Charging pile quantity configuration method based on big data analysis

Abstract

This invention discloses a method for configuring the number of charging piles based on big data analysis, belonging to the field of smart energy management technology. The method includes the following steps: collecting traffic flow data, vehicle charging record data, power grid load curve data, and meteorological time data; uniformly numbering the timestamps of all data to generate a time misalignment sensitive list; and comparing data from different sources at minute-level time intervals based on the time misalignment sensitive list, calculating the offset position of adjacent data in the time series, and forming a rhythm superposition contact point table. This invention achieves time unification of multi-source data and elimination of false peaks by establishing a time misalignment sensitive list, a rhythm superposition contact point table, and a false peak identification index, thereby improving the accuracy and spatial matching of charging pile configuration. Furthermore, by combining a high-mismatch area list and a dynamic sampling control mechanism, it performs quota rollback and optimization adjustments based on real-time traffic and charging demand, achieving intelligent and dynamic balance in charging pile configuration.

Description

Technical Field

[0001] This invention relates to the field of smart energy management technology, specifically to a method for configuring the number of charging piles based on big data analysis. Background Technology

[0002] Charging pile quantity allocation based on big data analytics refers to the process of scientifically determining the required number and type of charging piles for different regions by comprehensively analyzing multi-dimensional data such as urban traffic, user travel, energy supply, and grid load, including vehicle charging demand distribution, time-of-day characteristics, dwell time, electricity price fluctuations, and regional power supply capacity. This method utilizes big data technology to collect, clean, cluster, and predictively model historical and real-time data, thereby spatially identifying high-demand areas, temporally depicting peak charging patterns, and dynamically generating optimal allocation schemes by combining power supply capacity and land resource constraints. This data-driven approach enables a shift in charging infrastructure construction from experience-based planning to intelligent and refined allocation, improving energy efficiency and user charging convenience.

[0003] The existing technology has the following shortcomings:

[0004] In existing technologies, multi-source data is prone to periodic misalignment during time synchronization, especially in the fusion stage of multi-source information such as traffic flow, vehicle trajectories, charging orders, power grid load, and meteorological parameters. Different sampling frequencies and upload delays cause timestamp drift. This drift generates false peak charging demand signals at periodic overlap points. During the identification process, the model may misjudge these peaks as actual periods of concentrated charging, thus incorrectly increasing the number of charging piles in a certain area. Such misjudgments lead to some areas having charging piles idle for extended periods, while areas with genuine high demand have insufficient charging piles, resulting in a significant spatial mismatch.

[0005] The information disclosed in the background section is only intended to enhance the understanding of the background of this disclosure, and therefore may include information that does not constitute prior art known to those skilled in the art. Summary of the Invention

[0006] The purpose of this invention is to provide a method for configuring the number of charging piles based on big data analysis, so as to solve the problems in the background art mentioned above.

[0007] To achieve the above objectives, the present invention provides the following technical solution: a method for configuring the number of charging piles based on big data analysis, comprising the following steps:

[0008] Collect traffic flow data, vehicle charging record data, power grid load curve data and meteorological time data, assign a unified number to the timestamps of all data, and generate a time misalignment sensitive list to establish a time benchmark to support subsequent comparison and analysis.

[0009] Based on the time misalignment sensitive list, data from different sources are compared item by item at minute-level time intervals to calculate the offset position of adjacent data in the time series and form a rhythm superposition touch point table.

[0010] By using a rhythm overlay contact table, statistical analysis is performed on the recurrence patterns of each contact point to identify periodic echo points and locate peak locations of false charging demand, generating a false peak identification index to correct abnormal peak values ​​in the data.

[0011] Based on the false peak identification index, the regional charging pile configuration map is re-analyzed, the configuration quantity deviation of different regions is calculated, and a list of high misconfiguration areas is compiled as the target input for the dynamic adjustment stage.

[0012] Based on the list of high mismatch areas, combined with real-time traffic data and charging demand changes, dynamic adjustment operations are performed. The data sampling rhythm is corrected by alternating between forward and reverse rhythms and short-term suspension of data collection windows. Quota rollback is implemented according to demand fluctuations, and the number of charging piles is adjusted in real time to ensure that the charging pile configuration results are consistent with actual demand.

[0013] The preferred steps for generating the time-dislocation sensitive list are as follows:

[0014] Traffic flow data, vehicle charging record data, power grid load curve data, and meteorological time data are continuously collected at fixed time sampling intervals to form a multi-source raw time series data set containing timestamps.

[0015] Using a unified time base point as a reference, all timestamps are numbered, various types of data are mapped to the same time axis system, and missing points are filled in by time interpolation to form a unified time series dataset.

[0016] Perform time offset correlation analysis on the data after unified numbering to identify the time drift intervals of data from different sources and record the start and end numbers of the offset and the offset direction;

[0017] Summarize the offset information, generate a time misalignment sensitive list in chronological order, and record the offset source, offset magnitude and frequency for subsequent data comparison and analysis.

[0018] Preferably, the steps for forming the rhythm superimposed contact table are as follows:

[0019] Based on the unified time numbering of the time misalignment sensitive list, traffic flow data, vehicle charging record data, power grid load curve data and meteorological time data are aligned minute by minute, and the data from each source under each time number are combined into the same time unit.

[0020] The difference between data groups with adjacent time numbers is calculated according to the time number order, the offset position of data from different sources in the time series is identified, and the offset start number, offset end number and offset duration are recorded.

[0021] The identified time offset intervals are analyzed for patterns, and the offset direction, duration and frequency of occurrence of data from each source are statistically analyzed to form a time offset pattern table.

[0022] Based on the time offset pattern table, integrate time number intervals with synchronous or periodic characteristics, record the offset direction, offset magnitude and offset frequency of each source data, and generate a rhythm superposition touch point table.

