Distributed energy storage and flexible load complementary characteristic analysis method and system

Abstract

The invention discloses a distributed energy storage and flexible load complementary characteristic analysis method and system, and the method comprises the steps: collecting the related data of distributed energy storage, flexible load and local power grid, and removing abnormal data through screening to construct an original database; analyzing the incidence relation of the two to determine key parameters, and constructing an evaluation system containing indexes such as power matching degree; and a multi-time scale algorithm is adopted to quantitatively analyze complementary characteristics under a minute level, an hour level and a day level, and finally an analysis report of clear advantages and the to-be-optimized area is generated. The system comprises a multi-source data acquisition unit, a preprocessing unit, an incidence relation analysis unit, an evaluation index system construction unit, a multi-time scale quantitative analysis unit and an analysis report generation unit which are sequentially connected to cooperatively work. The method and the system solve the problems of single time scale and imperfect evaluation system in the prior art, improve analysis comprehensiveness and reliability, provide guidance for collaborative optimization of the two, and assist in enhancing power grid stability and energy utilization rate.

Description

Technical Field

[0001] This invention relates to the field of distributed energy storage technology, and in particular to a method and system for analyzing the complementary characteristics of distributed energy storage and flexible loads. Background Technology

[0002] With the increasing penetration rate of distributed energy resources, the scale of distributed energy storage devices and flexible loads in local power grids continues to expand, and the synergistic operation between the two has a more significant impact on grid stability and energy utilization. Currently, the power consumption of various flexible loads, such as industrial loads, HVAC loads, and charging loads, exhibits significant fluctuations and uncertainties, while the charging and discharging states of distributed energy storage devices also dynamically change with the grid operating conditions. How to accurately analyze the complementary characteristics of the two to optimize dispatch strategies and improve grid operating efficiency has become a key issue that urgently needs to be addressed in the energy sector. At the same time, the requirements for the stability of voltage, frequency, and power flow in local power grids are constantly increasing. It is necessary to use scientific methods to identify the advantageous areas and areas requiring optimization for the complementarity of distributed energy storage and flexible loads, providing a reliable basis for grid operation and regulation. Therefore, developing a systematic method for analyzing complementary characteristics and a corresponding system is of significant practical importance.

[0003] Existing technologies for analyzing the complementary characteristics of distributed energy storage and flexible loads suffer from two significant drawbacks. Firstly, existing analytical methods often focus on a single time scale, failing to incorporate quantitative analysis across different time dimensions such as minutes, hours, and days. This prevents a comprehensive capture of the dynamic changes in the complementary characteristics under various operating conditions, resulting in insufficient accuracy and comprehensiveness of the analysis results, making it difficult to support grid dispatch decisions in multiple scenarios. Secondly, the evaluation systems constructed by existing technologies are inadequate. When determining key parameters affecting complementary characteristics, they fail to fully integrate linear and nonlinear correlation characteristics, and the setting of indicator weights lacks a rigorous and scientific verification process. This makes the evaluation indicators unable to accurately reflect the core requirements of complementarity, thus affecting the reliability of the complementary characteristic analysis results and failing to provide effective guidance for the coordinated optimization of distributed energy storage and flexible loads. Summary of the Invention

[0004] In order to overcome the shortcomings and deficiencies of existing technologies, this invention provides a method and system for analyzing the complementary characteristics of distributed energy storage and flexible loads.

[0005] The technical solution adopted in this invention is as follows:

[0006] A method for analyzing the complementary characteristics of distributed energy storage and flexible loads includes the following steps:

[0007] S1 collects output power data, charging and discharging status data of distributed energy storage devices under different operating conditions, as well as real-time power consumption data and operating status parameters of industrial load, HVAC load and charging load in flexible loads, and simultaneously obtains voltage, frequency and power flow data of the local power grid.

[0008] S2, classify and filter the collected data, remove abnormal data, and establish the original database of distributed energy storage and flexible load;

[0009] S3. Based on the data in the original database, analyze the correlation between the charging and discharging characteristics of distributed energy storage devices and the power change characteristics of different types of flexible loads, and determine the calibration parameters that affect the complementary characteristics.

[0010] S4. Based on the calibration parameters, construct an evaluation index system for the complementary characteristics of distributed energy storage and flexible loads. This system includes indicators related to power matching degree, response synchronization, and capacity utilization.

[0011] S5 employs a multi-timescale analysis algorithm to quantitatively analyze the complementary characteristics of distributed energy storage and flexible loads at minute, hour, and day timescales, respectively.

[0012] S6 generates an analysis report on the complementary characteristics of distributed energy storage and flexible loads based on the quantitative analysis results at different time scales, clarifying the advantageous areas and areas to be optimized for complementarity under different operating conditions.

[0013] Furthermore, the following model is used when analyzing the correlation in S3:

[0014] ,

[0015] in, The combined power value of complementary characteristics at time t. For distributed energy storage power weighting coefficients, Let be the output power of the distributed energy storage at time t. This is the weighting factor for the total power of flexible loads. For the types and quantities of flexible loads, Let be the power consumption of the i-th type of flexible load at time t. For local power grid power weighting coefficients, Let t be the local power grid input power at time t.

[0016] Furthermore, when constructing the evaluation index system in S4, the power matching degree index adopts the following model:

[0017] ,

[0018] in, For power matching degree, These represent the start and end times of the analysis period, respectively. The value of the flexible load power fluctuation at time t;

[0019] The response synchronization index uses the following model:

[0020] ,

[0021] in, In response to synchronization, Let t be the distributed energy storage response delay time. Let t be the flexible load response delay time. This is the preset maximum response latency threshold.

[0022] Furthermore, the minute-scale timescale quantification analysis in S5 employs the following model:

[0023] ,

[0024] in, This is the quantitative analysis value for the k-th minute time period. This refers to the number of data collections within a minute-level time period. Let J be the weight of the data collected in the j-th time. Let j be the time interval for the j-th data collection. This represents the distributed energy storage power collected in the j-th time interval within the k-th minute period. This represents the total power of the flexible load collected in the j-th time interval within the k-th minute period.

[0025] Furthermore, the daily-scale timescale quantitative analysis in S5 adopts the following model:

[0026] ,

[0027] in, These are daily quantitative analysis values. The total number of seconds in a day. The time weighting coefficient for the h-th hour. Let t be the distributed energy storage capacity utilization rate. Let t be the utilization rate of flexible load capacity demand.

[0028] Furthermore, when generating the analysis report in S6, the following model is used to determine the advantageous regions:

[0029] ,

[0030] in, For the collection of advantageous areas, The complementary characteristic quantization value is given when the power is P at time t. For quantization threshold, These represent the minimum and maximum output power of distributed energy storage, respectively. These represent the minimum and maximum power consumption for flexible loads, respectively.

