum load item in each time period with the cycle factor and holiday correction parameters, optimizes the adaptability of the cycle factor and correction parameters at different times, determines the stability of the continuous change of the cycle factor, screens out abnormal fluctuation periods and marks them as adjustment nodes, and obtains a time series abnormality tag set. The load anomaly identification module analyzes the difference between the load forecast data and the real-time collected data of each adjustment node based on the time series anomaly tag set, calculates the difference trend between consecutive time periods, compares the degree of match between the trend change and the regional load variation range, identifies nodes with trend changes greater than the range, and establishes trend anomaly warning features based on the early warning prompts of the regional node load monitoring equipment; The energy storage optimization ranking module optimizes the remaining power and maximum output power of energy storage units in the associated area based on trend anomaly warning characteristics, calculates the load support capacity index of each energy storage unit per unit time, compares the ranking of each energy storage unit's capacity index in the node, selects the energy storage unit with the best capacity index, and obtains the dynamic load support sequence; The intelligent capacity dispatch module analyzes the current capacity utilization ratio and change rate of energy storage units based on the dynamic load support sequence, determines the difference in the amplitude of the two parameters, optimizes the dynamic balance between the capacity utilization ratio and the change rate, identifies energy storage units with low change rates as charging targets and energy storage units with high change rates as discharging targets, and generates a capacity balance instruction list; Based on the capacity balance instruction list, the collaborative scheduling execution module adjusts the task category and regional allocation of each energy storage unit, optimizes the charging and discharging capacity of the energy storage units within the adjustment sequence, determines the energy release order and instruction execution period between each energy storage unit, and synchronously issues adjustment commands based on the regional energy storage scheduling communication terminal to obtain the partition linkage execution results.

[0048] The above are merely preferred embodiments of the present invention and do not limit the present invention in any other form. Any technician familiar with the profession may use the technical content disclosed above to change or modify it into an equivalent embodiment with equivalent changes and apply it to other fields. However, any simple modification, equivalent change and modification made to the above embodiment based on the technical essence of the present invention without departing from the content of the technical solution of the present invention shall still fall within the scope of protection of the technical solution of the present invention.

Claims

1. A multi-region energy storage device charging and discharging intelligent regulation method, characterized in that: The following steps are involved: S1: Based on the energy users in each region, analyze the power changes during daily periods, compare the load optimization items, cycle factors, and holiday correction parameters, determine the stability of the cycle factors, screen out abnormal fluctuation periods, mark them as adjustment nodes, and obtain the time series abnormality mark set; S2: Based on the time series anomaly mark set, analyze the difference between the load forecast and the real-time collected data of the adjustment node, determine the difference trend between consecutive time periods, and establish trend anomaly warning features in combination with the warning information of the load monitoring equipment; S3: Based on the abnormal trend warning characteristics, analyze the remaining power and output power of the energy storage units, calculate their load support capabilities, and select the energy storage units with the best capabilities after sorting to obtain a dynamic load support sequence; S4: Based on the dynamic load support sequence, analyze the capacity utilization ratio and change rate of the energy storage unit, determine the amplitude difference, identify the energy storage unit with a low change rate as the charging target and the energy storage unit with a high change rate as the discharging target, and obtain a capacity balancing instruction list.

2. The multi-region energy storage device charge and discharge intelligent regulation method according to claim 1, characterized in that: The timing anomaly mark set includes a time period label, an abnormal fluctuation level, and an adjustment reference index; the trend anomaly warning feature includes an offset trend factor, an abnormality type classification, and an warning level parameter; the dynamic load support sequence includes a support capacity ranking, an energy allocation identifier, and a node response attribute; and the capacity balance instruction list includes a charging object number, a discharging object number, and a capacity balance level.

3. The multi-region energy storage device charge and discharge intelligent regulation method according to claim 1, characterized in that: The steps for obtaining the time series anomaly marker set are specifically as follows: S111: Based on the energy users in each region, analyze the power change trend of the users in each time period of the day, compare the relationship between the maximum load item in each time period and the cycle factor and holiday correction parameter, determine the synchronization and difference in the change process, and obtain the correlation trend feature sequence; S112: Based on the correlation trend feature sequence, determine whether the change of the periodic factor in the continuous time period is stable, calculate the fluctuation of the periodic factor in each continuous time period, screen the time period with abnormal fluctuation, optimize the time period division, and obtain the periodic continuous change pattern; S113: Based on the periodic continuous change pattern, screen the time nodes with abnormal fluctuation amplitudes, compare the distribution and change characteristics of each node in the sequence, calculate the abnormal fluctuation aggregation index, combine the index with the node trajectory, and obtain a time series abnormality label set.

4. The multi-region energy storage device charge and discharge intelligent regulation method according to claim 1, characterized in that: The steps for obtaining the abnormal trend warning feature are specifically as follows: S211: Based on the time series anomaly marker set, analyze and adjust the changes in the load forecast data and the real-time collected data of the node, calculate the forecast offset development trend between consecutive time periods, compare the change rate of each node in the continuous time with the change of the regional load change limit, screen the trend change of each node, and obtain the trend change response sequence; S212: Based on the trend change response sequence, determine the trend growth of each node, analyze the relationship between the node trend change and the upper and lower bounds of the regional load change allowable range, filter out nodes whose trend changes exceed the allowable range, optimize node classification, and obtain a trend deviation node set; S213: Based on the trend deviation node set, compare the trend change of each node with the alarm status of the monitoring equipment, screen the nodes with key trend warning signal strength, judge the trend direction, diffusion speed and device response of the node, and establish trend abnormality warning characteristics.