[0023] Preferably, the steps for generating the pseudo-peak identification index are as follows:

[0024] Based on the time number, offset direction, offset duration, offset amplitude, and offset frequency recorded in the rhythm superposition contact table, statistical analysis is performed on the repetition distribution of various types of contacts to form an offset data sequence arranged by time number;

[0025] Perform time interval analysis on the offset data sequence, extract periodic echo features, record the offset source combination, repetition interval, duration and frequency of occurrence, and form a list of periodic echo analysis.

[0026] Based on the periodic echo analysis list and the time misalignment sensitivity list, identify false charging demand peaks formed by time drift superposition, establish a false peak location list and record the time number range, offset direction consistency and echo period length;

[0027] Based on the list of false peak locations, all false peak information is integrated, and a false peak identification index is generated in chronological order, recording the offset source category, offset direction, echo period, offset duration, and frequency of occurrence.

[0028] Preferably, during the generation of the false peak identification index, the offset source categories recorded in the false peak location list include traffic flow data, vehicle charging record data, power grid load curve data, and meteorological time data. The offset direction consistency and echo period length are marked by corresponding time numbers, thereby achieving accurate location and periodic feature identification of false peaks.

[0029] The preferred steps for generating the list of high mismatch regions are as follows:

[0030] Based on the time number, offset source type, offset direction, echo period, offset duration and geographical location identifier in the false peak identification index, it is mapped with the spatial nodes of the regional charging pile configuration map to form a list of false peak affected areas, and records the area number, false peak time and configuration change trend corresponding to each false peak.

[0031] Based on the list of areas affected by false peaks, the configuration quantity of each area during the false peak period and the baseline period is compared and calculated. The configuration quantity deviation, load change magnitude and utilization rate change value are recorded to form a regional configuration quantity deviation table.

[0032] Based on the principles of geographical proximity and consistency of pseudo-peak cycles, spatial aggregation analysis is performed on the deviation results to generate a regional deviation aggregation table, which records the deviation type, duration, and pseudo-peak influence cycle.

[0033] Based on the regional deviation aggregation table, each aggregation unit is summarized and organized, and the configuration error type, error magnitude, error duration period and pseudo-peak interference source are recorded to generate a list of high misconfiguration regions.

[0034] Preferably, in the regional deviation aggregation analysis, the correlation judgment of geographically adjacent regions is made based on the power grid load transmission relationship. When the number of charging piles in a region increases while the load in adjacent regions decreases during the pseudo-peak period, it is recorded as a spatial resource migration phenomenon and marked as an uneven resource allocation region in the list of high mismatched regions, so as to be used for priority correction in the dynamic adjustment stage.

[0035] Preferably, the dynamic adjustment operation execution process is as follows:

[0036] Based on the list of high mismatched areas, real-time status identification is performed on areas identified as over-configured or under-configured. Real-time traffic data and real-time charging demand data are matched with area numbers to form a real-time status matching matrix that records traffic operation status and changes in energy demand.

[0037] Based on the real-time state matching matrix, the data sampling process in the high mismatch area is corrected by alternating forward and reverse rhythm sampling. By controlling the sampling method of increasing and decreasing time number, the data distribution is kept balanced, and the rise and fall patterns of demand changes are recorded.

[0038] During the alternating sampling process of forward and reverse rhythms, a short-term pause window is set to adjust the sampling rhythm by pausing, preventing time offset caused by high-frequency sampling and maintaining the continuity and stability of the data time axis;

[0039] Based on the mismatch direction and deviation range in the list of high mismatch areas, and combined with real-time traffic and charging demand changes, quota rollback is performed to adjust the number of charging piles in each area and record the adjustment results.

[0040] The technical effects and advantages provided by the present invention in the above technical solution are as follows:

[0041] This invention introduces a time misalignment sensitivity list, a rhythm overlay contact point table, and a spurious peak identification index during multi-source data fusion. This enables unified benchmark alignment of multi-source data in the time dimension, effectively identifying and eliminating false peak signals caused by time drift, thus fundamentally avoiding time mismatch problems during data fusion. This approach makes the distribution of charging demand more realistic and reliable, preventing the model from mistakenly identifying spurious peaks as real demand during peak identification. This improves the accuracy and spatial distribution rationality of charging pile allocation, achieving synchronous matching between energy allocation and actual travel patterns.

[0042] This invention utilizes a linkage analysis of a high-mismatch area list and real-time traffic and charging demand data, combined with a dynamic sampling control mechanism that alternates between forward and reverse traffic flows and short-term sampling pauses, to achieve real-time optimization and quota rollback of charging pile deployment. This allows deployment decisions to respond promptly to changes in traffic flow and fluctuations in power load. This method transforms charging infrastructure construction from static planning to an adaptive dynamic adjustment mode, improving the utilization efficiency and deployment flexibility of charging resources. It ensures that charging pile layout continuously aligns with actual usage needs, thereby maintaining a dynamic balance between urban energy supply and vehicle charging behavior. Attached Figure Description

[0043] To more clearly illustrate the technical solutions in the embodiments of this application or the prior art, the drawings used in the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments recorded in this invention. For those skilled in the art, other drawings can be obtained based on these drawings.

[0044] Figure 1 This is a flowchart of the charging pile quantity configuration method based on big data analysis according to the present invention. Detailed Implementation

[0045] Exemplary embodiments will now be described more fully with reference to the accompanying drawings. However, these exemplary embodiments can be implemented in many forms and should not be construed as limited to the examples set forth herein; rather, they are provided so that the description of this disclosure will be more complete and fully convey the concept of the exemplary embodiments to those skilled in the art.

