A method and system for energy storage scheduling based on multi-source data analysis

Abstract

This application provides an energy storage scheduling method and system based on multi-source data analysis, relating to the field of energy storage control technology. The method involves acquiring charge / discharge depth records and ambient temperature records; extracting the capacity cycle segments of each energy storage unit from all charge / discharge depth records; performing capacity correction on each capacity cycle segment based on all ambient temperature records to generate the capacity response integral value of each energy storage unit; determining the energy storage contribution of each energy storage unit based on the capacity response integral values; when the energy storage contribution exceeds a preset equilibrium threshold, dividing the energy storage units into a high-response capacity group and a low-response capacity group; configuring dynamic charge / discharge depths for the high-response capacity group and the low-response capacity group based on the capacity response integral values, and periodically swapping the roles of the two groups. This application can adjust the differences in capacity contribution of energy storage units based on multi-source operating data to improve the stability of energy storage scheduling response.

Description

Technical Field

[0001] This application relates to the field of energy storage control technology, and more specifically, to an energy storage scheduling method and system based on multi-source data analysis. Background Technology

[0002] Energy storage control is the link in which energy storage power stations participate in grid dispatch and realize the spatiotemporal transfer and management of power. It is a key control system that connects the upper-level grid dispatch instructions with the execution actions of the lower-level energy storage units. Energy storage control covers functions such as energy storage unit charging and discharging power management, charging and discharging depth limit setting, multi-unit collaborative dispatch strategy execution, real-time monitoring of operating status and closed-loop adjustment. It is widely used in typical energy storage dispatch scenarios such as grid peak-valley arbitrage, frequency and voltage regulation, and new energy consumption. Its goal is to accurately respond to dispatch instructions and maximize the grid support capability of the energy storage system while ensuring the safe and stable operation of the energy storage system. It is the foundation for the implementation of various energy storage dispatch methods.

[0003] However, traditional energy storage dispatching based on multi-source data analysis lacks a quantifiable assessment of capacity contribution and an adaptive grouping and control system. It also fails to establish a differentiated dynamic charge / discharge depth configuration and aging-balanced driving role-switching mechanism adapted to the individual operating characteristics of energy storage units. This results in severely uneven cycle life loss among energy storage units during dispatching, easily leading to premature failure due to excessive aging of some units. Therefore, how to regulate the differences in capacity contribution of energy storage units based on multi-source operating data to improve the stability of energy storage dispatching response is a problem facing the industry. Summary of the Invention

[0004] This application provides an energy storage scheduling method and system based on multi-source data analysis, which can adjust the differences in capacity contribution of energy storage units based on multi-source operating data to improve the stability of energy storage scheduling response.

[0005] In a first aspect, this application provides an energy storage scheduling method based on multi-source data analysis, the energy storage scheduling method comprising the following steps: Acquire the charge / discharge depth records and ambient temperature records for each energy storage unit during energy storage scheduling; The capacity cycle segment of each energy storage unit is extracted from all charge and discharge depth records, and the capacity of each capacity cycle segment is corrected according to all ambient temperature records to generate the capacity response integral value of each energy storage unit. The energy storage contribution of each energy storage unit is determined based on the capacity response integral value of all energy storage units. When the energy storage contribution exceeds a preset equalization threshold, the energy storage units are divided into a high response capacity group and a low response capacity group. Based on the capacity response integral value, dynamic charge and discharge depths are configured for the high-response capacity group and the low-response capacity group, and the roles of the two groups are periodically swapped.

[0006] In this embodiment, extracting the capacity cycle segment of each energy storage unit from all charge / discharge depth records specifically includes: Extract the capacity cycle boundary of each energy storage unit from all charge / discharge depth records; An adaptive phase segmentation window is constructed based on all capacity cycle boundaries. The charge-discharge depth curve is segmented into cycle segments characterized by the state-of-charge turning point through local integral charge matching. The capacity cycle segment of each energy storage unit is determined based on all cycle segments.

[0007] In this embodiment, capacity correction is performed on each capacity cycle segment based on all ambient temperature records to generate the capacity response integral value of each energy storage unit, specifically including: Align all ambient temperature records with the time axis of all capacity cycling segments to construct a thermally induced decay coefficient field and extract local fluctuation entropy. Extract the capacity degradation component of each energy storage unit from the thermally induced degradation coefficient field; The capacity response curve of each energy storage unit is determined based on all capacity degradation components and the local fluctuation entropy. The capacity response curve is converted to capacity to obtain the integral value of the capacity response of each energy storage unit.

[0008] In this embodiment, aligning all ambient temperature records with the time axis of all capacity cycling segments to construct a thermally induced decay coefficient field and extract local fluctuation entropy specifically includes: Bidirectional timestamp interpolation is performed on the time axis of all ambient temperature records and all capacity cycle segments to generate synchronized temperature sequences and synchronized capacity sequences; Based on the synchronized temperature sequence, a thermally induced decay coefficient field is constructed by mapping the capacity decay rate point by point. Local fluctuation entropy is extracted within a local sliding window of the thermally induced decay coefficient field.

[0009] In this embodiment, determining the energy storage contribution of a corresponding energy storage unit based on the integral value of the capacity response of all energy storage units specifically includes: Construct an initial contribution matrix based on the capacity response integral values ​​of all energy storage units, using weighted state of charge and corrected charge / discharge efficiency. Based on the initial contribution matrix, the energy storage contribution allocation tensor is determined by introducing voltage balance constraints and power response priorities among energy storage units. Extract the contribution coefficient of each energy storage unit from the energy storage contribution allocation tensor, and output the energy storage contribution of the corresponding energy storage unit.

[0010] In this embodiment, when the energy storage contribution exceeds a preset balancing threshold, dividing the energy storage units into a high-response capacity group and a low-response capacity group specifically includes: The energy storage contribution is compared with a preset equilibrium threshold to generate a contribution residual sequence. Based on the contribution residual sequence, high-response initial groups and low-response initial groups of energy storage units are constructed; Power coupling constraints and state of charge consistency checks are introduced between energy storage units, and neighborhood exchange checks are performed on the high-response initial group and the low-response initial group. Based on the results of the neighborhood exchange verification, the energy storage units are divided into high-response capacity group and low-response capacity group.

[0011] In this embodiment, power coupling constraints and state of charge consistency checks are introduced between energy storage units. The neighborhood exchange check for the high-response initial group and the low-response initial group specifically includes: Based on the high-response initial grouping and the low-response initial grouping, the power coupling strength matrix and state of charge deviation between adjacent energy storage units are determined; A bipartite graph minimum weight matching model is constructed using the power coupling strength matrix and the state of charge deviation, and candidate exchange unit pairs that satisfy the power coupling constraints and the state of charge consistency verification are determined. Based on the candidate exchange units, perform iterative neighborhood exchange verification, update the group labels until the cost converges, and output the neighborhood exchange verification results.