[0031] Furthermore, step S3 includes the following sub-steps:

[0032] S31. Extract the continuous 72-hour charging and discharging power data of distributed energy storage devices and the continuous 72-hour power consumption data of different types of loads in flexible loads from the original database, align them by timestamp, and form a paired data sequence.

[0033] S32, perform sliding window processing on the paired data sequence, with the window size set to 30 minutes, calculate the covariance between the distributed energy storage power and the flexible load power within each window, and preliminarily determine the degree of linear correlation between the two;

[0034] S33 introduces a nonlinear correlation analysis method to calculate the mutual information value between the rate of change of the charge and discharge state of distributed energy storage and the rate of change of the power of flexible load, and explores the nonlinear correlation characteristics between the two.

[0035] S34. Based on the combined results of linear covariance and nonlinear mutual information, parameters with a correlation degree higher than a preset threshold are selected and identified as calibration parameters that affect complementary properties.

[0036] Furthermore, step S4 includes the following sub-steps:

[0037] S41. Referring to the power system operation specifications and energy management standards, power balance, response speed consistency, capacity adaptability, operation stability, and economic indicators are initially selected as candidate evaluation indicators.

[0038] S42. Using the analytic hierarchy process, construct an indicator weight judgment matrix, invite 5-8 experts in the power system field to compare and score the importance of each indicator in the candidate set pairwise, and calculate the weight value of each indicator.

[0039] S43. Based on the weight values, indicators with weights below 0.05 are removed, and indicators with higher weights are retained to form a preliminary evaluation indicator system.

[0040] S44. Conduct a consistency test on the preliminary evaluation index system. If the test passes, it is determined as the final evaluation index system for the complementary characteristics of distributed energy storage and flexible load. If it fails, the indexes and weights are readjusted, and the test is repeated until it passes.

[0041] Furthermore, step S5 includes the following sub-steps:

[0042] S51, the data in the original database is divided into minute-level, hour-level, and day-level datasets according to the time scale. The minute-level dataset is collected every 5 minutes, the hour-level dataset is collected every hour, and the day-level dataset is summarized every 24 hours.

[0043] S52 uses the short-time Fourier transform algorithm to convert power data to the frequency domain for minute-level datasets, and analyzes the complementary characteristics of distributed energy storage and flexible load power in the high-frequency band.

[0044] S53, for hourly datasets, uses wavelet transform algorithm to decompose different frequency components of power data and analyzes the variation law of complementary characteristics in the mid-frequency band;

[0045] S54 uses a trend analysis algorithm to fit the power data change trend curve for daily-level datasets, analyzes the long-term complementary characteristics of low-frequency bands, and completes multi-timescale quantitative analysis.

[0046] A system for analyzing the complementary characteristics of distributed energy storage and flexible loads, the system being applied to the aforementioned method for analyzing the complementary characteristics of distributed energy storage and flexible loads, comprising:

[0047] The multi-source data acquisition unit is used to collect the output power and charging / discharging status data of distributed energy storage devices under different operating conditions, the real-time power consumption data and operating status parameters of industrial load, HVAC load and charging load in flexible loads, as well as the voltage, frequency and power flow data of the local power grid, and transmit the collected data to the data preprocessing unit.

[0048] The data preprocessing unit is connected to the multi-source data acquisition unit. It classifies and filters the acquired data, removes abnormal data, establishes a raw database, and sends the raw database to the correlation analysis unit.

[0049] The correlation analysis unit, connected to the data preprocessing unit, analyzes the correlation between distributed energy storage and flexible loads based on the original database, and determines the calibration parameters that affect the complementary characteristics.

[0050] The evaluation index system construction unit is connected to the correlation analysis unit. Based on the calibration parameters, an evaluation index system for the complementary characteristics of distributed energy storage and flexible load is constructed. This system includes indicators related to power matching degree, response synchronization, and capacity utilization.

[0051] The multi-timescale quantitative analysis unit is connected to the evaluation index system construction unit. It adopts a multi-timescale analysis algorithm to quantitatively analyze the complementary characteristics of distributed energy storage and flexible load at minute, hour, and day timescales respectively.

[0052] The analysis report generation unit is connected to the multi-timescale quantitative analysis unit. Based on the quantitative analysis results at different time scales, it generates an analysis report on the complementary characteristics of distributed energy storage and flexible load, clarifying the advantageous areas and areas to be optimized for the complementarity of the two under different operating conditions.

[0053] The present invention has the following beneficial effects:

[0054] 1. At the data processing and analysis level, multi-source data collection covers distributed energy storage devices, various flexible loads, and key data of local power grids. After classification, screening, and elimination of abnormal data, a high-quality raw database is constructed, laying a reliable foundation for subsequent analysis. With the help of multi-time scale analysis algorithms, quantitative analysis is carried out at the minute, hour, and day levels to comprehensively capture the dynamic changes of the complementary characteristics of the two under different operating conditions. This effectively solves the problem that the accuracy and comprehensiveness of the analysis results are insufficient due to the focus on a single time scale in existing technologies, and provides sufficient support for power grid dispatching decisions in multiple scenarios.

[0055] 2. At the evaluation and application level, an evaluation system including indicators such as power matching degree and response synchronization is constructed. When determining the key parameters affecting complementary characteristics, the linear and nonlinear correlation characteristics are comprehensively considered. The weights of the indicators are determined by the analytic hierarchy process and consistency verification is carried out to ensure that the evaluation indicators accurately reflect the core needs of complementarity. This makes up for the shortcomings of the existing technology evaluation system and the lack of scientific verification of the indicator weight settings. Finally, an analysis report is generated that clearly defines the advantageous areas and the areas to be optimized, providing effective guidance for the coordinated optimization of distributed energy storage and flexible loads, and helping to improve grid stability and energy utilization. Attached Figure Description

[0056] Figure 1 This is a flowchart illustrating the overall steps of the method for analyzing the complementary characteristics of distributed energy storage and flexible loads according to the present invention.

[0057] Figure 2 This is a flowchart of method step S3 of the present invention;

[0058] Figure 3 This is a flowchart of method step S4 of the present invention;

[0059] Figure 4 This is a flowchart of step S5 of the method of the present invention;

[0060] Figure 5 This is a unit composition diagram of a distributed energy storage and flexible load complementary characteristic analysis system according to the present invention. Detailed Implementation

[0061] It should be noted that, unless otherwise specified, the embodiments and features described in this application can be combined with each other. The application will be further described in detail below with reference to the accompanying drawings and specific embodiments.