5. The multi-region energy storage device charge and discharge intelligent regulation method according to claim 1, characterized in that: The steps for obtaining the dynamic load support sequence are specifically as follows: S311: Based on the abnormal trend warning characteristics, analyze the remaining power and sustainable output status of the regional energy storage unit, calculate its corresponding unit time support capacity and energy release limit range, adjust the parameter proportions and normalize them, and obtain the unit support capacity parameter; S312: Based on the unit support capacity parameters, compare the support differences of each unit at the abnormal node, calculate the regulation capacity strength of each energy storage unit, perform a sequence sorting operation, and obtain an energy storage unit sorting sequence; S313: Based on the energy storage unit sorting sequence, filter the units with priority, determine the matching order between them and each node, adjust the node mapping structure and rearrange the numbering correspondence, determine the dynamic linkage order of the regulation resources, and obtain the dynamic load support sequence.

6. The multi-region energy storage device charge and discharge intelligent regulation method according to claim 1, characterized in that: The steps for obtaining the capacity balancing instruction list are specifically as follows: S411: Based on the dynamic load support sequence, analyze the current capacity utilization ratio and change rate of the energy storage unit, compare the change range of each parameter within the same cycle, and determine the difference between the two by comparing their growth or decline trends to obtain a capacity parameter comparison feature; S412: Based on the capacity parameter comparison characteristics, classify the energy storage units according to their change rates, optimize the classification results, classify the units with low change rates as charging targets, and classify the units with high change rates as discharging targets, and obtain charge and discharge group identifiers; S413: Analyze the current capacity utilization ratio, change rate, output power, and regional load response of each energy storage unit based on the charge and discharge group identifier, determine the charge and discharge task category and allocation relationship of each energy storage unit in the current cycle, and obtain a capacity balancing instruction list.

7. The multi-region energy storage device charge and discharge intelligent regulation method according to claim 1, characterized in that: The steps also include: S5: Based on the capacity balancing instruction list, adjust the task category and regional allocation of each energy storage unit, optimize the charging and discharging capacity of the energy storage units in the adjustment sequence, determine the energy release order and instruction time period, issue adjustment commands, and obtain the partition linkage execution results; The partition linkage execution result includes a coordinated adjustment identifier, a regional energy allocation ratio, and an execution feedback signal.

8. The multi-region energy storage device charge and discharge intelligent regulation method according to claim 7, characterized in that: The steps for obtaining the partition linkage execution result are specifically as follows: S511: Based on the capacity balancing instruction list, analyze the charging object number, discharging object number, and capacity balancing level, optimize the task category allocation of each energy storage unit, compare the area number and node distribution corresponding to each unit, determine the adaptability of the task category and area division, and obtain the energy storage task partition mapping list; S512: Calculate the matching relationship between the task category of each energy storage unit and the energy storage capacity of the region in which it is located based on the energy storage task partition mapping list, optimize the adjustment order, adjust the correspondence between the region number and the task category, and obtain the energy storage task scheduling sorting identifier; S513: Call the energy storage task scheduling sorting identifier, analyze the adjustment command issued by the regional energy storage scheduling communication terminal, compare the energy storage unit feedback signal with the task category action, screen the task completion status and organize the communication results to obtain the partition linkage execution result.

9. A multi-region energy storage device charging and discharging intelligent regulation system, characterized in that: The system is used to implement the multi-region energy storage device charging and discharging intelligent regulation method according to any one of claims 1 to 8, and the system includes: The time series feature extraction module analyzes the power change trends in each time period of the day based on energy users in each region. It compares the correlation characteristics of the maximum load item in each time period with the cycle factor and holiday correction parameters, optimizes the adaptability of the cycle factor and correction parameters at different times, determines the stability of the continuous change of the cycle factor, screens out abnormal fluctuation periods and marks them as adjustment nodes, and obtains a time series abnormality tag set. The load anomaly identification module analyzes the difference between the load forecast data and the real-time collected data of each adjustment node based on the time series anomaly mark set, calculates the difference trend between consecutive time periods, compares the degree of matching between the trend change and the regional load variation range, identifies the nodes whose trend change is greater than the range, and establishes trend anomaly warning features in combination with the warning prompts of the regional node load monitoring equipment; The energy storage optimization sorting module optimizes the remaining power and maximum output power of the energy storage units in the associated area based on the abnormal trend warning characteristics, calculates the load support capacity index of each energy storage unit per unit time, compares the order of the capacity index of each energy storage unit in the node, selects the energy storage unit with the best capacity index, and obtains the dynamic load support sequence; The intelligent capacity dispatch module analyzes the current capacity utilization ratio and change rate of the energy storage unit based on the dynamic load support sequence, determines the difference between the two parameters, optimizes the dynamic balance between the capacity utilization ratio and the change rate, identifies the energy storage unit with a low change rate as the charging target and the energy storage unit with a high change rate as the discharging target, and obtains a capacity balance instruction list; Based on the capacity balancing instruction list, the collaborative scheduling execution module adjusts the task category and regional allocation of each energy storage unit, optimizes the charging and discharging capabilities of the energy storage units within the adjustment sequence, determines the energy release order and instruction execution period between the energy storage units, and synchronously issues adjustment commands based on the regional energy storage scheduling communication terminal to obtain the partition linkage execution results.

Images