[0012] In this embodiment, configuring dynamic charge / discharge depths for the high-response capacity group and the low-response capacity group based on the capacity response integral value specifically includes: Based on the capacity response integral value, the initial reference charge / discharge depths corresponding to the high response capacity group and the low response capacity group are calculated respectively, wherein the reference depth of the high response capacity group is greater than the reference depth of the low response capacity group. Based on the state of charge range and cycle life decay rate of each energy storage unit in each group, a depth adaptive correction factor is constructed to adjust the initial reference charge and discharge depth on a unit-by-unit basis. Output the dynamic charge / discharge depth of each energy storage unit in the high-response capacity group and the low-response capacity group.

[0013] In this embodiment, periodically exchanging the two sets of roles specifically includes: The aging balance of each group is determined based on the cumulative capacity throughput and health status decay rate of the high response capacity group and the low response capacity group. When the corresponding aging balance exceeds the preset threshold, swap the two sets of role labels and reset the cumulative counter, outputting the updated high response capacity group and low response capacity group.

[0014] Secondly, this application provides an energy storage scheduling system based on multi-source data analysis, used to execute an energy storage scheduling method based on multi-source data analysis, the energy storage scheduling system comprising: The acquisition module is used to acquire the charge / discharge depth records and ambient temperature records corresponding to each energy storage unit during energy storage scheduling; The correction module is used to extract the capacity cycle segment of each energy storage unit from all charge and discharge depth records, and to perform capacity correction on each capacity cycle segment based on all ambient temperature records, thereby generating the capacity response integral value of each energy storage unit. The capacity partitioning module is used to determine the energy storage contribution of the corresponding energy storage unit based on the capacity response integral value of all energy storage units. When the energy storage contribution exceeds the preset equalization threshold, the energy storage units are divided into a high response capacity group and a low response capacity group. The capacity configuration module is used to configure dynamic charge and discharge depths for the high-response capacity group and the low-response capacity group respectively based on the capacity response integral value, and periodically switch roles between the two groups.

[0015] The technical solutions provided by the embodiments disclosed in this application have the following beneficial effects: The system acquires charge / discharge depth records and ambient temperature records for each energy storage unit during energy storage scheduling. It extracts the capacity cycle segment of each energy storage unit from all charge / discharge depth records and performs capacity correction on each capacity cycle segment based on all ambient temperature records, generating the capacity response integral value for each energy storage unit. Based on the capacity response integral values ​​of all energy storage units, it determines the energy storage contribution of the corresponding energy storage unit. When the energy storage contribution exceeds a preset equilibrium threshold, the energy storage units are divided into a high-response capacity group and a low-response capacity group. Based on the capacity response integral values, it configures dynamic charge / discharge depths for the high-response capacity group and the low-response capacity group respectively, and periodically swaps the roles of the two groups.

[0016] Therefore, in this application, dynamic charge / discharge depths are configured for the high-response capacity group and the low-response capacity group based on the capacity response integral value, and the roles of the two groups are periodically exchanged. The generation of the capacity response integral value provides a normalized quantification benchmark for the effective capacity contribution of each energy storage unit within the scheduling cycle, thereby eliminating the interference of ambient temperature fluctuations, differences in charge / discharge efficiency, and cyclic aging losses on the quantification of capacity contribution. This accurately characterizes the actual response capability and effective capacity support level of each energy storage unit to scheduling commands. This quantification benchmark enables accurate identification and unbiased comparison of the capacity contribution differences among energy storage units, avoiding the problem of disconnect between traditional scheduling strategies based on rated capacity and the actual operating performance of the units, and providing a unified standard for differentiated scheduling and management. A compliant quantitative basis ensures the consistency and long-term stability of energy storage dispatch response. Dividing into high-response capacity groups and low-response capacity groups provides a differentiated dispatch and control platform adapted to the individual response characteristics of energy storage units. This enables hierarchical control and targeted adaptation of charging and discharging strategies for energy storage units with different contribution levels. It avoids the problems of uneven lifespan loss and increased differentiation in response capabilities among units caused by traditional unified dispatch strategies. Through grouped control, targeted regulation of the capacity contribution differences of energy storage units can be achieved. High-response units ensure rapid and accurate response to dispatch commands, while low-response units provide basic capacity support and lifespan buffering. This achieves synergy between dispatch response performance and overall system lifespan balance, continuously improving the long-term stability of energy storage dispatch response.

[0017] In summary, the technical solution adopted in this application can adjust the differences in capacity contribution of energy storage units based on multi-source operation data, so as to improve the stability of energy storage dispatch response. Attached Figure Description

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

[0019] Figure 1 This is an exemplary flowchart of an energy storage scheduling method based on multi-source data analysis provided in this application; Figure 2 This is a flowchart illustrating the process of generating capacity response integral values ​​provided in this application; Figure 3 This is a flowchart illustrating the division of high response capacity groups and low response capacity groups provided in this application; Figure 4 This is a schematic diagram of the dynamic charging and discharging depth configuration of the energy storage unit and the principle of periodic exchange of two sets of roles provided in this application; Figure 5This is a module structure diagram of an energy storage scheduling system based on multi-source data analysis provided in this application. Detailed Implementation

[0020] The technical solutions of the embodiments of this application will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of this application, and not all embodiments. Based on the embodiments of this application, all other embodiments obtained by those of ordinary skill in the art without creative effort are within the scope of protection of this application.

[0021] This application provides an energy storage scheduling method and system based on multi-source data analysis. The core of this method is to acquire charge / discharge depth records and ambient temperature records corresponding to each energy storage unit during energy storage scheduling; extract the capacity cycle segment of each energy storage unit from all charge / discharge depth records, and perform capacity correction on each capacity cycle segment based on all ambient temperature records to generate the capacity response integral value of each energy storage unit; determine the energy storage contribution of the corresponding energy storage unit based on the capacity response integral values ​​of all energy storage units; when the energy storage contribution exceeds a preset equilibrium threshold, divide the energy storage units into a high-response capacity group and a low-response capacity group; configure dynamic charge / discharge depths for the high-response capacity group and the low-response capacity group according to the capacity response integral values, and periodically exchange roles between the two groups.

[0022] Example 1: To better understand the above technical solution, the following will provide a detailed description of the technical solution in conjunction with the accompanying drawings and specific implementation methods. (Refer to...) Figure 1 As shown in the figure, this is an exemplary flowchart of an energy storage scheduling method based on multi-source data analysis according to this embodiment of the present application. The energy storage scheduling method includes the following steps: In step S1, the charge / discharge depth record and ambient temperature record corresponding to each energy storage unit are obtained during energy storage scheduling.