[0062] like Figure 1 As shown, a method for analyzing the complementary characteristics of distributed energy storage and flexible loads includes the following steps:

[0063] S1 collects output power data, charging and discharging status data of distributed energy storage devices under different operating conditions, as well as real-time power consumption data and operating status parameters of industrial load, HVAC load and charging load in flexible loads, and simultaneously obtains voltage, frequency and power flow data of the local power grid.

[0064] Specifically, step S1 completes the comprehensive collection of multi-dimensional data, providing basic data support for subsequent analysis. During implementation, for distributed energy storage devices, output power data needs to be collected under three operating conditions: charging, discharging, and standby. The collection frequency is set to once every 5 seconds to ensure the capture of short-term power fluctuations. Simultaneously, charging and discharging status data, including start and end times and the current charging / discharging rate, is recorded. The charging / discharging rate collection range covers 0.2C to 2C to include common operating states of the equipment. For flexible loads, real-time power consumption data needs to be collected separately for industrial loads, HVAC loads, and charging loads. Industrial load data collection focuses on the operating period of production equipment, collected once every 10 seconds. HVAC load data needs to be collected considering seasonal differences. Data is collected every 5 minutes in winter and summer, and every 15 minutes in spring and autumn. Charging load data is collected in real-time according to the charging pile's operating status. Power consumption is recorded every minute during each charging cycle. Simultaneously, operating status parameters of three types of flexible loads are collected: equipment start / stop status for industrial loads, temperature setpoints for HVAC loads, and charging gun connection status for charging loads. For the local power grid, voltage, frequency, and power flow data are collected. Voltage acquisition accuracy is controlled within ±0.01kV, frequency acquisition accuracy within ±0.01Hz, and power flow data is collected every minute, including the input and output power of each node in the power grid. This step, by refining the acquisition frequency and accuracy, ensures that the acquired data accurately reflects the operating status of each object, laying a high-quality data foundation for subsequent data processing and analysis.

[0065] S2. Classify and filter the collected data from distributed energy storage devices, flexible loads, and local power grids. After removing abnormal data, establish the original database of distributed energy storage and flexible loads.

[0066] Specifically, step S2 aims to filter and process the collected raw data to build a reliable raw database. In practice, the data is first categorized according to its source: distributed energy storage device data, flexible load data, and local power grid data. After categorization, the 3σ criterion is used to remove outlier data. This involves calculating the mean and standard deviation of each data category, and identifying data exceeding the mean ± 3 times the standard deviation as outlier. For example, in distributed energy storage output power data, if the data exceeds the device's rated power ± 30% at a certain moment and does not conform to the 3σ criterion, it is marked as outlier and removed. For flexible load data, in addition to the 3σ criterion, the load's own operating patterns are also considered to help determine outliers. For example, high power consumption data during non-production periods for industrial loads or power data during equipment shutdown periods for HVAC loads are both identified as outliers and removed. For local power grid data, the grid's rated parameters are considered; values ​​exceeding the rated voltage ± 10% or the frequency exceeding 50Hz ± 0.5Hz are considered outlier and removed. After removing outlier data, the remaining valid data is formatted uniformly by converting all data to CSV format and sorting it by timestamp. The timestamp precision is uniformly set to the millisecond level to ensure time consistency of data from different sources. Finally, the original database for distributed energy storage and flexible load is constructed. The database storage capacity must meet the data storage requirements of at least 3 months of historical data. At the same time, a data backup mechanism is set up to automatically back up the data of the day at 2:00 AM every day to prevent data loss and provide a stable and reliable data source for subsequent steps.

[0067] S3. Based on the data in the original database, analyze the correlation between the charging and discharging characteristics of distributed energy storage devices and the power change characteristics of different types of flexible loads, and determine the calibration parameters that affect the complementary characteristics.

[0068] Specifically, step S3, which identifies key parameters by analyzing data correlations, is a crucial step in connecting data with the evaluation system. During implementation, firstly, valid data for 72 consecutive hours is extracted from the original database. The charging and discharging characteristics data of distributed energy storage devices are then aligned with the power variation characteristics data of various flexible loads according to timestamps, with alignment accuracy controlled within 1 second to ensure matching in the time dimension. Next, correlation analysis is used to calculate the Pearson correlation coefficient between the charging and discharging power of distributed energy storage and the power variations of industrial loads, HVAC loads, and charging loads. The correlation coefficient calculation results are rounded to four decimal places. The magnitude of the coefficient is used to initially determine the degree of linear correlation: an absolute value greater than 0.6 is considered a strong linear correlation, between 0.3 and 0.6 is considered a moderate linear correlation, and less than 0.3 is considered a weak linear correlation. Subsequently, time series analysis is introduced... The time-lag correlation between the two was analyzed, with a lag time range of 0 to 30 minutes. The cross-correlation coefficients at different lag times were calculated to determine the lag time point with the strongest correlation. For example, if the cross-correlation coefficient between the change in the charging and discharging power of distributed energy storage and the change in the power of HVAC load is the largest at a lag of 10 minutes, then this lag time point is recorded. Finally, by combining the results of linear correlation analysis and time-lag correlation analysis, parameters that significantly affect complementary characteristics were selected, such as the charging and discharging power of distributed energy storage, the charging and discharging rate, the rate of change of HVAC load power, the start-stop frequency of industrial loads, and the charging duration of charging loads. These parameters were identified as key parameters affecting complementary characteristics, providing a clear parameter basis for the subsequent construction of the evaluation index system.

[0069] S4. Based on the calibration parameters, construct an evaluation index system for the complementary characteristics of distributed energy storage and flexible loads. This system includes indicators related to power matching degree, response synchronization, and capacity utilization.

[0070] Specifically, step S4 constructs a scientific and comprehensive evaluation index system for complementary characteristics, providing a standard framework for quantitative analysis. During implementation, firstly, based on the key parameters determined in step S3, preliminary evaluation indicators are selected, including three core categories: power matching degree, response synchronicity, and capacity utilization. The power matching degree indicator measures the degree of matching between the output power of distributed energy storage and the power demand of flexible loads; the response synchronicity indicator assesses the time synchronization between the response of distributed energy storage and the power changes of flexible loads; and the capacity utilization indicator reflects the degree of adaptation between the capacity of distributed energy storage and the capacity demand of flexible loads. Next, each type of indicator is further defined. The power matching degree indicator requires a clearly defined time window for calculation, set to 15 minutes; the response synchronicity indicator requires a calculation benchmark for the response delay time, using the moment of power change in flexible loads as the benchmark point; and the capacity utilization rate... The indicators need to define the statistical scope of capacity, including the rated capacity of distributed energy storage and the maximum capacity demand of flexible loads. Then, the calculation dimensions of each indicator are determined: power matching degree and response synchronization are calculated at the minute and hour levels respectively, and capacity utilization rate is calculated at the daily level. The value ranges for each indicator are also clearly defined: power matching degree ranges from 0 to 1, response synchronization rate ranges from 0 to 1, and capacity utilization rate ranges from 0 to 1. Values ​​closer to 1 indicate better characteristics. Finally, the completeness of the constructed indicator system is verified to check whether it covers the evaluation requirements corresponding to key parameters. If any omissions are found, they are supplemented and improved to ensure that the evaluation indicator system can comprehensively and accurately reflect the complementary characteristics of distributed energy storage and flexible loads, providing clear evaluation standards for subsequent quantitative analysis.