[0046] This invention provides, for example Figure 1 The method for configuring the number of charging piles based on big data analysis, as shown, includes the following steps:

[0047] Collect traffic flow data, vehicle charging record data, power grid load curve data and meteorological time data, assign a unified number to the timestamps of all data, and generate a time misalignment sensitive list to establish a time benchmark to support subsequent comparison and analysis.

[0048] To achieve unified and accurate correspondence of multi-source data across time dimensions, a multi-source time data synchronization construction technology path centered on time numbering is adopted. This process involves hierarchical collection, unified numbering, misalignment identification, and sensitive list generation of traffic flow data, vehicle charging record data, power grid load curve data, and meteorological time data. This ensures consistency of data from all sources across a unified timeline, laying the foundation for subsequent comparative analysis. The specific steps are as follows:

[0049] Continuous data collection is performed on traffic flow, vehicle charging records, power grid load curves, and meteorological time data. The collection process is based on fixed time sampling intervals, acquiring raw time-series data from monitoring terminals or data interfaces from multiple data sources. Traffic flow data is collected at road monitoring nodes, recording the number of vehicles passing through the monitoring section at different time points, average vehicle speed, vehicle type distribution, road occupancy, and traffic flow direction. Vehicle charging records are collected from the charging equipment platform database, recording the charging start time, charging end time, charging duration, single charging amount, charging pile location, and vehicle identification number. Power grid load curve data is collected by distribution area, recording load power values, voltage levels, current intensity, power supply stability, and instantaneous load change trends at various times. Meteorological time data is collected from meteorological observation stations or meteorological service platforms, recording temperature, humidity, wind speed, precipitation, air pressure, solar radiation intensity, and wind direction at corresponding time points. All collected data are acquired within the same time sampling interval to ensure sampling consistency in subsequent time numbering. After the data collection is completed, a raw dataset containing four types of time-series information is formed. Each type of data contains a complete timestamp record for subsequent unified numbering.

[0050] The four types of data collected were processed using a unified time numbering system. During the numbering process, a unified time base point was used as a reference to map all timestamps to the same timeline system. The starting point of the time numbering was determined by the earliest collection time, and the ending point was determined by the end time of the collection cycle. The numbering order was arranged chronologically, assigning consecutive numbers to each time node to ensure the consistency of the time sequence throughout the data series. Traffic flow data, vehicle charging record data, power grid load curve data, and meteorological time data were all mapped according to this time numbering system, ensuring that each record corresponds to a unique number. For data with different sampling time intervals, missing time points were filled in using time interpolation to maintain completeness and continuity within the time numbering system. After numbering, all source data were indexed using the unified numbering system, forming a unified time-series dataset that could be matched along the time dimension. The key to this process is unifying data sources from different sampling frequencies into the same numbering system, enabling traffic, energy, and meteorological data to achieve one-to-one correspondence in subsequent comparisons and analyses, thereby eliminating potential offset problems caused by differences in time sampling.

[0051] After establishing a unified time numbering system, time offset correlation analysis is performed on the multi-source data following the numbering to identify sensitive locations with time drift. The analysis uses the unified time numbering as the main thread, extracting the numerical changes of traffic flow data, vehicle charging record data, power grid load curve data, and meteorological time data under the same number, and comparing these trends horizontally. By analyzing the differences in data changes between consecutive numbers, regions exhibiting asynchronous phenomena within certain time periods are identified. For example, if traffic flow increases under a specific number but charging records do not change synchronously, or if the power grid load curve fluctuates within a certain number range but meteorological conditions do not change accordingly, a time offset can be identified for that time period. All detected offset number intervals are recorded to form a time misalignment identification record. This record includes the offset start number, offset end number, offset duration, offset source type, and offset direction information, thus accurately describing the differences in the time dimension between different data sources. This correlation analysis based on unified time numbering can effectively identify time drift points generated during periodic sampling, providing a complete characterization of time offset features and a data basis for the subsequent generation of a sensitive list.

[0052] After time offset identification is completed, all identification results are compiled and summarized to generate a time misalignment sensitive list. The list uses a unified time number as the primary key, summarizing the offset information of traffic flow, vehicle charging records, power grid load curves, and meteorological time data corresponding to each number. The list records the offset source, offset magnitude, offset frequency, and offset duration interval. The list is arranged chronologically, allowing for quick querying and tracking of the data offset characteristics corresponding to each number. The time misalignment sensitive list not only preserves the time offset distribution of various data types within the sampling period but also records the mutual influence relationships between different offset sources. For example, within the same number interval, if the offset direction of traffic flow and power grid load data is consistent, while the change in meteorological data lags, the offset in that interval may be related to the load response delay caused by concentrated vehicle travel. By explicitly marking such offset characteristics in the list, a direct reference for data comparison can be provided for subsequent analysis stages. The formation of the list provides a unified reference benchmark for the entire multi-source data system in the time dimension, allowing any subsequent rhythm overlay analysis, false peak identification, and regional configuration correction to be performed based on the number index in this list.

[0053] Based on the time misalignment sensitive list, data from different sources are compared item by item at minute-level time intervals to calculate the offset position of adjacent data in the time series and form a rhythm superposition touch point table.