[0023] In practical implementation, a time-series synchronous acquisition framework is built based on the globally unified clock used for energy storage scheduling. The acquisition cycle is adapted according to the electrochemical type of the energy storage unit: 100ms for lithium battery units and 500ms for lead-carbon battery units, matching the dynamic response characteristics of the units. A dual-channel synchronous acquisition module is deployed on the local controller of each energy storage unit: The first channel collects the terminal voltage and charging / discharging current in real time through a current sensor connected in series with the unit's main circuit and a voltage sensor connected in parallel. The charging / discharging depth is calculated in real time using the ampere-hour integration method, and a time-series record is generated by binding a global timestamp. The second channel deploys a high-precision temperature sensor at the unit's installation location to synchronously collect ambient temperature data that is completely aligned with the charging / discharging timeline. The corresponding record is generated by binding the same timestamp, and two levels of verification are set up for completeness and rationality. Abnormal data with mismatched timestamps or excessive data jumps are removed. The valid records are uploaded to the system's central database to complete the acquisition of the target data.

[0024] It should be noted that, in this application, an energy storage unit refers to the smallest controllable single unit that independently performs energy storage and release in an energy storage scheduling system; a charge / discharge depth record refers to a dataset that records the temporal changes in the charge / discharge depth of an energy storage unit; and an ambient temperature record refers to a temperature dataset that is synchronized with the time sequence of the charge / discharge process of the energy storage unit.

[0025] In step S2, the capacity cycle segment of each energy storage unit is extracted from all charge and discharge depth records, and the capacity of each capacity cycle segment is corrected according to all ambient temperature records to generate the capacity response integral value of each energy storage unit.

[0026] In this embodiment, the extraction of the capacity cycle segment of each energy storage unit from all charge-discharge depth records can be achieved using the following steps: Extract the capacity cycle boundary of each energy storage unit from all charge / discharge depth records; An adaptive phase segmentation window is constructed based on all capacity cycle boundaries. The charge-discharge depth curve is segmented into cycle segments characterized by the state-of-charge turning point through local integral charge matching. The capacity cycle segment of each energy storage unit is determined based on all cycle segments.

[0027] In practice, firstly, all charge and discharge depth records of a single energy storage unit are arranged in ascending order according to the global timestamp to generate a continuous charge and discharge depth time-series change curve. Using the sliding window extreme value identification method, the window width is set to 20 sampling points and the sliding step size is 1 sampling point. The entire curve is completely traversed along the time axis to identify the local maximum value point where the charge and discharge depth changes from rising to falling and the local minimum value point where the charge and discharge depth changes from falling to rising. The start and end nodes of the minimum-maximum-minimum combination of adjacent minimum values ​​with a total charge throughput difference not exceeding 2% of the rated capacity are marked as the capacity cycle boundary of the energy storage unit. Then, using the extracted capacity cycle boundaries as start and end anchor points, an adaptive phase segmentation window is constructed. The initial window width is the time span of adjacent capacity cycle boundaries, and the sliding step size is set to a single data acquisition cycle. The window width is dynamically adjusted according to the curve slope changes. Where the slope changes abruptly, the window width is reduced to half of the original width, while the original window width is maintained where the slope is gentle. During the traversal, the ampere-hour integration method is used to calculate the total charge and discharge amount within each window. When the difference between the two does not exceed 3% of the maximum charge amount within the window, the curve within the window is determined to be the target cycle segment, and the curve segmentation is completed. Finally, all cycle segments of a single energy storage unit are sorted in chronological order. Invalid segments with a duration less than the minimum rated charge and discharge cycle of the energy storage unit and a total charge throughput less than 5% of the unit's rated capacity are first removed. Valid cycle segments that meet the requirements of a complete charge and discharge cycle are retained. For each valid cycle segment, the minimum and maximum values ​​of the charge and discharge depth within the segment are extracted. The interval between the minimum and maximum values ​​is determined as the corresponding capacity cycle segment, and the time series data of the corresponding time period is bound to it, thus completing the determination of all capacity cycle segments.

[0028] It should be noted that, in this application, the capacity cycle boundary refers to the characteristic node that defines the start and end positions of a single complete charge-discharge cycle of the energy storage unit; the adaptive phase segmentation window refers to the time-series data segmentation window that dynamically adjusts the window width and sliding step size according to the changing characteristics of the charge-discharge depth curve; local integral charge matching refers to the method of verifying the charge throughput equivalence by comparing the ampere-hour integral results of charge and discharge within the time-series window; the charge-discharge depth curve refers to the curve plotted as the ratio of the current charged / discharged amount of energy to its rated available capacity; the state of charge inflection point refers to the extreme point in the charge-discharge depth curve where the charge-discharge state switches; the cycle segment refers to the curve segment corresponding to a single complete charge-discharge process with the state of charge inflection point as the start and end boundary; and the capacity cycle segment refers to the charge-discharge depth interval corresponding to a single complete charge-discharge cycle of the energy storage unit.

[0029] Preferably, in this embodiment, capacity correction is performed on each capacity cycle segment based on all ambient temperature records to generate the capacity response integral value of each energy storage unit, with reference to... Figure 2As shown in the figure, this is a schematic flowchart of generating capacity response integral values ​​in some embodiments of this application. In this embodiment, generating capacity response integral values ​​can be achieved by the following steps: In step S21, all ambient temperature records are aligned with the time axis of all capacity cycling segments to construct a thermal decay coefficient field and extract local fluctuation entropy. In step S22, the capacity degradation component of each energy storage unit is extracted from the thermally induced decay coefficient field; In step S23, the capacity response curve of each energy storage unit is determined based on all capacity degradation components and the local fluctuation entropy; In step S24, the capacity response curve is converted to capacity to obtain the integral value of the capacity response of each energy storage unit.

[0030] In practical implementation, firstly, using a globally unified clock as a reference, the start and end timestamps of each capacity cycle segment are mapped and aligned with the corresponding ambient temperature records to ensure a complete match between the time sequence nodes and temperature data for each segment. Based on the industry-known capacity-temperature characteristic model calibrated at the factory of the energy storage unit, the capacity thermal degradation coefficient corresponding to the temperature at each time node is calculated. A continuous thermal degradation coefficient field is constructed with time as the horizontal axis and degradation coefficient as the vertical axis. Using the Shannon entropy calculation method, with a single capacity cycle segment as the calculation unit, the local fluctuation entropy of temperature fluctuations within the segment is calculated. Next, for each capacity cycle segment of a single energy storage unit, the time sequence of the thermal degradation coefficient corresponding to the entire time period of that segment is extracted from the thermal degradation coefficient field. The average thermal degradation coefficient of that segment is calculated using the arithmetic mean method. Using the rated capacity of the energy storage unit at a standard reference temperature of 25℃ as the benchmark value, the benchmark value is multiplied by the average thermal degradation coefficient to obtain the capacity degradation component caused by ambient temperature in that segment. This completes the extraction of the capacity degradation components corresponding to all segments. Then, for each capacity cycle segment of a single energy storage unit, the measured charge / discharge capacity of that segment is used as the base value. The capacity degradation component of the corresponding segment is subtracted to obtain the effective capacity value corrected for temperature mean. Then, the local fluctuation entropy of that capacity cycle segment is used as the correction weight to perform a second dynamic correction on the effective capacity value, eliminating calculation errors caused by random temperature fluctuations, resulting in the final corrected capacity value for that capacity cycle segment. A continuous time-series curve is plotted with the global time axis as the horizontal axis and the corrected capacity values ​​for each time period as the vertical axis; this is the capacity response curve of the energy storage unit. Finally, a fixed statistical period matching the energy storage dispatch demand is preset. For the capacity response curve of a single energy storage unit, the Newton-Leibniz definite integral calculation method is used. The start and end times of the statistical period are used as the upper and lower limits of integration. The effective capacity values ​​of the capacity response curve within the integration interval are continuously accumulated and integrated. During the integration process, the values ​​of invalid capacity segments triggered by non-dispatch commands are removed. The final integral result is the capacity response integral value of the energy storage unit in the current statistical period.