[0071] S5 employs a multi-timescale analysis algorithm to quantitatively analyze the complementary characteristics of distributed energy storage and flexible loads at minute, hour, and day timescales, respectively.

[0072] Specifically, step S5 employs a multi-timescale analysis algorithm to achieve precise quantitative analysis of complementary characteristics. During implementation, the timescales are first defined, clarifying the specific time ranges for the three analysis dimensions: minute-level, hour-level, and day-level. The minute-level timescale uses one hour as an analysis cycle, with each cycle divided into one-minute units. The hour-level timescale uses one day as an analysis cycle, with each cycle divided into one-hour units. The day-level timescale uses one month as an analysis cycle, with each cycle divided into one-day units. For minute-level timescales, a sliding window algorithm is used for quantitative analysis, with a window size of 5 minutes and a sliding step of 1 minute. The average power matching degree and response synchronicity within each window are calculated, with the average rounded to four decimal places. The results of continuous window calculations are used to analyze short-term fluctuations in complementary characteristics. For hourly timescales, a weighted average algorithm is used, with the number of data collections per minute within each hour as the weight. The weighted average of power matching degree and response synchronicity for that hour is calculated, along with the instantaneous value of capacity utilization. Hourly data is used to observe the intraday trend of complementary characteristics. For daily timescales, a statistical analysis algorithm is used to calculate the daily average values ​​of power matching degree and response synchronicity, as well as the daily maximum, minimum, and average values ​​of capacity utilization. Daily data is used to understand the long-term changing patterns of complementary characteristics. Throughout the quantitative analysis process, the continuity of data calculations across timescales must be ensured. Minute-level data provides the foundation for hourly-level data, and hourly-level data supports daily-level data, resulting in multi-dimensional and multi-layered quantitative analysis results.

[0073] S6 generates an analysis report on the complementary characteristics of distributed energy storage and flexible loads based on the quantitative analysis results at different time scales, clarifying the advantageous areas and areas to be optimized for complementarity under different operating conditions.

[0074] Specifically, step S6 generates an analysis report based on the quantitative analysis results, clarifying the advantages and areas for optimization of complementary characteristics, providing guidance for practical applications. During implementation, the quantitative analysis results at different time scales are first summarized. Minute-level results focus on summarizing the fluctuation ranges of power matching degree and response synchronization, statistically analyzing the duration of fluctuations within the three ranges of 0.8 to 1, 0.5 to 0.8, and 0 to 0.5. Hourly-level results summarize the average change curves of power matching degree and response synchronization for each hour within the day, as well as the hourly distribution of capacity utilization. Daily-level results summarize the average, maximum, and minimum values ​​of various indicators for each day within the month, forming a monthly statistical report. Next, based on the summarized results, advantageous regions are determined. Time periods and operating conditions with average power matching degree ≥ 0.8, average response synchronization ≥ 0.8, and average capacity utilization ≥ 0.7 are defined as advantageous regions. For example, when distributed energy storage discharges at 1C rate, HVAC loads are in high demand periods, and the local power grid voltage is stable, if all three indicators meet the above thresholds, it is marked as an advantageous region, and the corresponding equipment operating parameters and power grid conditions are recorded. At the same time, regions to be optimized are determined. Time periods and operating conditions where any one of the following indicators is met (average power matching degree < 0.5, average response synchronization < 0.5, average capacity utilization < 0.4) are defined as regions to be optimized. Common characteristics of data in regions to be optimized are analyzed, such as excessively high distributed energy storage charge / discharge rates and excessively large power fluctuations in flexible loads. Finally, an analysis report is generated according to the standard report format. The report includes four parts: a data summary table, indicator change curves, a distribution map of advantageous areas, and an analysis of problems in areas to be optimized. The report clearly marks the time range, operating parameters, and operating conditions corresponding to each area, providing specific and operable guidance for the scheduling optimization and parameter adjustment of distributed energy storage and flexible loads.

[0075] Preferably, the following model is used when analyzing the correlation in step S3: ,in, The combined power value of complementary characteristics at time t. For distributed energy storage power weighting coefficients, Let be the output power of the distributed energy storage at time t. This is the weighting factor for the total power of flexible loads. For the types and quantities of flexible loads, Let be the power consumption of the i-th type of flexible load at time t. For local power grid power weighting coefficients, Let t be the local power grid input power at time t.

[0076] Specifically, the correlation analysis model in step S3 quantifies the comprehensive power value of the complementary characteristics of distributed energy storage and flexible loads by integrating multiple types of power data. During implementation, the value range of each weight coefficient in the model is first determined. The distributed energy storage power weight coefficient needs to be set based on its capacity proportion in the local power grid, typically ranging from 0.3 to 0.6. When the total capacity of distributed energy storage accounts for more than 50% of the adjustable capacity of the local power grid, this coefficient is set to 0.5 to 0.6. The flexible load total power weight coefficient is determined based on the proportion of the total power of the flexible load to the grid load, generally between 0.2 and 0.5. If the proportion of the flexible load exceeds 40%, the coefficient is set to 0.4 to 0.5. The local power grid power weight coefficient represents the remaining complementary portion, ranging from 0.1 to 0.3, ensuring that the sum of the three coefficients is 1. In the data input stage, real-time power data for 24 consecutive hours must be selected, with the time interval consistent with the data collection frequency in step S1. Specifically, distributed energy storage output power is collected at one data point every 5 seconds, flexible load power at one data point every 1 to 15 minutes, and local grid input power at one data point every minute. The comprehensive power value of complementary characteristics calculated by the model can intuitively reflect the strength of the synergistic effect of the three at different times. When the comprehensive power value is between 0 and 30% of the rated power, it indicates a weak complementary effect; between 30% and 70%, the complementary effect is moderate; and above 70%, the complementary effect is strong. This result provides a quantitative basis for subsequent key parameter selection and helps to accurately identify factors that significantly affect complementary characteristics.

