A charging pile background management system and method

Abstract

The present invention relates to the field of new energy supply management technology, specifically a charging pile background management system and method. The method includes: obtaining photovoltaic output, energy storage status and charging load data, completing time alignment and normalization processing, and constructing a unified data input structure; configuring photovoltaic priority strategy and energy storage scheduling parameters based on the structure, constructing and solving a multi-energy power supply optimization model, and generating power supply scheduling output results; combining electricity price sequences to identify peak and valley periods, generate arbitrage labels, and then generate energy storage charging and discharging control strategies and energy storage control instruction structures; issuing control instructions and collecting execution data to generate energy consumption deviation feedback structures; and finally realizing adaptive closed-loop correction of the scheduling model based on the feedback structure. The present invention can improve photovoltaic utilization, realize peak-valley arbitrage and energy storage coordinated control, significantly reduce the power supply cost of charging stations, and enhance system scheduling accuracy and operational stability.

Description

Technical Field

[0001] The present invention relates to the technical field of new energy supply management, and in particular to a charging pile background management system and method. Background Art

[0002] As electric vehicle penetration continues to rise, the density of charging piles, essential energy supply infrastructure, continues to increase. The power supply systems they support are also trending towards a multi-energy convergence, including a hybrid power supply network comprised of centralized grid power, distributed photovoltaic power generation, and local energy storage units. However, current mainstream charging pile management systems generally rely on a single grid power supply model, failing to fully utilize local photovoltaic resources or integrate the adjustable characteristics of energy storage systems for intelligent scheduling. This results in high peak electricity prices and increased operating costs, limiting the large-scale green deployment and economic feasibility of charging infrastructure.

[0003] In practical applications, due to the intermittent and volatile nature of photovoltaic power generation, the charging and discharging strategies of energy storage systems are also highly dependent on the peak and valley fluctuations of grid electricity prices. Therefore, how to perform unified time mapping and optimized scheduling modeling based on photovoltaic output data, energy storage status data, charging load curves, and real-time electricity price series at different time scales has become a key issue in improving the economic and green efficiency of charging stations.

[0004] Traditional strategies often rely on static rule-based control, which is unable to dynamically respond to actual load changes and fluctuations in renewable energy generation. They also suffer from coarse scheduling granularity and low energy efficiency. Especially in scenarios where multiple energy sources are connected simultaneously, achieving optimal systemic operation of multi-energy synergy will be difficult if this multi-source data cannot be structurally integrated and a unified model with mathematical constraints and economic optimization objectives constructed.

[0005] Therefore, there is an urgent need for a charging pile backend management method that combines a mixed integer programming model, a time-series interpolation alignment mechanism, and energy consumption deviation feedback capabilities to dynamically implement a collaborative energy supply strategy with photovoltaic priority, energy storage assistance, and electricity price guidance, so as to minimize the comprehensive electricity cost and improve the utilization rate of green electricity in complex scenarios. Summary of the Invention

[0006] The present invention provides a charging pile backend management system and method to solve the problem of how to build a mixed integer programming model based on photovoltaic output data, energy storage status information, charging load data and real-time electricity price series after unified time mapping and structural integration, generate photovoltaic-priority energy storage charging and discharging coordinated scheduling in real time, and perform closed-loop adaptive correction of energy consumption deviation to minimize the comprehensive power supply cost of charging stations in multi-energy scenarios.

[0007] In order to solve the above technical problems, the present invention provides an analysis method for livestock claims data based on artificial intelligence, comprising:

[0008] Acquire photovoltaic output data, energy storage status information, and charging load data, complete time alignment, normalization, and structural integration, and generate a unified data input structure;

[0009] Obtain a unified data input structure, configure photovoltaic priority strategies and energy storage scheduling constraint parameters, build and solve a multi-energy power supply optimization model, and generate power supply scheduling output results;

[0010] The expression of the power supply scheduling output result is:

[0011]

[0012] in, Output structure for power supply scheduling; is the optimal charge and discharge power; is the optimal power purchase power; is the energy storage state trajectory at each moment; is the unit energy consumption scheduling cost; K is the time index variable; N is the total number of time steps in the scheduling period;

[0013] Obtain power supply scheduling output and electricity price series, identify peak and valley periods, generate arbitrage tags, and build an arbitrage reference structure;