[0031] It should be noted that in this application, capacity correction refers to the process of eliminating the influence of ambient temperature on the actual usable capacity of the energy storage unit and converting the measured capacity into an equivalent capacity under a unified standard; thermal degradation coefficient field refers to the time-series distribution dataset of the degree of influence of ambient temperature on the capacity degradation of the energy storage unit; local fluctuation entropy refers to the quantitative index of the degree of disorder of ambient temperature fluctuation within a single capacity cycle segment; capacity degradation component refers to the amount of reduction in the actual usable capacity of the energy storage unit relative to the standard rated capacity caused by ambient temperature; capacity response curve refers to the time-series curve of the actual effective capacity throughput of the energy storage unit after temperature correction changing continuously over time; capacity conversion refers to the process of normalizing and integrating the dynamic capacity response value after temperature correction according to a unified benchmark to form a standard quantitative value for contribution assessment; capacity response integral value refers to the index of the total effective capacity throughput contribution of the energy storage unit after temperature correction within the statistical period.

[0032] In addition, in this embodiment, aligning all ambient temperature records with the time axis of all capacity cycling segments to construct the thermally induced decay coefficient field and extract local fluctuation entropy can be achieved through the following steps: Bidirectional timestamp interpolation is performed on the time axis of all ambient temperature records and all capacity cycle segments to generate synchronized temperature sequences and synchronized capacity sequences; Based on the synchronized temperature sequence, a thermally induced decay coefficient field is constructed by mapping the capacity decay rate point by point. Local fluctuation entropy is extracted within a local sliding window of the thermally induced decay coefficient field.

[0033] In practice, firstly, using the timestamp of the capacity cycling segment as the target time axis, the original timestamps of the ambient temperature records are traversed. A linear interpolation method is used to interpolate the temperature data to ensure that its time points are completely consistent with the time points of the capacity cycling segment. Simultaneously, using the original timestamp of the ambient temperature as a reference, the capacity data is subjected to reverse linear interpolation to ensure that the two sets of data have the same number and position of time points on the same time axis. Through bidirectional interpolation processing, synchronized temperature sequences and synchronized capacity sequences with completely overlapping time axes are generated. Then, the temperature-capacity decay characteristic curve in the energy storage field is used. This curve is obtained from the battery's factory standard test. Each temperature value in the synchronized temperature sequence is substituted into the temperature-capacity decay characteristic function, and the capacity decay rate at the corresponding temperature is calculated point by point. That is, the ratio of the current temperature capacity to the standard temperature capacity. The capacity decay rates of all time points are arranged in chronological order to form a continuous distribution data set with time as the horizontal axis and decay coefficient as the vertical axis. This set is the thermally induced decay coefficient field. Finally, a local sliding window of fixed length is set on the thermal decay coefficient field and slides point by point along the time axis. For the thermal decay coefficient data in each sliding window, the normalization process is first performed to map the coefficient value to the probability distribution interval. Then, the Shannon entropy calculation formula is used to substitute the probability value of each data point in the sliding window into the calculation to obtain the local fluctuation entropy that characterizes the degree of data fluctuation disorder. The entire thermal decay coefficient field is traversed in turn to obtain the local fluctuation entropy corresponding to each time position.

[0034] It should be noted that, in this application, bidirectional timestamp interpolation refers to a processing method that ensures that environmental temperature records with different sampling frequencies and time sequences correspond completely to the capacity cycle segments on the time axis, eliminating time sequence misalignment and achieving data synchronization; synchronized temperature sequence refers to a set of temperature data that corresponds one-to-one with the time points of the capacity sequence; synchronized capacity sequence refers to a set of capacity data that corresponds one-to-one with the time points of the temperature sequence; point-by-point mapping capacity decay rate refers to the process of converting the temperature of each time point in the synchronized temperature sequence into the capacity decay ratio at the corresponding time; thermal decay coefficient field refers to a set of capacity decay coefficients distributed point-by-point with time and temperature as dimensions; local sliding window refers to a fixed-length time sequence interval that slides along the time axis on the thermal decay coefficient field, used for local feature extraction and calculation of local fluctuation entropy analysis region.

[0035] In step S3, the energy storage contribution of the corresponding energy storage unit is determined based on the capacity response integral value of all energy storage units. When the energy storage contribution exceeds the preset equalization threshold, the energy storage units are divided into a high response capacity group and a low response capacity group.

[0036] In this embodiment, determining the energy storage contribution of a corresponding energy storage unit based on the integral value of the capacity response of all energy storage units can be achieved through the following steps: Construct an initial contribution matrix based on the capacity response integral values ​​of all energy storage units, using weighted state of charge and corrected charge / discharge efficiency. Based on the initial contribution matrix, the energy storage contribution allocation tensor is determined by introducing voltage balance constraints and power response priorities among energy storage units. Extract the contribution coefficient of each energy storage unit from the energy storage contribution allocation tensor, and output the energy storage contribution of the corresponding energy storage unit.

[0037] In practice, firstly, the capacity response integral value of each energy storage unit is obtained, and at the same time, the real-time state of charge data of each unit is collected. Using a linear weighting method, the state of charge value is mapped to a weight coefficient in the 0-1 interval. The closer the state of charge is to the ideal interval, the closer the weight coefficient is to 1. Using the charge and discharge efficiency test data of the energy storage unit, the ratio of the actual efficiency to the rated efficiency is calculated as the efficiency correction coefficient. The capacity response integral value of each unit is multiplied by the corresponding state of charge weight coefficient, and then multiplied by the efficiency correction coefficient to obtain the initial contribution value of a single unit. The initial contribution values ​​of all units are arranged in rows and columns to construct an initial contribution value matrix. Then, the maximum allowable range of voltage deviation between energy storage units is set as a voltage balancing constraint. When the voltage deviation between units exceeds this range, the contribution allocation ratio of high-voltage units is reduced and the ratio of low-voltage units is increased. Based on parameters such as the rated power and remaining cycle life of the energy storage units, the power response priority of each unit is determined using the analytic hierarchy process (AHP). The higher the priority, the larger the allocation weight coefficient. The initial contribution matrix is ​​fused with the voltage balancing constraint parameters and the power response priority weight coefficient to construct a three-dimensional energy storage contribution allocation tensor, where the dimensions correspond to the energy storage unit number, constraint type, and contribution quantification value, respectively. Finally, the energy storage contribution allocation tensor is dimensionality reduced, the constraint type dimension is fixed, and the contribution quantification value of each energy storage unit under each constraint condition is extracted. The arithmetic mean method is used to calculate the comprehensive quantification value of each unit. The comprehensive quantification values ​​of all energy storage units are normalized to obtain the contribution coefficient of each unit. The sum of the coefficients is 1. The contribution coefficient is multiplied by 100% and converted into a percentage form, which is the energy storage contribution of the corresponding energy storage unit, and output according to the unit number.