[0054] To ensure synchronized comparison of traffic flow data, vehicle charging records, power grid load curves, and meteorological data from different sources across time dimensions, a minute-by-minute analysis and offset identification is performed on the multi-source data based on a time misalignment sensitivity list. The specific implementation steps are as follows:

[0055] Based on the unified time number recorded in the time misalignment sensitive list, traffic flow data, vehicle charging record data, power grid load curve data, and meteorological time data are aligned minute by minute. This process uses the earliest time number in the list as the starting point and arranges the data from each source in chronological order, with minutes as the interval unit. Traffic flow data is extracted according to minute number, including the number of vehicles passing through the road, average vehicle speed, lane utilization, road traffic direction ratio, and road segment traffic density. Vehicle charging record data is extracted according to minute number, including the number of charging starts, the number of charging completions, the duration of a single charging session, the charging amount per unit time, the geographical location of the charging pile, and the charging vehicle identification. Power grid load curve data is extracted according to minute number, including load power, voltage level, current intensity, load response delay time, power supply fluctuation amplitude, and distribution area identification. Meteorological time data is extracted according to minute number, including temperature, humidity, wind speed, precipitation, air pressure, wind direction, and solar radiation intensity. After all data is extracted, the different source data under each time number are combined into data groups of the same time unit, forming a minute-level multi-source data sequence indexed by the time number. When a certain type of data has missing records at a specific time number, it is linearly extended according to the data trends of adjacent time numbers, ensuring the continuity and comparability of the time series of various data types. This process ensures that all source data establish a complete correspondence on a unified timeline, with minutes as the basic unit.

[0056] After completing minute-level time alignment, based on the numbering order in the time misalignment sensitivity list, the differences between adjacent data groups of each time number are calculated to identify the temporal offset positions of data from different sources. This step uses two consecutively numbered adjacent data groups as analysis units to calculate the trends and directions of change in traffic flow, charging records, power grid load, and meteorological data between adjacent time numbers. For traffic flow, the trend of increase or decrease in the number of vehicles passing through, the direction of change in lane utilization, and the rate of change in average vehicle speed are calculated between adjacent time numbers. For vehicle charging records, the trend of change in the number of charging starts and finishes, the direction of increase or decrease in charging amount per unit time, and the trend of change in charging duration are calculated between adjacent time numbers. For power grid load curves, the direction of change in load power, the trend of change in voltage fluctuations, the magnitude of change in current intensity, and the delay in load response are calculated between adjacent time numbers. For meteorological time data, the direction of temperature change, the rate of change in humidity, the trend of change in wind speed and direction, the trend of increase or decrease in precipitation, and the direction of change in air pressure are calculated between adjacent time numbers. By comparing the changing trends of these data from different sources under adjacent time numbers, when the changing direction of a certain type of data is inconsistent with the changing direction of other data, this time number interval is marked as a time offset interval. The identification result of the offset interval uses the time number as the core recording unit, clearly indicating the source of the offset, the start number of the offset, the end number of the offset, and the duration of the offset.

[0057] After identifying the offset intervals, a pattern analysis is performed on all identified time offset intervals to extract the rhythmic superposition characteristics of data from different sources in the time series. This process is based on the offset relationship of consecutive numbering, statistically classifying the offset direction according to time sequence to discover periodic synchronization patterns between data from different sources. For example, when traffic flow data shows an increasing trend in multiple consecutive numbered intervals, while power grid load data shows a delayed increase in intervals several numbers after these intervals, it indicates a stable rhythmic superposition relationship between traffic and power load. When the direction of change in vehicle charging records in several consecutive numbered intervals is synchronized with the direction of temperature change in meteorological data, it indicates that the charging demand in this area is periodically affected by meteorological factors. In this process, the direction of change, duration, and frequency of occurrence of each source of data are comprehensively analyzed, and the number of offset repetitions and time intervals between data pairs are statistically analyzed to form a time offset pattern table. This table records the superposition relationship of data from different sources in the time dimension, providing a regular basis for the generation of rhythmic touchpoints.

[0058] Based on the statistical results of the time offset pattern table, all time number intervals with synchronous or periodic characteristics are integrated and recorded to generate a rhythm overlay contact table. The rhythm overlay contact table uses the time number as its core field, recording the time offset direction, duration, magnitude, and frequency of occurrence between traffic flow, vehicle charging records, power grid load curves, and meteorological time data under each time number. Each row in the table corresponds to a time number interval, detailing the direction of change of each data source within that interval and their synchronous relationships. For example, when traffic flow increases within a time number interval while power grid load increases two time number delays later, the rhythm overlay contact table records the delay correlation between traffic and power within that interval; when vehicle charging records and meteorological data change in the same direction within the same time number interval, the rhythm overlay contact table records their synchronous relationship and the duration of the time number interval. In this way, the rhythm overlay contact table preserves the time offset relationships of data from different sources in a structured form, allowing the time coupling patterns between multi-source data to be directly referenced. After the rhythm overlay contact table is formed, the multi-source offset information for each time number can be quickly located and analyzed, providing a reliable time series basis for the subsequent pseudo-peak identification stage.

[0059] By using a rhythm overlay contact table, statistical analysis is performed on the recurrence patterns of each contact point to identify periodic echo points and locate peak locations of false charging demand, generating a false peak identification index to correct abnormal peak values ​​in the data.

[0060] To fully utilize the time offset information of multi-source data recorded in the rhythm superposition contact table, thereby enabling the identification of periodic echo points, the location of false charging demand peaks, and the generation of a false peak identification index, the specific implementation steps are as follows:

[0061] Based on the time number, offset direction, offset duration, offset amplitude, and offset frequency recorded in the rhythm overlay contact table, statistical analysis was performed on the repetitive distribution of various contact types. The rhythm overlay contact table is derived from minute-level time comparison results of multi-source data, including the offset relationships of traffic flow data, vehicle charging record data, power grid load curve data, and meteorological time data under different time numbers. To identify the repetitive patterns of these offsets in the time dimension, all contact points need to be categorized and summarized using time numbers as the main line. Specifically, the delay relationship between traffic flow and power grid load is extracted separately to form a traffic and power offset data sequence; the synchronous change records between vehicle charging records and meteorological time data are extracted to form a charging and meteorological offset data sequence; and the synchronous trend between traffic flow and vehicle charging records is extracted to form a traffic and charging offset data sequence. Each type of offset data sequence is arranged in time number order, forming a continuous time distribution structure. During the statistical process, the number of contact occurrences, the duration of each occurrence, the time interval between different offsets, and their distribution frequency throughout the entire sampling period were calculated for each type of offset sequence. This statistical analysis can preliminarily determine which offset relationships in multi-source data have recurring characteristics, and provide a basis for subsequent periodic pattern identification.