[0014] Obtain the arbitrage reference structure and power supply dispatch output results, combine them with the load data in the unified data input structure, generate the energy storage charging and discharging control strategy and build the energy storage control instruction structure;

[0015] Issue energy storage control instruction structure and power supply scheduling output results, collect execution data of energy control module, and generate energy consumption deviation feedback structure;

[0016] Obtain energy consumption deviation feedback structure, analyze and adjust scheduling parameters in the multi-energy power supply optimization model, and complete model adaptive closed-loop correction;

[0017] The analysis and adjustment of the scheduling parameters in the multi-energy power supply optimization model include:

[0018] The peak deviation and average deviation are mapped to the maximum power constraint of energy storage charging and discharging; the cumulative deviation is mapped to the energy storage cycle number constraint; the number of switching failures and switching delays are mapped to the mode switching delay margin; the maximum state of charge offset is mapped to the energy storage efficiency coefficient; the average state of charge offset and state recovery time are mapped to the state of charge continuity constraint;

[0019] The model adaptive closed-loop correction includes:

[0020] The system performs fast consistency checks, including supply and demand balance constraint checks, energy storage power constraint checks, and charge state continuity checks.

[0021] Furthermore, the acquisition of photovoltaic output data, energy storage status information and charging load data, and the completion of time alignment, normalization and structural integration, include:

[0022] Time-align PV output data, energy storage status information, and charging load data;

[0023] Normalize and perform sliding average processing on photovoltaic output data, energy storage status information, and charging load data;

[0024] The normalization and sliding average processing results are structurally integrated to generate a unified data input structure.

[0025] Furthermore, the configuration of the photovoltaic priority strategy and energy storage scheduling constraint parameters includes:

[0026] Obtain a unified data input structure and configure photovoltaic priority strategy and energy storage scheduling constraint parameters.

[0027] Furthermore, the construction and solution of the multi-energy power supply optimization model includes:

[0028] Construct a multi-energy power supply optimization model, and the objective function includes the power purchase cost of the power grid and the energy loss of energy storage scheduling.

[0029] Furthermore, the obtaining of the power supply scheduling output result and the electricity price sequence, identifying the peak period and the valley period, and generating the arbitrage label includes:

[0030] Obtain power supply scheduling output results and electricity price series, and mark the peak and valley distribution of electricity prices;

[0031] Analyze and process the peak and valley distribution of electricity prices to identify peak and valley periods;

[0032] Generate arbitrage labels based on peak and valley periods to build an arbitrage reference structure.

[0033] Furthermore, the generation of the energy storage charge and discharge control strategy and the construction of the energy storage control instruction structure include:

[0034] Determine the energy storage charging and discharging periods based on the arbitrage reference structure and power supply scheduling output results;

[0035] Generate energy storage charging and discharging control strategies based on load data in a unified data input structure;

[0036] The energy storage charging and discharging control strategy is structured to generate the energy storage control instruction structure.

[0037] Furthermore, the steps of issuing the energy storage control instruction structure and the power supply scheduling output result and collecting the energy control module execution data include:

[0038] Synchronously send the energy storage control instruction structure and power supply dispatch output results to the energy control module;

[0039] Collect actual execution data from the energy control module and compare and analyze it with the issued instructions.

[0040] Furthermore, the step of generating the energy consumption deviation feedback structure includes:

[0041] The comparative analysis results are structured to generate an energy consumption deviation feedback structure.

[0042] A charging pile backend management system, applied to any of the above methods, comprising:

[0043] Data acquisition module, used to obtain real-time photovoltaic output data, energy storage status information, charging load data and electricity price series from the photovoltaic array controller, energy storage management system, smart charging pile controller and billing platform;

[0044] The data preprocessing module is used to perform time alignment, normalization, sliding average and structure integration processing on the raw data and output a unified data input structure;

[0045] An optimization modeling module, which extracts available capacity and power ranges based on a unified data input structure, configures PV priority strategy weights, energy storage scheduling constraints, and state-of-charge continuity constraints;

[0046] Arbitrage analysis module, which receives the power supply scheduling output and electricity price series, marks the peak and valley distribution of electricity prices, identifies peak and valley periods, and calculates derived indicators;