[0038] It should be noted that, in this application, "weighted state of charge" refers to the processing method of adjusting the weight of the capacity response integral value based on the real-time state of charge of the energy storage unit; "charge and discharge efficiency correction" refers to the correction process to eliminate the influence of energy loss during charging and discharging of the energy storage unit on the contribution calculation; "initial contribution matrix" refers to a two-dimensional data matrix that integrates the capacity response integral value, the state of charge weight, and the charge and discharge efficiency correction; "voltage balance constraint" refers to the constraint conditions that limit the voltage deviation between energy storage units and ensure safe operation; "power response priority" refers to the order rule of response scheduling instructions set according to the performance parameters of the energy storage unit; "energy storage contribution allocation tensor" refers to a three-dimensional quantitative data set formed after integrating multiple constraints; "contribution coefficient" refers to the parameter of the contribution ratio of a single energy storage unit; and "energy storage contribution" refers to the evaluation index of the actual effective capacity contribution ratio of a single energy storage unit in the overall scheduling.

[0039] Preferably, in this embodiment, when the energy storage contribution exceeds a preset balancing threshold, the energy storage units are divided into a high-response capacity group and a low-response capacity group, with reference to... Figure 3 As shown in the figure, this is a schematic diagram of the process of dividing the high response capacity group and the low response capacity group in some embodiments of this application. In this embodiment, the division of the high response capacity group and the low response capacity group can be achieved by the following steps: In step S31, the energy storage contribution is compared with a preset equilibrium threshold to generate a contribution residual sequence; In step S32, high-response initial groups and low-response initial groups of energy storage units are constructed based on the contribution residual sequence; In step S33, power coupling constraints and state of charge consistency checks between energy storage units are introduced, and neighborhood exchange checks are performed on the high-response initial group and the low-response initial group. In step S34, the energy storage units are divided into high response capacity group and low response capacity group according to the neighborhood exchange verification results.

[0040] In practice, firstly, a preset equilibrium threshold is established based on the average energy storage contribution of all effective energy storage units. A floating range is set in conjunction with the system lifetime equilibrium target to determine the final threshold value. The energy storage contribution of each energy storage unit is then calculated by subtracting it from the preset equilibrium threshold to obtain the contribution residual for each unit. A positive residual indicates that the contribution exceeds the threshold, while a negative residual indicates that the contribution is below the threshold. The contribution residuals of all energy storage units are arranged in a fixed order according to their unit numbers, generating an ordered sequence of contribution residuals. Next, all residual data in the contribution residual sequence are traversed, and each residual is matched to its corresponding energy storage unit. Energy storage units with residual values ​​greater than 0 are all assigned to the high-response initial group; energy storage units with residual values ​​less than or equal to 0 are all assigned to the low-response initial group. After grouping, the number of energy storage units, total rated capacity, and total power capability within each initial group are statistically analyzed to form the basic attribute set of the two initial groups, thus completing the initial grouping construction. Then, a minimum total power threshold required for system scheduling is preset as a power coupling constraint, and the maximum allowable deviation of the state of charge (SOC) of energy storage units within a group is preset as a consistency verification standard. Several energy storage units with the smallest absolute residual values ​​at the boundaries of two initial groups are selected as units to be exchanged. After the units to be exchanged are exchanged across groups, it is verified whether the total system power after grouping meets the power coupling constraint and whether the SOC within the group passes the consistency verification. Exchange schemes that meet the constraint requirements are recorded. Finally, from all exchange schemes that meet the constraint requirements in the neighboring exchange verification record, the optimal exchange scheme with the smallest sum of squared residuals of the unit contribution of the two groups after grouping is selected. The cross-group adjustment of boundary energy storage units is completed according to the optimal scheme. After the adjustment is completed, the optimized set of high-response units is determined as the high-response capacity group, and the optimized set of low-response units is determined as the low-response capacity group. The unit composition and grouping attributes of the two groups are locked, and the final grouping is completed.

[0041] It should be noted that, in this application, the preset equilibrium threshold refers to the benchmark judgment boundary used to classify the contribution level of energy storage units and distinguish between high-response and low-response units; the contribution residual sequence refers to the ordered data set of the degree of deviation between the energy storage contribution of each energy storage unit and the preset equilibrium threshold; the high-response initial grouping refers to the set of energy storage units whose contribution meets the high-response judgment requirements after preliminary screening; the low-response initial grouping refers to the set of energy storage units whose contribution meets the low-response judgment requirements after preliminary screening; the power coupling constraint refers to the constraint condition that limits the total power output capability of energy storage units after grouping to meet the scheduling requirements; the state of charge consistency verification refers to the verification process of verifying that the state of charge deviation of energy storage units within the group is within the safe allowable range; the neighborhood exchange verification refers to the grouping optimization process of performing cross-group exchange verification of energy storage units at the initial grouping boundary; the high-response capacity group refers to the set of energy storage units undertaking high-intensity charge and discharge scheduling tasks; and the low-response capacity group refers to the set of energy storage units undertaking low-intensity charge and discharge scheduling tasks.

[0042] In addition, in this embodiment, power coupling constraints and state of charge consistency checks are introduced between energy storage units. The neighborhood exchange check for the high-response initial group and the low-response initial group can be implemented by the following steps: Based on the high-response initial grouping and the low-response initial grouping, the power coupling strength matrix and state of charge deviation between adjacent energy storage units are determined; A bipartite graph minimum weight matching model is constructed using the power coupling strength matrix and the state of charge deviation, and candidate exchange unit pairs that satisfy the power coupling constraints and the state of charge consistency verification are determined. Based on the candidate exchange units, perform iterative neighborhood exchange verification, update the group labels until the cost converges, and output the neighborhood exchange verification results.