[0062] After obtaining the statistical results of repeated contact points, a thorough analysis of the temporal patterns of contact point distribution is conducted to extract periodic echo characteristics. Periodic echo refers to the phenomenon that a specific type of offset relationship appears multiple times in a time series at fixed or similar intervals. To identify this pattern, the time numbering intervals of each type of offset sequence need to be organized, and a time interval sequence is established based on the time number difference. This time interval sequence reflects the interval length and occurrence pattern of the same offset relationship in different time periods. When multiple identical or similar interval lengths exist in the time interval sequence, it can be determined that the offset relationship has periodic echo characteristics. For example, in the offset sequence of traffic flow and power grid load, if the offset delay of both appears at fixed numbering intervals in multiple numbering intervals, it indicates that changes in traffic flow have a periodic impact on power grid load; in the offset sequence of vehicle charging records and meteorological data, if the synchronous rise of both reappears after the same time interval, it indicates that periodic changes in meteorological conditions affect charging demand. Through the time interval analysis of various types of offset sequences, multiple periodic echo patterns can be extracted. Each pattern includes the combination of offset sources, recurrence interval, duration, average offset magnitude, and frequency of occurrence, forming a list of periodic echo analysis patterns.

[0063] After identifying the periodic echo characteristics, these periodic patterns need to be screened and analyzed to determine which periodic echoes are false charging demand peaks caused by time drift. This step uses the periodic echo analysis list and the time misalignment sensitivity list as references to compare the synchronization relationship of data from various sources one by one. When traffic flow, vehicle charging records, grid load curves, and meteorological time data all show an upward trend within a certain time interval, while meteorological conditions remain stable within that interval, traffic fluctuations show no abnormal changes, and grid load increases after a delay, it can be determined that the peak is a false peak signal caused by the superposition of data upload delay and sampling time drift. If this phenomenon occurs repeatedly in the periodic echo pattern, it indicates that the false peak has a fixed time periodic characteristic. In the process of false peak location, a complete record needs to be established for each false peak, including the time number range, the data sources involved, the consistency of the offset direction, the offset duration, the echo period length, and the number of times the false peak occurs. By summarizing these characteristic information to form a false peak location list, the specific location and formation pattern of false peaks appearing on the time axis of multi-source data can be clearly described. The generation of this list allows false peaks to be separated from ordinary time fluctuations, providing a precise basis for the subsequent establishment of a false peak index.

[0064] Based on the list of false peak locations, all identified false peaks are integrated to establish a false peak identification index. The index uses a time number as a unique identifier and records various characteristic information of false peaks in a structured manner. Each index entry corresponding to a time number includes the offset source category, offset direction, echo period, offset duration, offset amplitude, frequency of occurrence, and the time difference between the offset and the true peak value. For ease of subsequent analysis, the false peak identification index is arranged in chronological order of time numbers and the duration of false peaks is marked by time span. When the same offset relationship repeatedly occurs in different time intervals, the index records the number of repetitions and their time intervals, representing the periodic stability of the false peak. After the false peak identification index is established, the temporal location and source relationship of all false peaks in the multi-source data are recorded by the system. This index can be directly called in subsequent data correction stages. By processing the time intervals marked in the index, false peaks can be removed or weakened in the overall data analysis, thereby improving the accuracy and reliability of multi-source data fusion analysis.

[0065] Based on the false peak identification index, the regional charging pile configuration map is re-analyzed, the configuration quantity deviation of different regions is calculated, and a list of high misconfiguration areas is compiled as the target input for the dynamic adjustment stage.

[0066] Based on the pseudo-peak identification index, a comprehensive re-analysis of the charging pile configuration maps for each region is conducted to accurately calculate the configuration quantity deviation in different regions. This data is then compiled into a list of high-mismatch areas, providing precise target input for the dynamic adjustment phase. The entire process includes mapping the pseudo-peak index to the regional configuration map, calculating the configuration quantity deviation, aggregating and analyzing regional deviations, and generating the list of high-mismatch areas. Through this series of operations, the impact of false peaks in the temporal dimension and the resource allocation status in the spatial dimension can be integrated for analysis, revealing the resource mismatch phenomenon caused by pseudo-peak interference and presenting the configuration deviation in each region in a quantitative form. The specific steps are as follows:

[0067] Based on the time index, offset source type, offset direction, echo period, offset duration, and geographic location identifier contained in the false peak identification index, and by referring to the spatial node information of the regional charging pile configuration map, the mapping operation between the false peak index and the configuration map is completed. The regional charging pile configuration map records the geographic boundaries, charging pile distribution density, number of installed charging piles, number of operating charging piles, charging load per unit time, number of vehicles, grid power supply capacity, and average utilization rate of each zone within the city. During the mapping process, the geographic coordinates in the false peak identification index are first spatially aligned with the regional boundaries in the configuration map, and the time index corresponding to the false peak is matched with the time period in the configuration map, thereby locating the false peak impact event to a specific regional node. After alignment, time series correlation is performed on each marked regional node to extract the configuration change data of the region in the time periods before and after the appearance of the false peak, including the number of newly built charging piles, changes in the number of charging piles already in operation, fluctuations in average charging power, and changes in vehicle usage load. Through this mapping method, a correspondence between false peak events and regional configuration changes can be established at the spatial level. After mapping is completed, a list of areas affected by false peaks is generated. This list details the area number corresponding to each false peak, the time of occurrence of the false peak, the trend of changes in the number of charging piles in the area, the number of charging piles affected, and the direction of load fluctuation. This list provides specific temporal and spatial references for subsequent calculations of configuration quantity deviations.