[0047] The strategy generation module is used to obtain the arbitrage reference structure and power supply scheduling output results, and determine the energy storage charging and discharging switching points based on the load power requirements in the unified data input structure;

[0048] The instruction execution module is used to synchronously send the energy storage control instruction structure and the power supply scheduling output results to the energy control module;

[0049] The model update module is used to analyze the deviation of energy storage power constraints, mode switching margins, state of charge continuity constraints, and photovoltaic priority weight model parameters based on the energy consumption deviation feedback structure generated by the instruction execution module;

[0050] The system monitoring and visualization module is used to monitor the operating status of each functional module in real time, displaying the photovoltaic-energy storage-grid power flow, peak and valley identification results, strategy execution trajectory and model parameter evolution curve.

[0051] The key innovations of the present invention include:

[0052] (1) Construct a unified mapping and normalized structural integration mechanism for multi-source heterogeneous data to achieve the integrated input of photovoltaic output, energy storage status, electricity price and load information, and effectively solve the problems of asynchrony and incompatibility of multi-dimensional data.

[0053] (2) Construct a hybrid integer programming scheduling framework with photovoltaic priority and energy storage participation, support dynamic configuration of scheduling constraints, peak and valley identification and arbitrage label generation, and adapt to the multi-variable and strong coupling characteristics of multi-energy regulation.

[0054] (3) A scheduling closed-loop mechanism with execution data feedback and model parameter self-correction capabilities is proposed, so that the scheduling model is no longer static and rigid, but has dynamic adaptability and continuous optimization potential.

[0055] The following are its main beneficial effects:

[0056] (1) Significantly reduce the comprehensive power supply cost of charging stations. This invention obtains photovoltaic output data, energy storage status information, charging load data and real-time electricity price series, and for the first time proposes an input mechanism based on unified time mapping and normalized structure integration, breaking through the interaction bottleneck of multi-source asynchronous data. By constructing a mixed integer programming model with scheduling constraints, the system can dynamically generate photovoltaic-priority collaborative energy supply scheduling results, effectively improving the utilization rate of green electricity, reducing dependence on high-priced grid electricity, achieving peak-valley arbitrage and significantly reducing power supply costs.

[0057] (2) Improve the real-time performance and stability of the multi-energy scheduling model. Traditional charging pile systems mostly adopt static strategies or coarse-grained load forecasting methods, lacking a dynamic scheduling mechanism that adapts to the volatility of renewable energy. This invention introduces the power supply scheduling output structure and electricity price identification mechanism, constructs an arbitrage label and energy storage charging and discharging strategy generation model, and realizes the flexible scheduling of photovoltaic, energy storage and city power by charging stations while ensuring the continuity of energy supply. This method fully utilizes the buffering role of the energy storage system in load shifting and peak shaving, so that the scheduling system has stronger dynamic response capabilities and strategy robustness.

[0058] (3) Implementing a closed-loop adaptive correction model to improve scheduling accuracy and economy. In actual operation, due to the volatility of renewable energy power generation and the uncertainty of load demand, it is difficult for the optimization model to maintain the optimal operating state for a long time. To this end, the present invention proposes an energy consumption deviation feedback mechanism. By collecting the execution data of the energy control module, an energy consumption deviation feedback structure is constructed, and the parameters in the scheduling model are periodically adaptively corrected to achieve continuous dynamic optimization of the scheduling strategy. This closed-loop control design not only improves the stability and self-learning ability of the overall model, but also significantly improves the accuracy and profitability of the scheduling output. BRIEF DESCRIPTION OF THE DRAWINGS

[0059] Figure 1A flowchart of a charging pile background management method provided in an embodiment of the present application;

[0060] Figure 2 This is a structural block diagram of a charging pile background management system provided in an embodiment of the present application. DETAILED DESCRIPTION

[0061] Example 1: Reference Figure 1 , is a flow chart of a charging pile background management method provided by an embodiment of the present invention, the process may include at least steps S100-S600:

[0062] S100: Acquire photovoltaic output data, energy storage status information, and charging load data, complete time alignment, normalization, and structure integration, and generate a unified data input structure.

[0063] S200: Obtain a unified data input structure, configure photovoltaic priority strategy and energy storage scheduling constraint parameters, build and solve a multi-energy power supply optimization model, and generate power supply scheduling output results.