[0043] In specific implementation, firstly, using the boundary cells of the high-response initial group and the low-response initial group as a benchmark, all adjacent energy storage units connected in parallel on the electrical topology are paired. An internal resistance power coupling calculation method is used to collect the DC internal resistance and rated power parameters of each unit, calculating the power coupling strength between each pair of adjacent units. The value range is 0 to 1. A power coupling strength matrix is ​​constructed by arranging all paired coupling strengths in rows and columns. Simultaneously, the absolute value of the difference in state of charge (SOC) between each pair of units is calculated to obtain the corresponding SOC deviation. Then, using the boundary cells of the high-response initial group as the left vertex set of the bipartite graph and the boundary cells of the low-response initial group as the right vertex set, a minimum weighted matching model of the bipartite graph is constructed using the weighted sum of the power coupling strength and SOC deviation of the unit pair as the edge weight. The weighting coefficients are set according to the constraint priority. The Kuhn-Munkres algorithm is used to solve the model to obtain the initial matched unit pairs. The system verifies whether the unit pairs satisfy the power coupling constraints and SOC consistency checks after exchange, retaining the pairs that meet the requirements as candidate exchange unit pairs. Finally, a target cost function based on the sum of squared residuals of the energy storage contribution of the two groups of units and a cost convergence threshold are preset. Cross-group exchanges are performed sequentially according to the candidate exchange unit pairs, and the grouping labels of the corresponding units are updated synchronously. After each exchange, the cost value of the current grouping scheme is calculated, and the decrease rate is judged by comparing it with the cost value of the previous round. Iterative exchanges and cost calculations are repeated until the difference between the cost values ​​of two consecutive rounds is less than the preset convergence threshold. The cost convergence is determined, the iteration is terminated, and the final optimized grouping scheme is output.

[0044] It should be noted that in this application, the power coupling strength matrix refers to a set of quantified data on the degree of mutual influence of power output between adjacent energy storage units; the state of charge deviation refers to a quantified parameter on the difference in the state of charge values ​​between different energy storage units; the bipartite graph minimum weight matching model refers to a graph theory optimization model used to match the optimal cross-group exchange unit pairs; the candidate exchange unit pair refers to a set of energy storage unit pairs belonging to two initial groups that meet the cross-group exchange conditions after constraint verification; the iterative neighborhood exchange verification refers to the grouping process of iteratively executing cross-group exchanges of units and verifying constraints; the group label refers to the attribute mark that identifies the high-response or low-response group to which the energy storage unit belongs; cost convergence is the state that characterizes the objective cost function of grouping optimization decreasing to a stable threshold, and is the core judgment condition for terminating iterative optimization; the neighborhood exchange verification result refers to the judgment basis used to regulate the adjustment of energy storage unit group boundaries and ensure that the groups meet the safety constraints and lifetime balance objectives.

[0045] Additionally, it should be noted that in this application, when the energy storage contribution does not exceed the preset equilibrium threshold and is not lower than the lower limit of the preset equilibrium threshold, it is determined that the capacity contribution distribution of each unit in the current energy storage system is within a reasonable range that meets the lifetime equilibrium target. At this time, the current division process between the high-response capacity group and the low-response capacity group is terminated, the established charging and discharging scheduling strategy and dynamic charging and discharging depth limit of each energy storage unit remain unchanged, the periodic exchange operation between the two groups of roles is not performed, and the floating range of the preset equilibrium threshold for the next statistical period is finely adjusted and updated based on the capacity response integral value and energy storage contribution distribution of the current statistical period. The operating data of each energy storage unit is continuously collected synchronously, and the energy storage contribution calculation and threshold comparison are re-completed in the next statistical period to maintain the closed-loop operation of the system's equilibrium control.

[0046] In step S4, dynamic charge and discharge depths are configured for the high-response capacity group and the low-response capacity group based on the capacity response integral value, and the roles of the two groups are periodically swapped.

[0047] In this embodiment, configuring dynamic charge / discharge depths for the high-response capacity group and the low-response capacity group based on the capacity response integral value can be achieved through the following steps: Based on the capacity response integral value, the initial reference charge / discharge depths corresponding to the high response capacity group and the low response capacity group are calculated respectively, wherein the reference depth of the high response capacity group is greater than the reference depth of the low response capacity group. Based on the state of charge range and cycle life decay rate of each energy storage unit in each group, a depth adaptive correction factor is constructed to adjust the initial reference charge and discharge depth on a unit-by-unit basis. Output the dynamic charge / discharge depth of each energy storage unit in the high-response capacity group and the low-response capacity group.

[0048] In specific implementation, firstly, the capacity response integral values ​​of all effective energy storage units in the high-response capacity group and the low-response capacity group are statistically analyzed. Using an arithmetic mean method, the group average capacity response integral values ​​for the two groups are calculated. The upper limit is set at the factory-calibrated maximum charge / discharge depth of the energy storage unit, and the lower limit is set at the minimum allowable charge / discharge depth to ensure system dispatch response capability. A linear mapping method is used to map the group average capacity response integral values ​​of the two groups to the upper and lower limit intervals, obtaining the initial reference charge / discharge depths for the two groups, ensuring that the reference depth of the high-response capacity group is greater than that of the low-response capacity group. Then, the real-time state of charge interval of a single energy storage unit within the group for the current statistical period is collected, along with the cycle life decay rate calculated based on the number of completed cycles and the rated cycle life. Both parameters are known and directly obtainable operating parameters in the energy storage field. A linear weighting method is used to map the deviation of the state of charge from the ideal operating range and the cycle life decay rate to sub-correction coefficients within the range of 0.8-1.2. The two sub-correction coefficients are multiplied to obtain a depth adaptive correction factor. The initial reference charge / discharge depth of the corresponding group is multiplied by this factor to complete the weighted adjustment. Finally, the weighted adjustment values ​​of charge and discharge depths for each unit are checked for compliance with the boundary to ensure that the adjusted values ​​do not exceed the maximum allowable charge and discharge depth specified by the energy storage unit at the factory and are not lower than the minimum allowable charge and discharge depth required by the system scheduling. Values ​​exceeding the boundary are replaced with the corresponding boundary limit. The checked charge and discharge depth values ​​are bound and associated with the unique number and grouping attribute of the corresponding energy storage unit. According to the scheduling instruction issuance cycle, they are synchronously output to the local controller of each energy storage unit in the two groups as the execution control parameters for charge and discharge actions.

[0049] It should be noted that, in this application, the initial reference depth of charge / discharge refers to the basic limit values ​​of the depth of charge / discharge set for the high response capacity group and the low response capacity group respectively; the reference depth refers to the basic limit values ​​of the depth of charge / discharge set for the high response capacity group and the low response capacity group respectively; the state of charge range refers to the state parameter of the range of changes in the state of charge during the charge / discharge cycle of the energy storage unit; the cycle lifetime decay rate refers to the parameter that quantifies the degree of life loss of the energy storage unit after cycle use; the depth adaptive correction factor refers to the weighting coefficient that adapts to the operating characteristics of a single energy storage unit and makes fine adjustments to the reference depth of charge / discharge; the unit-by-unit weighted adjustment refers to the process of making personalized adjustments to the reference depth of charge / discharge for the real-time operating characteristics of a single energy storage unit; and the dynamic depth of charge / discharge refers to the maximum allowable depth limit of a single energy storage unit in a single charge / discharge cycle.

[0050] In this embodiment, the periodic swapping of the two sets of roles can be achieved using the following steps: The aging balance of each group is determined based on the cumulative capacity throughput and health status decay rate of the high response capacity group and the low response capacity group. When the corresponding aging balance exceeds the preset threshold, swap the two sets of role labels and reset the cumulative counter, outputting the updated high response capacity group and low response capacity group.