[0068] Based on the list of areas affected by false peaks, deviation calculations were performed on the number of charging piles configured in each affected area. Deviation calculations were performed on a time-period basis, determining the degree of interference of false peaks on regional configuration results by comparing the configuration quantity during the false peak period with the baseline period. The baseline period was a stable period before the false peak occurred, using the average configuration quantity during that period as a reference value. For each region, the total number of charging piles, the number of operational piles, the average utilization rate, the charging power per unit time, the vehicle charging demand, the grid load change value, and the average waiting time during the false peak period were extracted and compared with the corresponding values ​​during the baseline period to calculate the magnitude of each change. If the configuration quantity during the false peak period was higher than the baseline value, and the charging load in the region did not increase accordingly, it indicated that the false peak led to over-configuration of resources; if the configuration quantity during the false peak period was lower than the baseline value, and the charging load and the number of vehicles in the region increased simultaneously, it indicated that the region experienced insufficient resource allocation. For each region, these deviation data were recorded to form a regional configuration quantity deviation table. The deviation table included the region number, the false peak time number, the configuration quantity deviation value, the load change magnitude, the charging pile utilization rate change value, and the comparison result with the grid power supply capacity. This calculation process can quantify the degree of configuration deviation in each region after being affected by spurious peaks, laying the foundation for judging misconfiguration trends.

[0069] After calculating the regional deviations, spatial aggregation analysis is performed on the deviation results for all regions to identify the set of regions where resource allocation errors occur simultaneously in both time and space. The aggregation analysis is based on the principles of geographical proximity and consistency of pseudo-peak cycles, grouping regions with adjacent geographical locations and the same or similar pseudo-peak numbers into regional aggregation units. For each regional aggregation unit, the average deviation, deviation direction consistency, deviation duration, and frequency of pseudo-peak interference within each region are calculated. If multiple regions within an aggregation unit continuously show an increasing trend in the number of allocated resources during the same pseudo-peak cycle, the unit is classified as an over-allocated aggregation area; if multiple regions within an aggregation unit continuously show a decreasing trend in the number of allocated resources during the same pseudo-peak cycle, the unit is classified as an under-allocated aggregation area. During the aggregation analysis, correlation analysis is also performed on the grid load transmission relationship between regions to determine whether the load transmission effect between adjacent regions is caused by pseudo-peaks. For example, when the number of charging piles increases in one region during a pseudo-peak period, while the load in adjacent regions decreases, it indicates that resources have undergone irrational spatial migration, a typical misallocation phenomenon. After the aggregation analysis is completed, a regional deviation aggregation table is generated. This table records in detail the geographical range, number of regions, deviation type, average deviation magnitude, deviation duration, spurious peak influence period, and corresponding region number range for each aggregation unit.

[0070] Based on the regional deviation aggregation table, each aggregation unit is summarized and organized to generate a list of high misconfiguration areas. The list is indexed by region number, and the system records the configuration error type, error magnitude, error duration, source of pseudo-peak interference, and number of affected charging piles for each region. For over-configured regions, the list records the total number of charging piles, utilization rate decrease percentage, load decrease per unit time, grid load redundancy percentage, and resource idle time during pseudo-peak periods. For under-configured regions, the list records the total number of charging piles, utilization rate increase percentage, average waiting time increase, trend of increasing number of vehicles queuing for charging, and local grid load concentration. The list is sorted by both geographical location and pseudo-peak number, allowing for continuous display of misconfigured regions within the same pseudo-peak period, facilitating unified processing during subsequent dynamic adjustment phases. After the list is generated, all regions experiencing configuration deviations due to pseudo-peak interference are clearly identified, forming a list of regions that can be directly used for adjustment decisions.

[0071] Based on the list of high mismatch areas, combined with real-time traffic data and charging demand change information, dynamic adjustment operations are performed. The data sampling rhythm is corrected by alternating forward and reverse rhythms and short-term suspension of data collection windows. Quota rollback is implemented according to demand fluctuations, and the number of charging piles is adjusted in real time to ensure that the charging pile configuration results are consistent with the actual demand.

[0072] To achieve dynamic adjustment of the list of high mismatched areas, and to continuously correct the number of charging piles in each area based on real-time traffic data and charging demand changes, the data sampling rhythm is optimized through alternating forward and reverse sampling methods and short-term data cessation windows. Furthermore, quota rollback is implemented based on real-time demand fluctuations to achieve a precise match between the number of charging piles and actual demand. The specific implementation steps are as follows:

[0073] Based on the list of high mismatch areas, real-time status identification is performed on each area identified as over- or under-configured, and this status is dynamically matched with traffic data and charging demand data. The list of high mismatch areas records the area number, mismatch type, mismatch magnitude, mismatch duration, source of false peak interference, geographical range of the area, and number of affected charging piles. In this step, real-time traffic data and real-time charging demand information are extracted for each area in the list, within the corresponding geographical coordinate range. Real-time traffic data includes the number of vehicle passages, average speed, lane occupancy rate, vehicle density, road segment travel time, and vehicle dwell distribution within a specific time interval. Real-time charging demand information includes the number of vehicle charging requests, charging start and end times, average charging power per unit time, single charging duration, charging pile utilization rate, and average waiting time. By mapping these two types of data one-to-one with the area numbers in the high mismatch area list, a real-time status matching matrix is ​​formed. This matrix uses the time number as an index to record the traffic operation status, vehicle energy demand changes, and power distribution load response of each area at the current time. This process enables a direct correspondence between spatial mismatch information and real-time dynamic data, providing a data foundation for subsequent sampling rhythm correction.