[0064] S300: Obtain power supply scheduling output results and electricity price sequences, identify peak periods and valley periods, generate arbitrage tags, and construct an arbitrage reference structure.

[0065] S400: Obtain the arbitrage reference structure and the power supply scheduling output result, combine the load data in the unified data input structure, generate the energy storage charging and discharging control strategy and construct the energy storage control instruction structure.

[0066] S500: Sending energy storage control instruction structure and power supply scheduling output results, collecting execution data of energy control module, and generating energy consumption deviation feedback structure.

[0067] S600: By obtaining the energy consumption deviation feedback structure, the scheduling parameters in the multi-energy power supply optimization model are analyzed and adjusted to complete the model adaptive closed-loop correction.

[0068] Step S100 at least includes steps S110-S130:

[0069] S110: Acquire photovoltaic output data, energy storage status information, and charging load data, and perform time alignment processing.

[0070] First, the system accesses the charging station's photovoltaic power generation system, energy storage management system, and intelligent charging pile controller through a data acquisition module to obtain photovoltaic output data, energy storage status information, and charging load data. The photovoltaic output data is the change in photovoltaic power generation within a unit time interval. The energy storage status information includes the current state of charge of the energy storage battery, the current charge and discharge power, and the cumulative charge and discharge time. The charging load data is the power demand and instantaneous load curve of all terminal devices in the charging station within a unit time window.

[0071] Furthermore, to ensure consistency across multiple data sources during optimization, time alignment is required. Specifically, based on a unified timestamp generation mechanism, the system parses the original time tags of each data source and performs interpolation, cropping, or delay correction on data records with sampling frequency discrepancies or delays. Photovoltaic output data uses a minute-level sampling cycle, energy storage status information uses second-level status updates, and charging load data has a millisecond-level sampling granularity to respond to demand fluctuations, requiring aggregation according to a unified cycle.

[0072] Furthermore, the time alignment process is accomplished through a sliding window mechanism and queue buffer management, ensuring a synchronized relationship between PV output, energy storage status, and load data within each time slice. Once completed, three sets of time-aligned data sequences with consistent structure are generated.

[0073] This time alignment result will directly serve as input data for the subsequent normalization and sliding average processing in S120, forming the "original aligned data set." Each time node in the original aligned data set contains a set of photovoltaic power, energy storage status values, and load values, laying the data time series foundation for building a unified data input structure.

[0074] S120 , normalizing and performing sliding average processing on the photovoltaic output data, energy storage status information, and charging load data.

[0075] After completing the time alignment operation in S110, this step performs normalization and sliding average processing on the three types of data to improve data consistency, smooth volatility, and construct a standard input space for the optimization modeling stage.

[0076] Specifically, the system first reads the original aligned data set and constructs normalization processes for PV output data, energy storage status information, and charging load data. This normalization process uses a component range compression method to uniformly map each indicator value to an interval range, eliminating weight bias caused by differences in device rated power, time scale, or status units.

[0077] Furthermore, PV output data is normalized based on the PV array's rated capacity, energy storage status data is normalized based on its maximum storage capacity and maximum power, and charging load data is normalized using a table of factors based on the system's rated load limit. This normalization process ensures that all three data channels have equal input capabilities for numerical features.

[0078] After normalization, the system performs a sliding average based on a sliding window strategy with equal lengths. This process uses a fixed window width and step size to perform a weighted or simple average of multiple data sets before and after each time point, thereby reducing high-frequency noise caused by environmental disturbances, load jitter, or sensor errors.

[0079] For example, short-term fluctuations in photovoltaic output data are smoothed through averaging, spikes in energy storage capacity caused by switching between charging and discharging states are adjusted, and abnormal pulse interference in load data is reduced. The sliding average window length is independently set based on the fluctuation characteristics of each data channel, with a longer window used for the photovoltaic channel and a shorter window used for the load channel to maintain response sensitivity.

[0080] The outputs of normalization and sliding average processing form a "standardized photovoltaic data set," a "standardized energy storage data set," and a "standardized load data set," respectively. These three standardized data structures have unified timestamps and numerical dimension formats, forming a standardized data processing result.