[0051] In specific implementation, firstly, the cumulative capacity throughput of the high-response capacity group and the low-response capacity group within the current role cycle is calculated separately. Using the ampere-hour integral method, the total effective charging and discharging capacity of the two groups in response to scheduling commands throughout the entire cycle is accumulated. The health status decay rate of all energy storage units within the two groups is collected, and the group average health status decay rate of the two groups is calculated using the arithmetic mean method. Using the relative ratio of the cumulative capacity throughput of the two groups and the absolute difference of the group average health status decay rate as parameters, the aging balance degree of the two groups is calculated using a linear weighted method, with the weighting coefficient set according to the system lifespan balance target. Then, a preset threshold is set based on the system's full lifespan balance target. This preset threshold is the maximum allowable limit for the difference in aging loss between the two groups. The calculated aging balance degree is compared with the preset threshold in real time. When the aging balance degree exceeds the preset threshold, it is determined that the difference in aging loss between the two groups exceeds the allowable balance range. A full swap operation of the role labels of the two groups is performed, and the cumulative capacity throughput counters and health status statistical cycles of the two groups are reset simultaneously. Finally, the updated unit sets of the high-response capacity group and the low-response capacity group are output.

[0052] It should be noted that, in this application, cumulative capacity throughput refers to a quantitative indicator of the total effective charge and discharge capacity completed by the high response capacity group and the low response capacity group within the current role cycle; health status decay rate refers to a parameter that quantifies the degree of health status loss of the energy storage unit after cycle use; aging balance refers to an indicator that measures the degree of difference in cycle aging loss between the high response capacity group and the low response capacity group; preset threshold is a benchmark limit used to limit the maximum allowable deviation of aging loss between the two groups; role label refers to an attribute mark that identifies the group category to which the energy storage unit belongs; and cumulative counter refers to a counting carrier that counts the operating data of cumulative capacity throughput and effective cycle count within the two role cycles.

[0053] In this embodiment, reference Figure 4As shown in the diagram, this is a schematic diagram of the dynamic charge / discharge depth configuration of energy storage units and the principle of periodic role exchange between two groups. The diagram calculates the energy storage contribution based on the capacity response integral value of each energy storage unit, thereby dividing it into a high-response capacity group and a low-response capacity group. Differentiated dynamic charge / discharge depths D% are configured for each group: the high-response capacity group undertakes deep charge / discharge scheduling tasks, while the low-response capacity group uses low-intensity charge / discharge to reduce the aging rate. The system uses a preset exchange period T as the control window, calculating the aging balance based on the cumulative capacity throughput and health state decay rate of the two groups. When the difference in aging loss between the two groups exceeds a preset threshold, the role labels of the two groups are fully swapped, and the dynamic charge / discharge depth configuration is updated synchronously. This allows the two groups to alternately undertake high- and low-intensity charge / discharge tasks, ultimately achieving balanced cycle life loss of energy storage units across the entire system while ensuring stable system scheduling response capabilities.

[0054] Therefore, in this application, dynamic charge / discharge depths are configured for the high-response capacity group and the low-response capacity group based on the capacity response integral value, and the roles of the two groups are periodically exchanged. The generation of the capacity response integral value provides a normalized quantification benchmark for the effective capacity contribution of each energy storage unit within the scheduling cycle, thereby eliminating the interference of ambient temperature fluctuations, differences in charge / discharge efficiency, and cyclic aging losses on the quantification of capacity contribution. This accurately characterizes the actual response capability and effective capacity support level of each energy storage unit to scheduling commands. This quantification benchmark enables accurate identification and unbiased comparison of the capacity contribution differences among energy storage units, avoiding the problem of disconnect between traditional scheduling strategies based on rated capacity and the actual operating performance of the units, and providing a unified standard for differentiated scheduling and management. A compliant quantitative basis ensures the consistency and long-term stability of energy storage dispatch response. Dividing into high-response capacity groups and low-response capacity groups provides a differentiated dispatch and control platform adapted to the individual response characteristics of energy storage units. This enables hierarchical control and targeted adaptation of charging and discharging strategies for energy storage units with different contribution levels. It avoids the problems of uneven lifespan loss and increased differentiation in response capabilities among units caused by traditional unified dispatch strategies. Through grouped control, targeted regulation of the capacity contribution differences of energy storage units can be achieved. High-response units ensure rapid and accurate response to dispatch commands, while low-response units provide basic capacity support and lifespan buffering. This achieves synergy between dispatch response performance and overall system lifespan balance, continuously improving the long-term stability of energy storage dispatch response.

[0055] In summary, the technical solution adopted in this application can adjust the differences in capacity contribution of energy storage units based on multi-source operation data, so as to improve the stability of energy storage dispatch response.

[0056] Example 2: This application provides an energy storage dispatching system based on multi-source data analysis, referring to... Figure 5 As shown in the figure, this is a modular structure diagram of an energy storage scheduling system based on multi-source data analysis according to this embodiment of the present application. The energy storage scheduling system includes: The acquisition module 100 is used to acquire the charge and discharge depth records and ambient temperature records corresponding to each energy storage unit during energy storage scheduling. The correction module 200 is used to extract the capacity cycle segment of each energy storage unit from all charge and discharge depth records, and to perform capacity correction on each capacity cycle segment according to all ambient temperature records, thereby generating the capacity response integral value of each energy storage unit. The capacity division module 300 is used to determine the energy storage contribution of the corresponding energy storage unit based on the capacity response integral value of all energy storage units. When the energy storage contribution exceeds the preset equalization threshold, the energy storage units are divided into a high response capacity group and a low response capacity group. The capacity configuration module 400 is used to configure dynamic charge and discharge depths for the high-response capacity group and the low-response capacity group respectively according to the capacity response integral value, and periodically exchange the roles of the two groups.

[0057] This application is described with reference to flowchart illustrations and / or block diagrams of methods, apparatus (systems), and computer program products according to embodiments of this application. It will be understood that each block of the flowchart illustrations and / or block diagrams, and combinations of blocks in the flowchart illustrations and / or block diagrams, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, special-purpose computer, embedded processor, or other programmable data processing apparatus to produce a machine, such that the instructions, which execute via the processor of the computer or other programmable data processing apparatus, generate instructions for implementing the process. Figure 1 One or more processes and / or boxes Figure 1 A device that provides the functions specified in one or more boxes.

[0058] Those skilled in the art will understand that all or part of the steps in the various methods of the above embodiments can be implemented by a program instructing related hardware. The program can be stored in a computer-readable storage medium, including read-only memory (ROM), random access memory (RAM), programmable read-only memory (PROM), erasable programmable read-only memory (EPROM), one-time programmable read-only memory (OTPROM), electrically-Erasable Programmable Read-Only Memory (EEPROM), compactdisc read-only memory (CD-ROM) or other optical disc storage, disk storage, magnetic tape storage, or any other computer-readable medium capable of carrying or storing data.