[0077] Preferably, when constructing the evaluation index system in step S4, the power matching degree index adopts the following model: ,in, For power matching degree, These represent the start and end times of the analysis period, respectively. The value of flexible load power fluctuation at time t; the response synchronicity index adopts the following model: ,in, In response to synchronization, Let t be the distributed energy storage response delay time. Let t be the flexible load response delay time. This is the preset maximum response latency threshold.

[0078] Specifically, the power matching degree and response synchronization index model in step S4 of the evaluation index system quantifies the indexes through specific calculation logic. When calculating the power matching degree, the start and end times of the analysis period need to be determined according to the actual application scenario. If it is used for daily scheduling analysis, the period is set to 24 hours; if it is used for specific operating condition analysis, the period can be set to 1 to 4 hours. The power fluctuation value of the flexible load needs to be calculated by the power difference between adjacent times. The time interval is consistent with the power acquisition frequency of the flexible load. For example, the fluctuation value is calculated once every 10 seconds for industrial loads and once every 5 to 15 minutes for HVAC loads. The calculation result ranges from 0 to the rated power. The smaller the value, the higher the matching degree between the distributed energy storage output power and the power fluctuation of the flexible load. When the value is less than 10% of the rated power, it is judged as excellent power matching. When calculating the response synchronization, the preset maximum response delay time threshold needs to be set with reference to industry standards, usually 5 to 10 seconds. The response delay time between distributed energy storage and flexible load needs to be obtained through actual testing, that is, recording the time difference between the issuance of the power change command and the actual power adjustment. In the model calculation, the absolute value of the time difference of each response delay within the time period is first integrated, then divided by the product of the time period length and the maximum response delay time threshold. Finally, the result is subtracted from 1 to obtain the response synchronicity. Its value ranges from 0 to 1. A value greater than 0.8 indicates that the two responses have good synchronicity and can quickly and collaboratively adjust to adapt to the power grid demand. This indicator provides a core quantitative standard for the evaluation system.

[0079] Preferably, the minute-scale timescale quantification analysis in S5 adopts the following model: ,in, This is the quantitative analysis value for the k-th minute time period. This refers to the number of data collections within a minute-level time period. Let J be the weight of the data collected in the j-th time. Let j be the time interval for the j-th data collection. This represents the distributed energy storage power collected in the j-th time interval within the k-th minute period. This represents the total power of the flexible load collected in the j-th time interval within the k-th minute period.

[0080] Specifically, in step S5, the minute-scale quantitative analysis model reflects short-term complementary characteristics through weighted calculation of multiple data points. In implementation, the first step is to determine the division of minute-scale time periods, dividing one hour into 60 one-minute time periods, each serving as an independent analysis unit. The number of data collections is determined based on the collection frequency of the corresponding data in step S1. If distributed energy storage power is collected once every 5 seconds and flexible load power is collected once every 10 seconds, then within each one-minute time period, distributed energy storage power has 12 data points and flexible load power has 6 data points. The smaller of the two data point counts is taken as the collection count for that time period to ensure data matching. The weight of each collected data is set based on the distance between the data collection time and the midpoint of the time period; the closer to the midpoint, the greater the weight. For example, the weight of data near the 30th second within a one-minute time period is set to 0.2, and the weight of data at both ends of the time period is set to 0.05, with the sum of all weights being 1. The calculation first calculates the difference between the distributed energy storage power and the total power of the flexible load for each data point. Then, the difference is squared, multiplied by the corresponding weight, and summed to obtain the quantitative analysis value for that minute-level time period. The value ranges from 0 to the square of the rated power. The smaller the value, the better the complementary characteristics of the two within that minute. By continuously calculating the values ​​for 60 time periods, a minute-level complementary characteristic change curve can be plotted, clearly showing the short-term fluctuation pattern and providing data support for minute-level grid dispatch.

[0081] Preferably, the daily-scale timescale quantitative analysis in S5 adopts the following model: ,in, These are daily quantitative analysis values. The total number of seconds in a day. The time weighting coefficient for the h-th hour. Let t be the distributed energy storage capacity utilization rate. Let t be the utilization rate of flexible load capacity demand.

[0082] Specifically, in step S5, the daily-scale quantitative analysis model reflects the capacity utilization synergy in the long-term complementary characteristics through daily average calculations. During implementation, the total number of seconds in a day is fixed at 86,400 seconds, serving as the base time for model calculations. The time weighting coefficient for the h-th hour needs to be set in conjunction with the grid load characteristics of that period. The weighting coefficient for peak electricity consumption periods (e.g., 8:00-12:00, 18:00-22:00) is set to 0.05-0.06, for average periods (e.g., 6:00-8:00, 12:00-18:00, 22:00-24:00) to 0.03-0.04, and for off-peak periods (e.g., 0:00-6:00) to 0.01-0.02, ensuring the sum of the weighting coefficients for 24 hours is 1. The distributed energy storage capacity utilization rate is calculated as the ratio of real-time capacity to rated capacity. Real-time capacity needs to be collected every minute, resulting in 60 data points per hour. The flexible load capacity demand utilization rate is calculated as the ratio of real-time capacity demand to maximum capacity demand. The real-time capacity demand collection frequency is consistent with the flexible load power collection frequency. The calculation first integrates the capacity utilization difference for each hour to obtain the hourly integral result, then multiplies it by the corresponding hourly weighting coefficient. Finally, the results for 24 hours are summed and divided by 86400 to obtain the daily quantitative analysis value. This value ranges from 0 to 1. A value greater than 0.7 indicates good synergy between the capacity utilization of distributed energy storage and flexible loads on that day. Maintaining this value in the long term can improve the overall capacity utilization efficiency of the power grid and provide a basis for formulating daily dispatch plans.

[0083] Preferably, when generating the analysis report in step S6, the following model is used to determine the advantageous regions: ,in, For the collection of advantageous areas, The complementary characteristic quantization value is given when the power is P at time t. For quantization threshold, These represent the minimum and maximum output power of distributed energy storage, respectively. These represent the minimum and maximum power consumption for flexible loads, respectively.

[0084] Specifically, the model for determining the advantageous region in step S6 precisely defines the complementary advantageous region by setting a quantification threshold and power range. During implementation, the quantification threshold for complementary characteristics is first determined. This threshold needs to be set based on historical data analysis. If the quantification value is less than 0.3 for more than 70% of the excellent complementary periods in historical data, the threshold is set to 0.3; if the quantification value is less than 0.4 for 60% to 70% of the excellent periods, the threshold is set to 0.4. The minimum and maximum output power of distributed energy storage need to refer to the rated parameters of the equipment. The minimum output power is typically 10% to 20% of the rated power, and the maximum output power is 90% to 100% of the rated power, ensuring that it is within the safe operating range of the equipment. The minimum and maximum power consumption of flexible loads are determined based on historical load data statistics. The minimum value is taken as 80% to 90% of the minimum load of the same period in history, and the maximum value is taken as 110% to 120% of the maximum load of the same period in history, covering possible load fluctuations. In the data screening stage, 24-hour data from seven consecutive days must be selected. Each data point corresponds to a specific time and power value. The time and power combinations with quantified values ​​below a threshold and power values ​​within the corresponding range are selected to form a set of advantageous regions. This set of advantageous regions needs to be statistically analyzed separately along the time and power dimensions. The time dimension shows the proportion of advantageous regions in each time period, and the power dimension shows the proportion of advantageous regions in each power range. This result clarifies the complementary advantages under different time periods and power levels, providing precise guidance for optimizing distributed energy storage charging and discharging strategies and flexible load control schemes.