[0074] After real-time status matching is completed, the data sampling process in high mismatch areas undergoes alternating forward and reverse sampling rhythms to eliminate the cumulative effect of time drift during data collection. This step uses minutes as the smallest sampling unit and time numbers as the sequential identifier, achieving a balanced data time distribution through bidirectional sampling rhythm control. Forward sampling extracts data in ascending time number order, focusing on recording real-time status changes in traffic flow, vehicle charging power, grid load, and weather conditions, ensuring data collection covers the entire demand-rising phase. Reverse sampling extracts data in descending time number order, recording the changing patterns during the demand-falling phase and capturing the recovery trend at the end of demand fluctuations. Alternating forward and reverse sampling is executed in time segments, with each rhythm segment maintaining a fixed sampling duration. When sudden fluctuations in traffic flow, charging demand, or grid load are detected, the sampling rhythm direction is immediately switched to maintain a stable sampling frequency during periods of rapid demand change. This alternating method ensures symmetrical data distribution over time, avoiding the accumulation of time drift bias caused by continuous unidirectional sampling. This rhythm correction method ensures that the sampling data can fully reflect the real changes in the high mismatch area, providing a stable basis for setting subsequent short-term sampling stop windows.

[0075] To prevent excessively high data density or information lag caused by continuous sampling during the alternating forward and reverse sampling rhythm correction process, a short-term sampling pause window is set to adjust the sampling rhythm. The short-term sampling pause window is a sampling rhythm suspension mechanism for a specific time period, used to eliminate rhythm imbalances caused by high-frequency sampling or data upload delays. During implementation, based on the time number, data sampling is paused at the end of each pseudo-peak cycle or during periods of drastic change in traffic flow. The duration of the pause is determined based on the data density in the previous sampling cycle. No new data is collected during the pause; instead, the sampling data from the previous period is retained as a smooth transition, ensuring the continuity of the sampling timeline and data stability. After the sampling pause window closes, sampling resumes in the forward rhythm direction to capture data changes in subsequent periods. By adding a short-term sampling pause window during the sampling process, the cumulative risk of sampling rhythm deviation can be effectively reduced while ensuring data continuity, preventing sampling misalignment caused by time differences in multiple source data within the same time period. The short-term sampling pause window, combined with alternating forward and reverse sampling rhythms, creates a balanced closed loop in the time dimension, ensuring the consistency of the temporal correlation of multi-source data and providing reliable real-time data input for subsequent configuration adjustments.

[0076] After completing the sampling rhythm correction and suspension window adjustment, based on the mismatch direction and deviation magnitude recorded in the high mismatch area list, and combined with real-time traffic and charging demand changes, a quota rollback operation is executed to adjust the number of charging piles in real time. The quota rollback operation is implemented on a regional basis, balancing supply and demand by adjusting the effective number of charging piles within the region. For regions identified as over-configured in the list, based on the real-time decreasing trend of charging demand, the number of idle charging piles is reduced, temporarily removing them from operation or transferring them to adjacent areas with increasing demand. For regions identified as under-configured, based on real-time traffic flow density and the rate of increase in vehicle charging requests, the number of available charging piles is increased or the charging power allocation ratio is improved to meet instantaneous demand growth. During the quota rollback process, the configuration status of each region is monitored at minute intervals, recording the quantity changes, adjustment time, configuration difference before and after adjustment, regional power grid load changes, and vehicle waiting time changes for each adjustment. After each quota rollback cycle, the system stores the adjustment results in a dynamic adjustment record table for reference in subsequent optimization phases. The quota rollback mechanism ensures that high mismatch areas can correct configuration deviations in a timely manner when demand fluctuates, so that the number of charging piles in each area remains synchronized with real-time demand.

[0077] This invention introduces a time misalignment sensitivity list, a rhythm overlay contact point table, and a spurious peak identification index during multi-source data fusion. This enables unified benchmark alignment of multi-source data in the time dimension, effectively identifying and eliminating false peak signals caused by time drift, thus fundamentally avoiding time mismatch problems during data fusion. This approach makes the distribution of charging demand more realistic and reliable, preventing the model from mistakenly identifying spurious peaks as real demand during peak identification. This improves the accuracy and spatial distribution rationality of charging pile allocation, achieving synchronous matching between energy allocation and actual travel patterns.

[0078] This invention utilizes a linkage analysis of a high-mismatch area list and real-time traffic and charging demand data, combined with a dynamic sampling control mechanism that alternates between forward and reverse traffic flows and short-term sampling pauses, to achieve real-time optimization and quota rollback of charging pile deployment. This allows deployment decisions to respond promptly to changes in traffic flow and fluctuations in power load. This method transforms charging infrastructure construction from static planning to an adaptive dynamic adjustment mode, improving the utilization efficiency and deployment flexibility of charging resources. It ensures that charging pile layout continuously aligns with actual usage needs, thereby maintaining a dynamic balance between urban energy supply and vehicle charging behavior.

[0079] The foregoing has only described certain exemplary embodiments of the present invention by way of illustration. Undoubtedly, those skilled in the art can modify the described embodiments in various ways without departing from the spirit and scope of the present invention. Therefore, the foregoing drawings and descriptions are illustrative in nature and should not be construed as limiting the scope of protection of the claims of the present invention.

Claims

1. A method for configuring the number of charging piles based on big data analysis, characterized in that, Includes the following steps: Collect traffic flow data, vehicle charging record data, power grid load curve data and meteorological time data, assign a unified number to the timestamps of all data, and generate a time misalignment sensitive list. Based on the time misalignment sensitive list, data from different sources are compared item by item at minute-level time intervals to calculate the offset position of adjacent data in the time series and form a rhythm superposition touch point table. By using a rhythm overlay contact table, statistical analysis is performed on the recurrence patterns of each contact point to identify periodic echo points and locate peak locations of false charging demand, generating a false peak identification index. Based on the false peak identification index, the regional charging pile configuration map is re-analyzed, the configuration quantity deviation of different regions is calculated, and a list of high misconfiguration areas is compiled. Based on the list of high mismatch areas, combined with real-time traffic data and charging demand changes, dynamic adjustment operations are performed. The data sampling rhythm is corrected by alternating between forward and reverse rhythms and short-term suspension of data collection windows. Quota rollback is implemented according to demand fluctuations, and the number of charging piles is adjusted in real time.