[0059] It should also be noted that the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such process, method, article, or apparatus. Unless otherwise specified, an element defined by the phrase "comprising one..." does not exclude the presence of other identical elements in the process, method, article, or apparatus that includes that element.

Claims

1. An energy storage scheduling method based on multi-source data analysis, characterized in that, The energy storage dispatch method includes the following steps: Acquire the charge / discharge depth records and ambient temperature records for each energy storage unit during energy storage scheduling; The capacity cycle segment of each energy storage unit is extracted from all charge and discharge depth records, and the capacity of each capacity cycle segment is corrected according to all ambient temperature records to generate the capacity response integral value of each energy storage unit. The energy storage contribution of each energy storage unit is determined based on the capacity response integral value of all energy storage units. When the energy storage contribution exceeds a preset equalization threshold, the energy storage units are divided into a high response capacity group and a low response capacity group. Based on the capacity response integral value, dynamic charge and discharge depths are configured for the high-response capacity group and the low-response capacity group, and the roles of the two groups are periodically swapped.

2. The energy storage scheduling method based on multi-source data analysis as described in claim 1, characterized in that, The capacity cycling segment extracted from all charge / discharge depth records specifically includes: Extract the capacity cycle boundary of each energy storage unit from all charge / discharge depth records; An adaptive phase segmentation window is constructed based on all capacity cycle boundaries. The charge-discharge depth curve is segmented into cycle segments characterized by the state-of-charge turning point through local integral charge matching. The capacity cycle segment of each energy storage unit is determined based on all cycle segments.

3. The energy storage scheduling method based on multi-source data analysis as described in claim 1, characterized in that, Based on all ambient temperature records, capacity correction is performed for each capacity cycle segment to generate the capacity response integral value for each energy storage unit, specifically including: Align all ambient temperature records with the time axis of all capacity cycling segments to construct a thermally induced decay coefficient field and extract local fluctuation entropy. Extract the capacity degradation component of each energy storage unit from the thermally induced degradation coefficient field; The capacity response curve of each energy storage unit is determined based on all capacity degradation components and the local fluctuation entropy. The capacity response curve is converted to capacity to obtain the integral value of the capacity response of each energy storage unit.

4. The energy storage scheduling method based on multi-source data analysis as described in claim 3, characterized in that, Aligning all ambient temperature records with the time axis of all capacity cycling segments, constructing a thermally induced decay coefficient field, and extracting local fluctuation entropy specifically includes: Bidirectional timestamp interpolation is performed on the time axis of all ambient temperature records and all capacity cycle segments to generate synchronized temperature sequences and synchronized capacity sequences; Based on the synchronized temperature sequence, a thermally induced decay coefficient field is constructed by mapping the capacity decay rate point by point. Local fluctuation entropy is extracted within a local sliding window of the thermally induced decay coefficient field.

5. The energy storage scheduling method based on multi-source data analysis as described in claim 1, characterized in that, The energy storage contribution of each energy storage unit is determined based on the integral value of its capacity response. This includes: Construct an initial contribution matrix based on the capacity response integral values ​​of all energy storage units, using weighted state of charge and corrected charge / discharge efficiency. Based on the initial contribution matrix, the energy storage contribution allocation tensor is determined by introducing voltage balance constraints and power response priorities among energy storage units. Extract the contribution coefficient of each energy storage unit from the energy storage contribution allocation tensor, and output the energy storage contribution of the corresponding energy storage unit.

6. The energy storage scheduling method based on multi-source data analysis as described in claim 1, characterized in that, When the energy storage contribution exceeds a preset balancing threshold, the energy storage units are divided into a high-response capacity group and a low-response capacity group, specifically including: The energy storage contribution is compared with a preset equilibrium threshold to generate a contribution residual sequence. Based on the contribution residual sequence, high-response initial groups and low-response initial groups of energy storage units are constructed; Power coupling constraints and state of charge consistency checks are introduced between energy storage units, and neighborhood exchange checks are performed on the high-response initial group and the low-response initial group. Based on the results of the neighborhood exchange verification, the energy storage units are divided into high-response capacity group and low-response capacity group.

7. The energy storage scheduling method based on multi-source data analysis as described in claim 6, characterized in that, Introducing power coupling constraints and state-of-charge consistency checks among energy storage units, the neighborhood exchange check for high-response initial groups and low-response initial groups specifically includes: Based on the high-response initial grouping and the low-response initial grouping, the power coupling strength matrix and state of charge deviation between adjacent energy storage units are determined; A bipartite graph minimum weight matching model is constructed using the power coupling strength matrix and the state of charge deviation, and candidate exchange unit pairs that satisfy the power coupling constraints and the state of charge consistency verification are determined. Based on the candidate exchange units, perform iterative neighborhood exchange verification, update the group labels until the cost converges, and output the neighborhood exchange verification results.

8. The energy storage scheduling method based on multi-source data analysis as described in claim 1, characterized in that, The specific configuration of dynamic charge / discharge depth for the high-response capacity group and the low-response capacity group based on the capacity response integral value includes: Based on the capacity response integral value, the initial reference charge / discharge depths corresponding to the high response capacity group and the low response capacity group are calculated respectively, wherein the reference depth of the high response capacity group is greater than the reference depth of the low response capacity group. Based on the state of charge range and cycle life decay rate of each energy storage unit in each group, a depth adaptive correction factor is constructed to adjust the initial reference charge and discharge depth on a unit-by-unit basis. Output the dynamic charge / discharge depth of each energy storage unit in the high-response capacity group and the low-response capacity group.

9. The energy storage scheduling method based on multi-source data analysis as described in claim 1, characterized in that, The regular swapping of the two sets of roles specifically includes: The aging balance of each group is determined based on the cumulative capacity throughput and health status decay rate of the high response capacity group and the low response capacity group. When the corresponding aging balance exceeds the preset threshold, swap the two sets of role labels and reset the cumulative counter, outputting the updated high response capacity group and low response capacity group.

10. An energy storage scheduling system based on multi-source data analysis, used to execute an energy storage scheduling method based on multi-source data analysis as described in any one of claims 1 to 9, characterized in that, The energy storage dispatch system includes: The acquisition module is used to acquire the charge / discharge depth records and ambient temperature records corresponding to each energy storage unit during energy storage scheduling; The correction module is used to extract the capacity cycle segment of each energy storage unit from all charge and discharge depth records, and to perform capacity correction on each capacity cycle segment based on all ambient temperature records, thereby generating the capacity response integral value of each energy storage unit. The capacity partitioning module is used to determine the energy storage contribution of the corresponding energy storage unit based on the capacity response integral value of all energy storage units. When the energy storage contribution exceeds the preset equalization threshold, the energy storage units are divided into a high response capacity group and a low response capacity group. The capacity configuration module is used to configure dynamic charge and discharge depths for the high-response capacity group and the low-response capacity group respectively based on the capacity response integral value, and periodically switch roles between the two groups.

Images