[0085] Preferred, such as Figure 2 As shown, step S3 includes the following sub-steps: S31, extracting 72-hour continuous charging and discharging power data of distributed energy storage devices and 72-hour continuous power consumption data of different types of loads in flexible loads from the original database, aligning them by timestamps to form a paired data sequence; S32, performing sliding window processing on the paired data sequence, setting the window size to 30 minutes, calculating the covariance between the distributed energy storage power and the flexible load power within each window, and initially determining the degree of linear correlation between the two; S33, introducing a nonlinear correlation analysis method to calculate the mutual information value between the rate of change of the charging and discharging state of distributed energy storage and the rate of change of the power of flexible loads, and mining the nonlinear correlation characteristics between the two; S34, combining the results of linear covariance and nonlinear mutual information value, selecting parameters with a correlation degree higher than a preset threshold, and determining them as calibration parameters affecting complementary characteristics.

[0086] Specifically, the correlation analysis in step S3 achieves precise determination of key parameters through four sub-steps. In implementation, S31 extracts data from the original database, explicitly selecting 72 consecutive hours of charging and discharging power data from distributed energy storage devices and 72 consecutive hours of power consumption data from various flexible loads. This duration covers operating conditions at different times, comprehensively reflecting characteristics. After extraction, data is aligned by timestamps with a precision down to the second level, ensuring complete temporal consistency for each paired data sequence and laying the foundation for subsequent analysis. S32 performs sliding window processing, setting the window size to 30 minutes. This duration captures characteristic changes within a certain time period while avoiding loss of detail due to an excessively large window. The covariance of the power of the two components within each window is calculated. The covariance calculation must be based on all data points within the window. The results initially determine the degree of linear correlation; the larger the absolute value of the covariance, the stronger the linear correlation. S33 introduces a nonlinear correlation analysis method to calculate the mutual information value of the rates of change between the two data points. The rate of change is calculated as the ratio of the difference between adjacent data points to the time interval. The mutual information value calculation must include all possible combinations of rates of change to uncover complex correlation features that linear analysis cannot capture. S34 synthesizes the results of the two analyses and presets a correlation threshold. This threshold is set based on historical data analysis and industry experience, typically set to 0.6. Parameters with a correlation higher than this threshold are selected and identified as key parameters. This process ensures that key parameters comprehensively reflect the core factors affecting complementary characteristics, providing an accurate basis for the subsequent construction of the evaluation system.

[0087] Preferred, such as Figure 3 As shown, step S4 includes the following sub-steps: S41, referring to power system operation specifications and energy management standards, initially select power balance, response speed consistency, capacity adaptability, operational stability, and economic indicators as candidate evaluation indicators; S42, using the analytic hierarchy process (AHP), construct an indicator weight judgment matrix, invite 5-8 power system experts to conduct pairwise comparisons and scores of the importance of each indicator in the candidate set, and calculate the weight value of each indicator; S43, based on the weight value, eliminate indicators with a weight lower than 0.05, retain indicators with higher weights, and form a preliminary evaluation indicator system; S44, conduct a consistency test on the preliminary evaluation indicator system. If the test passes, it is determined as the final evaluation indicator system for the complementary characteristics of distributed energy storage and flexible loads; if it fails, the indicators and weights are readjusted, and the test is repeated until it passes.

[0088] Specifically, the construction of the evaluation index system in step S4 involves four sub-steps to form a scientific and comprehensive index system. During implementation, S41 initially selects candidate indicators, referencing power system operation specifications and energy management standards, and explicitly includes indicators related to power balance, response speed consistency, capacity adaptability, operational stability, and economic efficiency. These indicators cover the core evaluation dimensions of complementary characteristics, ensuring the comprehensiveness of the candidate set. S42 uses the analytic hierarchy process (AHP) to construct an index weight judgment matrix, inviting 5-8 experts in the power system field to participate in scoring. Experts must have more than 5 years of relevant research or practical experience. Scoring uses a 1-9 scale to compare the importance of candidate indicators pairwise, and calculates the weight value of each indicator based on the scoring results. The weight calculation must be performed through matrix operations to ensure accuracy. S43... The indicators are selected based on their weight values, with a weight threshold of 0.05. Indicators with weights lower than this value are eliminated. This threshold effectively excludes indicators with a weak impact on complementary characteristics, retaining core indicators with higher weights to form a preliminary evaluation indicator system. S44 performs a consistency test using the consistency ratio indicator specific to the analytic hierarchy process (AHP), with a consistency ratio threshold of 0.1. If the calculated consistency ratio is less than this threshold, the test passes, and the system is confirmed as the final system. If it is greater than this threshold, experts need to be invited to adjust the indicator importance scores and correct the weight values, and the test is repeated until it passes. This process ensures the scientific validity and rationality of the indicator system, providing a reliable standard for subsequent quantitative analysis.

[0089] Preferred, such as Figure 4 As shown, step S5 includes the following sub-steps: S51, dividing the data in the original database into minute-level, hour-level, and daily-level datasets according to time scales, wherein the minute-level dataset is collected every 5 minutes, the hour-level dataset is collected every hour, and the daily-level dataset is summarized every 24 hours; S52, for the minute-level dataset, using the short-time Fourier transform algorithm, converting the power data to the frequency domain, and analyzing the complementary characteristics of distributed energy storage and flexible load power in the high-frequency band; S53, for the hour-level dataset, using the wavelet transform algorithm, decomposing the different frequency components of the power data, and analyzing the variation law of complementary characteristics in the mid-frequency band; S54, for the daily-level dataset, using the trend analysis algorithm, fitting the trend curve of power data variation, analyzing the long-term complementary characteristics in the low-frequency band, and completing the multi-time-scale quantitative analysis.