[0081] The above-mentioned standardized data structure will be structurally integrated into a unified data input structure in S130, serving as the basic input variable set for the subsequent optimization modeling stage (S200 module), and will be directly called when used in S220 to build a multi-energy power supply optimization model.

[0082] S130: Structurally integrate the normalization and sliding average processing results to generate a unified data input structure.

[0083] Specifically, after completing the preprocessing of various energy-related data, this step structurally integrates the standardized photovoltaic data set, the standardized energy storage data set, and the standardized load data set to ultimately form a unified data input structure.

[0084] Specifically, the structural integration operation constructs a unified data row and column representation template, combining the normalized PV output value, energy storage state of charge, charge and discharge status labels, and load power demand values ​​within each time slice into structured records. These records are then sorted chronologically to form a sequential matrix structure. This structure supports sliding searches by time period and field calls by energy type, adapting to various optimization constraint forms.

[0085] Furthermore, each record item in the unified data input structure contains a data identifier field, a data value field, and a data source type field, which are used to describe the data source, numerical content, and data dimension classification, respectively, providing structural support for subsequent multi-energy scheduling strategy configuration.

[0086] Furthermore, after integration is complete, the unified data input structure is output as a single structure, directly passed to S210 for configuring the PV priority strategy and energy storage scheduling constraint parameters, and then used as the model construction input vector in S220. Based on this structure, the optimization module can identify current energy availability, load trends, and energy storage capacity, forming the temporal foundation for scheduling scenario modeling.

[0087] Through the detailed implementation of this step, the system achieves time-series synchronization, numerical unification, and structural integration of heterogeneous energy data, providing standardized input for subsequent multi-energy optimization and scheduling models. This process significantly improves model solution efficiency, scheduling strategy accuracy, and system execution stability, providing a high-quality data foundation for the multi-energy collaborative optimization goals of this invention.

[0088] Step S200 at least includes steps S210-S230:

[0089] S210: Obtain a unified data input structure and configure photovoltaic priority strategy and energy storage scheduling constraint parameters.

[0090] Specifically, this step first obtains the data input structure that has completed alignment, normalization, and structural integration, including:

[0091] Photovoltaic output time series: ;

[0092] Energy storage state time series: ;

[0093] Charging load demand sequence: ;

[0094] Real-time electricity price time series: ;

[0095] Unit energy consumption scheduling cost series: .

[0096] Since the above data have different time sampling sources, in order to achieve unified modeling and scheduling processing, they need to be mapped to a unified scheduling time. The process steps are as follows:

[0097]

[0098] in, Indicates interpolation or sliding window alignment operations to ensure that all indicators are aligned at the same scheduling time , meeting the model consistency input requirements.

[0099] Furthermore, the system sets the following parameters according to the configuration policy:

[0100] PV priority weight parameters:

[0101] Maximum power of energy storage charging and discharging: ;

[0102] Energy storage capacity upper and lower limits: .

[0103] S220. Construct a multi-energy power supply optimization model based on photovoltaic priority strategy and energy storage scheduling constraint parameters.

[0104] Specifically, based on the acquired and configured parameters, the following optimization objective function is constructed to minimize the combined dispatch loss of power supply cost and energy storage depreciation:

[0105] Formula ①: Multi-energy power supply optimization objective function:

[0106]

[0107] in:

[0108] :Energy storage at all times Discharge power (negative value indicates charging);

[0109] :express , used to indicate the positive compensation for the load notch;

[0110] : unit electricity price;

[0111] : Energy storage loss weight, controlling its dispatch frequency;

[0112] : The total number of time steps in the optimization cycle.

[0113] The objective function consists of two parts: the first is the power purchase cost of the power grid, and the second is the energy loss of energy storage scheduling.

[0114] Furthermore, in order to ensure the stable operation of the power supply system, the following constraints need to be added:

[0115] Formula ②: Load balancing constraint:

[0116]

[0117] in, For the moment Electricity purchased from the grid.

[0118] Formula ③: Energy storage power constraint and energy state evolution formula:

[0119]

[0120]

[0121] in, : Charge and discharge efficiency; ; : time interval; ;

[0122] S230: Solve the multi-energy power supply optimization model and generate a power supply scheduling output result.