2. The method for configuring the number of charging piles based on big data analysis according to claim 1, characterized in that, The steps for generating a time-misalignment sensitive list are as follows: Traffic flow data, vehicle charging record data, power grid load curve data, and meteorological time data are continuously collected at fixed time sampling intervals to form a multi-source raw time series data set containing timestamps. Using a unified time base point as a reference, all timestamps are numbered, various types of data are mapped to the same time axis system, and missing points are filled in by time interpolation to form a unified time series dataset. Perform time offset correlation analysis on the data after unified numbering to identify the time drift intervals of data from different sources and record the start and end numbers of the offset and the offset direction; Summarize the offset information, generate a time misalignment sensitive list in chronological order, and record the offset source, offset magnitude, and frequency.

3. The method for configuring the number of charging piles based on big data analysis according to claim 2, characterized in that, The steps for forming the rhythm superposition contact table are as follows: Based on the unified time numbering of the time misalignment sensitive list, traffic flow data, vehicle charging record data, power grid load curve data and meteorological time data are aligned minute by minute, and the data from each source under each time number are combined into the same time unit. The difference between data groups with adjacent time numbers is calculated according to the time number order, the offset position of data from different sources in the time series is identified, and the offset start number, offset end number and offset duration are recorded. The identified time offset intervals are analyzed for patterns, and the offset direction, duration and frequency of occurrence of data from each source are statistically analyzed to form a time offset pattern table. Based on the time offset pattern table, integrate time number intervals with synchronous or periodic characteristics, record the offset direction, offset magnitude and offset frequency of each source data, and generate a rhythm superposition touch point table.

4. The method for configuring the number of charging piles based on big data analysis according to claim 3, characterized in that, The steps for generating the pseudo-peak identification index are as follows: Based on the time number, offset direction, offset duration, offset amplitude, and offset frequency recorded in the rhythm superposition contact table, statistical analysis is performed on the repetition distribution of various types of contacts to form an offset data sequence arranged by time number; Perform time interval analysis on the offset data sequence, extract periodic echo features, record the offset source combination, repetition interval, duration and frequency of occurrence, and form a list of periodic echo analysis. Based on the periodic echo analysis list and the time misalignment sensitivity list, identify false charging demand peaks formed by time drift superposition, establish a false peak location list and record the time number range, offset direction consistency and echo period length; Based on the list of false peak locations, all false peak information is integrated, and a false peak identification index is generated in chronological order.

5. The method for configuring the number of charging piles based on big data analysis according to claim 4, characterized in that, During the generation of the false peak identification index, the offset source categories recorded in the false peak location list include traffic flow data, vehicle charging record data, power grid load curve data, and meteorological time data. The consistency of offset direction and echo period length are marked by corresponding time numbers.

6. The method for configuring the number of charging piles based on big data analysis according to claim 4, characterized in that, The steps for generating the list of high mismatched regions are as follows: Based on the time number, offset source type, offset direction, echo period, offset duration and geographical location identifier in the false peak identification index, it is mapped with the spatial nodes of the regional charging pile configuration map to form a list of false peak affected areas, and records the area number, false peak time and configuration change trend corresponding to each false peak. Based on the list of areas affected by false peaks, the configuration quantity of each area during the false peak period and the baseline period is compared and calculated. The configuration quantity deviation, load change magnitude and utilization rate change value are recorded to form a regional configuration quantity deviation table. Based on the principles of geographical proximity and consistency of pseudo-peak cycles, spatial aggregation analysis is performed on the deviation results to generate a regional deviation aggregation table, which records the deviation type, duration, and pseudo-peak influence cycle. Based on the regional deviation aggregation table, each aggregation unit is summarized and organized, and the configuration error type, error magnitude, error duration period and pseudo-peak interference source are recorded to generate a list of high misconfiguration regions.

7. The method for configuring the number of charging piles based on big data analysis according to claim 6, characterized in that, In the regional deviation aggregation analysis, the correlation between geographically adjacent regions is determined based on the power grid load transmission relationship. When the number of charging piles in a region increases during a pseudo-peak period while the load in adjacent regions decreases, it is recorded as a spatial resource migration phenomenon and marked as an unevenly distributed resource region in the list of high mismatched regions.

8. The method for configuring the number of charging piles based on big data analysis according to claim 6, characterized in that, The dynamic adjustment operation execution process is as follows: Based on the list of high mismatched areas, real-time status identification is performed on areas identified as over-configured or under-configured. Real-time traffic data and real-time charging demand data are matched with area numbers to form a real-time status matching matrix that records traffic operation status and changes in energy demand. Based on the real-time state matching matrix, the data sampling process in the high mismatch area is corrected by alternating forward and reverse rhythm sampling. By controlling the sampling method of increasing and decreasing time number, the data distribution is kept balanced, and the rise and fall patterns of demand changes are recorded. During the alternating sampling process of forward and reverse rhythms, a short-term pause window is set to adjust the sampling rhythm by pausing, preventing time offset caused by high-frequency sampling and maintaining the continuity and stability of the data time axis; Based on the mismatch direction and deviation range in the list of high mismatch areas, and combined with real-time traffic and charging demand changes, quota rollback is performed to adjust the number of charging piles in each area and record the adjustment results.

Images