[0090] Specifically, step S5, multi-timescale quantitative analysis, achieves characteristic analysis across different dimensions through four sub-steps. In implementation, S51 divides the dataset, explicitly categorizing the raw data into minute-level, hourly-level, and daily-level datasets based on time scale. The minute-level dataset is collected every 5 minutes, a frequency that accurately captures short-term power fluctuations. The hourly dataset is collected every hour, reflecting average characteristics within a time period. The daily dataset is summarized every 24 hours, showcasing long-term trends. After division, it is ensured that each dataset is complete, without missing or anomalies. S52 processes the minute-level dataset using a short-time Fourier transform algorithm to convert the power data to the frequency domain. During the transformation, a Hanning window is selected, with a window length set to 256 points. Frequency domain analysis captures the complementary characteristics of high-frequency bands, which correspond to rapid changes within short time periods and can reflect… The system reflects the immediate synergy between the two; S53 processes hourly datasets, employing wavelet transform algorithms and selecting the db4 wavelet basis function to decompose the power data into different frequency components. The decomposition level is set to 4 levels. Mid-frequency band analysis is used to understand the characteristic change patterns, and the mid-frequency band corresponds to the stable changes within the time period, reflecting the medium-term synergistic effect between the two; S54 processes daily datasets, employing trend analysis algorithms and selecting linear or exponential fitting methods to fit the trend curve of the power data. The goodness of fit must be greater than 0.8. Low-frequency band analysis is used to analyze the long-term complementary characteristics, and the low-frequency band corresponds to the long-term change trend, reflecting the overall synergistic level between the two. The four steps combined achieve multi-dimensional quantitative analysis, comprehensively covering the complementary characteristics at different time scales.

[0091] like Figure 5 As shown, this embodiment of the invention also provides a distributed energy storage and flexible load complementarity characteristic analysis system. This system is applied to the aforementioned distributed energy storage and flexible load complementarity characteristic analysis method, and includes:

[0092] The multi-source data acquisition unit is used to collect the output power and charging / discharging status data of distributed energy storage devices, the power consumption and operating status parameters of flexible loads, as well as the voltage, frequency, and power flow data of the local power grid, and transmit the collected data to the data preprocessing unit.

[0093] The data preprocessing unit is connected to the multi-source data acquisition unit. It classifies and filters the received data, removes abnormal data, establishes a raw database, and sends the raw database to the correlation analysis unit.

[0094] The correlation analysis unit, connected to the data preprocessing unit, analyzes the correlation between distributed energy storage and flexible loads based on the original database, determines calibration parameters, and transmits the calibration parameters to the evaluation index system construction unit.

[0095] The evaluation index system construction unit is connected to the correlation analysis unit. It constructs the evaluation index system based on the calibration parameters and sends the evaluation index system to the multi-time-scale quantitative analysis unit. The multi-time-scale quantitative analysis unit is connected to the evaluation index system construction unit. It uses a multi-time-scale analysis algorithm to perform quantitative analysis on complementary characteristics and transmits the quantitative analysis results to the analysis report generation unit.

[0096] The analysis report generation unit is connected to the multi-timescale quantitative analysis unit. It generates analysis reports based on the quantitative analysis results. This unit also has report storage and export functions, which can store the analysis reports locally or export them to external devices.

[0097] This invention comprehensively covers the output power and charging / discharging status data of distributed energy storage devices, the real-time power consumption and operating status parameters of flexible loads, and the voltage, frequency, and power flow data of the local power grid through multi-source data acquisition. After classifying, filtering, and eliminating abnormal data, a high-quality raw database is constructed to provide reliable data support for subsequent analysis. Simultaneously, multi-timescale analysis algorithms are employed to quantitatively analyze complementary characteristics at the minute, hour, and day levels. This not only captures the complementary patterns under short-term power fluctuations but also grasps the characteristic changes during long-term operation. It completely solves the problem of existing technologies having limited analytical results due to a single timescale, which cannot support multi-scenario scheduling decisions. This provides comprehensive data support for the operation and control of the power grid at different times.

[0098] This analytical method and system, when determining key parameters affecting complementary characteristics, integrates the results of linear and nonlinear correlation analyses to fully explore the complex correlation characteristics between distributed energy storage and flexible loads, avoiding the omission of key parameters due to a single analytical approach. In constructing the evaluation index system, candidate indicators are first selected with reference to industry standards. Then, domain experts are invited to determine the index weights using the analytic hierarchy process (AHP). Consistency checks are also conducted to ensure the scientific nature of the weight settings. Ultimately, a comprehensive system of core indicators, including power matching degree, response synchronization, and capacity utilization, is formed. This addresses the problems of incomplete index settings and lack of scientific verification of weights in existing technologies. The generated analysis report clearly identifies complementary advantage areas and areas requiring optimization, providing precise guidance for the synergistic optimization of both, and helping to improve grid operation stability and energy utilization efficiency.

[0099] In the description of this invention, it should be noted that, unless otherwise explicitly specified and limited, the terms "set," "install," "connect," "link," and "fix" should be interpreted broadly. For example, they can refer to a fixed connection, a detachable connection, or an integral connection; they can refer to a mechanical connection or an electrical connection; they can refer to a direct connection or an indirect connection through an intermediate medium; and they can refer to the internal communication between two components. Those skilled in the art will understand the specific meaning of the above terms in this invention based on the specific circumstances.

[0100] Although embodiments of the invention have been shown and described, it will be understood by those skilled in the art that various equivalent changes, modifications, substitutions and alterations can be made to these embodiments without departing from the principles and spirit of the invention, the scope of which is defined by the appended claims and their equivalents.

Claims

1. A method for analyzing the complementary characteristics of distributed energy storage and flexible loads, characterized in that, Includes the following steps: S1 collects output power data, charging and discharging status data of distributed energy storage devices under different operating conditions, as well as real-time power consumption data and operating status parameters of industrial load, HVAC load and charging load in flexible loads, and simultaneously obtains voltage, frequency and power flow data of the local power grid. S2, classify and filter the collected data, remove abnormal data, and establish the original database of distributed energy storage and flexible load; S3. Based on the data in the original database, analyze the correlation between the charging and discharging characteristics of distributed energy storage devices and the power change characteristics of different types of flexible loads, and determine the calibration parameters that affect the complementary characteristics. S4. Based on the calibration parameters, construct an evaluation index system for the complementary characteristics of distributed energy storage and flexible loads. This system includes indicators related to power matching degree, response synchronization, and capacity utilization. S5 employs a multi-timescale analysis algorithm to quantitatively analyze the complementary characteristics of distributed energy storage and flexible loads at minute, hour, and day timescales, respectively. S6 generates an analysis report on the complementary characteristics of distributed energy storage and flexible loads based on the quantitative analysis results at different time scales, clarifying the advantageous areas and areas to be optimized for complementarity under different operating conditions.