[0123] After constructing the objective function and constraints, the system calls a mixed integer linear programming (MILP) or gradient-based optimizer to iteratively solve the following optimization problem:

[0124] The system , , Given a known input, the optimal variable sequence is solved and the system encapsulates the optimization result into a power supply scheduling output structure:

[0125]

[0126] in, Output structure for power supply scheduling; is the optimal charge and discharge power; is the optimal power purchase power; is the energy storage state trajectory at each moment; is the unit energy consumption scheduling cost; K is a time index variable, representing the kth moment in the scheduling period, with a value range of k=1,2,…,N; N is the total number of time steps in the scheduling period, a positive integer, the unit is step, which is set by the system or dynamically determined.

[0127] The structure serves as the core input for subsequent modules S300 (arbitrage strategy generation) and S400 (energy storage control strategy), supporting the scheduling logic closed loop of the entire system.

[0128] Through the design and implementation of this S200 module, the present invention can effectively solve the real-time scheduling optimization problem of the charging pile system in a multi-energy access scenario, and realize an efficient power supply strategy with photovoltaic priority, energy storage assistance, and grid backup. The proposed unified time mapping mechanism, cost-loss dual-objective optimization function, dynamic energy balance constraint construction and feasible domain scheduling strategy joint modeling significantly improve the system's responsiveness to load changes and its adaptability to peak and valley fluctuations in electricity prices. Under the premise of ensuring the continuity of charging services, the dual goals of minimizing comprehensive energy consumption costs and extending the life of energy storage are achieved, which has good engineering promotion value and practical application prospects.

[0129] Step S300 at least includes steps S310-S330:

[0130] S310: Obtain the power supply scheduling output results and electricity price sequence, and mark the peak and valley distribution of electricity prices.

[0131] First, the system calls the power scheduling output generated in step S230 in the arbitrage analysis module and simultaneously calls the billing interface to obtain the electricity price sequence. Both data sets contain a time stamp field. The system performs integrity checks, duplication checks, and sequence consistency checks on the time stamp field. If missing records or time misalignments are found, log recovery or elimination is performed to ensure the reliability of the data link entering the labeling process.

[0132] Furthermore, the system aligns the two types of data based on a unified dispatch time. If the price sequence resolution is higher than the power dispatch time granularity, the system compresses the price data using a sliding window average. If the resolution is lower than the power dispatch granularity, the price entries are extended through time interpolation. Once this processing is complete, each power dispatch record has a matching price record.

[0133] Furthermore, the system calculates the statistical distribution of electricity prices for the same region, season, and week based on a database of similar historical electricity prices. It then uses the percentile interval method to determine the high and low price thresholds. Continuous segments above the high price threshold and lasting at least as long as the minimum high price window are labeled as peaks, while continuous segments below the low price threshold and lasting at least as long as the minimum low price window are labeled as valleys. The remaining segments are labeled as flat. The labeled results are written back to the electricity price sequence as newly added fields, forming a set of labeled peak and valley distributions for electricity prices.

[0134] The electricity price peak and valley distribution annotation set is written into the buffer database and generates a unique task identifier for subsequent sub-steps to call.

[0135] S320: Analyze and process the peak-valley distribution of electricity prices to identify peak and valley periods.

[0136] Specifically, once the peak-valley distribution annotation set for electricity prices is placed, the system immediately enters the continuous segment aggregation logic. The system scans the peak-valley fields of adjacent entries in the annotation set and merges records with the same identifier and continuous time to form a high-price continuous segment set and a low-price continuous segment set.

[0137] Furthermore, the system performs continuous filtering on the set of consecutive segments. Any high-price or low-price segment that lasts less than the minimum response window is merged with the adjacent segment if the time interval between the adjacent segment is less than the minimum gap threshold; if the time interval is greater than the threshold, it is re-labeled as a flat segment.

[0138] After obtaining a set of valid peak and valley time periods, the system further calculates nine derived metrics for each continuous segment: maximum unit price, minimum unit price, average unit price, price spread, fluctuation slope, historical percentile position, duration, forward gap, and backward gap. These nine metrics are then appended to the continuous segment structure. The processed peak and valley time period sets are written to the analysis buffer, and a statistical summary, including the number of segments, total duration, and maximum price spread, is recorded in the system log, providing traceability clues for subsequent model revisions.

[0139] S330: Generate arbitrage labels according to peak time periods and valley time periods, and construct an arbitrage reference structure.