2. The method for analyzing the complementary characteristics of distributed energy storage and flexible loads according to claim 1, characterized in that, When analyzing the correlation in S3, the following model is used: ， in, The combined power value of complementary characteristics at time t. For distributed energy storage power weighting coefficients, Let be the output power of the distributed energy storage at time t. This is the weighting factor for the total power of flexible loads. For the types and quantities of flexible loads, Let be the power consumption of the i-th type of flexible load at time t. For local power grid power weighting coefficients, Let t be the local power grid input power at time t.

3. The method for analyzing the complementary characteristics of distributed energy storage and flexible loads according to claim 1, characterized in that, When constructing the evaluation index system in S4, the power matching degree index adopts the following model: ， in, For power matching degree, These represent the start and end times of the analysis period, respectively. The value of the flexible load power fluctuation at time t; The response synchronization index uses the following model: ， in, In response to synchronization, Let t be the distributed energy storage response delay time. Let t be the flexible load response delay time. This is the preset maximum response latency threshold.

4. The method for analyzing the complementary characteristics of distributed energy storage and flexible loads according to claim 1, characterized in that, The model used for minute-scale time-scale quantitative analysis in S5 is as follows: ， in, This is the quantitative analysis value for the k-th minute time period. This refers to the number of data collections within a minute-level time period. Let J be the weight of the data collected in the j-th time. Let j be the time interval for the j-th data collection. This represents the distributed energy storage power collected in the j-th time interval within the k-th minute period. This represents the total power of the flexible load collected in the j-th time interval within the k-th minute period.

5. The method for analyzing the complementary characteristics of distributed energy storage and flexible loads according to claim 1, characterized in that, The model used for the daily-scale timescale quantitative analysis in S5 is as follows: ， in, These are daily quantitative analysis values. The total number of seconds in a day. The time weighting coefficient for the h-th hour. Let t be the distributed energy storage capacity utilization rate. Let t be the utilization rate of flexible load capacity demand.

6. The method for analyzing the complementary characteristics of distributed energy storage and flexible loads according to claim 1, characterized in that, When generating the analysis report, S6 uses the following model to determine the advantageous regions: ， in, For the collection of advantageous regions, The complementary characteristic quantization value is given when the power is P at time t. For quantization threshold, These represent the minimum and maximum output power of distributed energy storage, respectively. These represent the minimum and maximum power consumption for flexible loads, respectively.

7. The method for analyzing the complementary characteristics of distributed energy storage and flexible loads according to claim 1, characterized in that, S3 includes the following steps: S31. Extract the continuous 72-hour charging and discharging power data of distributed energy storage devices and the continuous 72-hour power consumption data of different types of loads in flexible loads from the original database, align them by timestamp, and form a paired data sequence. S32, Perform sliding window processing on the paired data sequence, with the window size set to 30 minutes, calculate the covariance between the distributed energy storage power and the flexible load power within each window, and preliminarily determine the degree of linear correlation between the two; S33 introduces a nonlinear correlation analysis method to calculate the mutual information value between the rate of change of the charge and discharge state of distributed energy storage and the rate of change of the power of flexible load, and explores the nonlinear correlation characteristics between the two. S34. Based on the combined results of linear covariance and nonlinear mutual information, parameters with a correlation degree higher than a preset threshold are selected and identified as calibration parameters that affect complementary properties.

8. The method for analyzing the complementary characteristics of distributed energy storage and flexible loads according to claim 1, characterized in that, S4 includes the following sub-steps: S41. Referring to the power system operation specifications and energy management standards, power balance, response speed consistency, capacity adaptability, operation stability, and economic indicators are initially selected as candidate evaluation indicators. S42. Using the analytic hierarchy process, construct an indicator weight judgment matrix, invite 5-8 experts in the power system field to compare and score the importance of each indicator in the candidate set pairwise, and calculate the weight value of each indicator. S43. Based on the weight values, indicators with weights below 0.05 are removed, and indicators with higher weights are retained to form a preliminary evaluation indicator system. S44. Conduct a consistency test on the preliminary evaluation index system. If the test passes, it will be determined as the final evaluation index system for the complementary characteristics of distributed energy storage and flexible load. If it fails, the indicators and weights will be readjusted, and the test will be repeated until it passes.

9. The method for analyzing the complementary characteristics of distributed energy storage and flexible loads according to claim 1, characterized in that, S5 includes the following steps: S51, the data in the original database is divided into minute-level, hour-level, and day-level datasets according to the time scale. The minute-level dataset is collected every 5 minutes, the hour-level dataset is collected every hour, and the day-level dataset is summarized every 24 hours. S52 uses the short-time Fourier transform algorithm to convert power data to the frequency domain for minute-level datasets, and analyzes the complementary characteristics of distributed energy storage and flexible load power in the high-frequency band. S53, for hourly datasets, uses wavelet transform algorithm to decompose different frequency components of power data and analyzes the variation law of complementary characteristics in the mid-frequency band; S54 uses a trend analysis algorithm to fit the power data change trend curve for daily datasets, analyzes the long-term complementary characteristics of low-frequency bands, and completes multi-timescale quantitative analysis.

10. A system for analyzing the complementary characteristics of distributed energy storage and flexible loads, characterized in that, The system is applied to the method for analyzing the complementary characteristics of distributed energy storage and flexible loads as described in any one of claims 1-9, comprising: The multi-source data acquisition unit is used to collect the output power and charging / discharging status data of distributed energy storage devices under different operating conditions, the real-time power consumption data and operating status parameters of industrial load, HVAC load and charging load in flexible loads, as well as the voltage, frequency and power flow data of the local power grid, and transmit the collected data to the data preprocessing unit. The data preprocessing unit is connected to the multi-source data acquisition unit. It classifies and filters the acquired data, removes abnormal data, establishes a raw database, and sends the raw database to the correlation analysis unit. The correlation analysis unit, connected to the data preprocessing unit, analyzes the correlation between distributed energy storage and flexible loads based on the original database, and determines the calibration parameters that affect the complementary characteristics. The evaluation index system construction unit is connected to the correlation analysis unit. Based on the calibration parameters, an evaluation index system for the complementary characteristics of distributed energy storage and flexible load is constructed. This system includes indicators related to power matching degree, response synchronization, and capacity utilization. The multi-timescale quantitative analysis unit is connected to the evaluation index system construction unit. It adopts a multi-timescale analysis algorithm to quantitatively analyze the complementary characteristics of distributed energy storage and flexible load at minute, hour, and day timescales respectively. The analysis report generation unit is connected to the multi-timescale quantitative analysis unit. Based on the quantitative analysis results at different time scales, it generates an analysis report on the complementary characteristics of distributed energy storage and flexible load, clarifying the advantageous areas and areas to be optimized for the complementarity of the two under different operating conditions.

Images