[0140] Specifically, once the peak and valley time periods are cached, the system begins energy storage capacity matching. The system first reads the energy storage status trajectory and energy storage power constraint fields from the power scheduling output in step S230, calculates the charge capacity that can be accommodated in each valley period, and eliminates valley periods with insufficient capacity.

[0141] After determining the available valley periods, the system searches for the nearest peak period for each valley period in chronological order. If multiple peak periods meet the requirements, they are matched sequentially, based on the price difference from highest to lowest. A successful match generates a charge-discharge pair. The system creates an arbitrage tag for each pair, writing nine fields: charge start time, charge end time, expected charge power, discharge start time, discharge end time, expected discharge power, achievable price difference profit, execution priority, and a unique tag number. If a valley period forms multiple pairs with multiple peak periods, the system sorts them according to the priority field.

[0142] After all arbitrage tags are encapsulated as record items, the system generates an arbitrage reference structure set. The arbitrage reference structure set follows the time format and field naming rules of the unified data input structure to ensure that it can be directly parsed in step S410.

[0143] Furthermore, the system writes the arbitrage reference structure set into the shared cache, and simultaneously pushes the reference address and task identifier to the energy storage control strategy generation module to complete the entire processing flow of this module.

[0144] Through the implementation of this S300 module, the system can accurately identify peak and valley periods and generate executable arbitrage tags while ensuring data alignment. This provides a quantifiable and verifiable decision-making basis for the energy storage system to initiate valley charging and peak discharging, significantly improving operating profits and laying a solid foundation for subsequent control strategy generation.

[0145] Step S400 at least includes steps S410-S430:

[0146] S410: Obtain an arbitrage reference structure and power supply scheduling output results, and determine an energy storage charging and discharging switching point.

[0147] Specifically, when the strategy generation module starts, the system reads the arbitrage reference structure set output by step S330 from the shared cache and reads the power supply scheduling output result output by step S230 in parallel. Both types of data structures use the unified scheduling time as the time base, and the internal field names remain consistent.

[0148] Specifically, the system first performs a data integrity check to confirm that the charging start time, charging end time, discharge start time, and discharge end time in the arbitrage reference structure all fall within the time range of the power supply scheduling output result; if it is found that the time is out of bounds or missing, the system reports an error to the upper level according to the log specification and suspends the execution of the tag to avoid sending unexecutable instructions to the energy storage device.

[0149] The system then reads the expected charging power and expected discharging power fields for each entry in the arbitrage reference structure and compares the difference with the energy storage power field in the power supply scheduling output. As you can see, if there's a significant discrepancy between the expected power and the scheduled power, the system invokes the model update module for a second evaluation. If the discrepancy is within a manageable threshold, the system proceeds to the switch point determination process.

[0150] During the switching point determination process, the system subdivides the charging and discharging time periods contained in each arbitrage tag, ensuring that the switching instant is fully aligned with the unified scheduling moment. Specifically, the system inserts the "energy storage charge and discharge switching point" marker between the charging end moment and the discharging start moment, using the unified scheduling moment as the scale, and writes the switching type field at the same moment. The switching type field is used to indicate the switch from "charging mode" to "discharging mode" or from "discharging mode" to "standby mode." The energy storage charge and discharge switching points are appended to the arbitrage reference structure to form an updated extended set of switching points.

[0151] The switch point extension set is written into the strategy buffer as the output content of S410 , and is accompanied by a task identification code and a timestamp for subsequent strategy generation.

[0152] S420: Generate an energy storage charge and discharge control strategy according to the energy storage charge and discharge switching point and the load data in the unified data input structure.

[0153] Specifically, after reading the extended set of switching points, the system immediately calls the unified data input structure generated in step S130 to extract the load power demand field at each moment. Specifically, during the charging period, the system calculates the load remaining capacity, i.e., the power headroom available for energy storage charging without impacting normal charging operations; during the discharging period, the system calculates the load compensation demand, i.e., the power gap that needs to be filled by energy storage discharge.

[0154] Understandably, the system also needs to combine the PV power field and the grid power field at the corresponding moment in the power dispatch output to obtain the real-time renewable energy utilization status and the grid's backup capacity. If the PV output exceeds the load demand during the charging period, the system will adjust the expected charging power and prioritize the use of remaining PV power for charging. If the PV output is insufficient, the system will evaluate whether the available charging power meets the expected value based on the grid power limit, and if appropriate, defer the remaining charge to the next available off-peak period.

[0155] Furthermore, during the discharge period, the system first determines whether the load power gap is sufficient to absorb the desired discharge power. If the gap is insufficient, the excess discharge power is adjusted down to the load gap limit. If the gap is greater than the desired discharge power, the system replenishes the difference as needed and outputs a margin indicator for the model update module to record. This dynamic power adjustment ensures that the energy storage power command is implemented safely and does not violate supply and demand balance constraints.

[0156] After the load coupling calculations described above, the system generates an "Energy Storage Charge and Discharge Control Strategy" record for each arbitrage tag. This record contains fields for the strategy number, charging period, charging power, discharging period, discharging power, switching point, load margin, compensation gap, and execution priority. All strategy records are sorted by execution priority and merged into a set of energy storage charge and discharge control strategies. This set of control strategies is frozen in read-only memory for subsequent instruction generation and recall.

[0157] S430: Convert the energy storage charge and discharge control strategy into an energy storage control instruction structure.

[0158] Specifically, after reading the energy storage charge and discharge control strategy set, the system calls the instruction template library to generate an instruction structure that can be directly parsed by the energy storage management system. The template library presets the instruction message format based on the energy storage device communication protocol, including a message header, timestamp, power command, mode command, priority command, checksum, and message trailer.

[0159] The system first populates the command message header for each control strategy entry, writing the global task identifier and the current generation time. It then maps the charging period, charging power, discharging period, and discharging power into the power command segment, the energy storage charge and discharge switching points into the mode command segment, and the execution priority into the priority command segment. After writing, the system calls the encryption module to generate a checksum and appends it to the end of the message.

[0160] To meet the energy storage management system's batch execution requirements, the system concatenates all command messages in chronological order and writes a field indicating the total number of commands in the message header, forming an "energy storage control command structure." This command structure is cached in binary form in a high-priority channel, awaiting dispatch in step S510. A command summary is also written to the log for audit purposes.

[0161] Through the implementation of this S400 module, the system, while ensuring data consistency, deeply couples arbitrage tags with real-time load status, dynamically generates energy storage charging and discharging control strategies, and then outputs energy storage control instruction structures in the form of protocolized messages. This process significantly improves strategy accuracy and execution reliability, laying a solid data and control foundation for subsequent energy consumption deviation feedback and model adaptive correction.

[0162] Step 500 at least includes steps S510-S530:

[0163] S510: The energy storage control instruction structure and the power supply scheduling output result are synchronously sent to the energy control module.

[0164] In the instruction execution module, the system first loads the energy storage control instruction structure generated in S430 and simultaneously loads the power supply scheduling output generated in S230. Both structures have a unified scheduling time field and a task identification field.

[0165] Specifically, the system requests a high-priority channel for this dispatch operation at the top of the communication scheduling queue and writes a synchronization flag into the channel header, enabling the energy control module to implement frame-level alignment on the receiving end. The system then writes the energy storage control instruction structure one by one into the transmission buffer. Each instruction is written in the following order: message header, timestamp, power instruction segment, mode instruction segment, priority instruction segment, checksum segment, and message trailer. Once written, the system immediately appends the power, state of charge, and cost fields of the power scheduling output to form a command-scheduling composite message frame.

[0166] Furthermore, to ensure time consistency across field devices, the system adds a clock synchronization segment to the end of the composite message frame and writes a second-level timestamp corresponding to the unified scheduling time. After completing frame verification at the receiving end, the energy control module returns an acknowledgment signal. Upon receiving the acknowledgment signal, the system transfers the message frame to the link log and marks it as "sent." If the acknowledgment signal does not return within the specified timeout, the system immediately triggers a retransmission process, with up to three retries. If the acknowledgment signal still fails, the message is entered into the exception queue and an alarm is generated.

[0167] S520: Collect actual execution data from the energy control module and compare and analyze it with the issued instructions.

[0168] Specifically, when the energy control module enters the execution phase, the system activates the telemetry channel and cyclically pulls actual execution data, using the unified scheduling time as the index period. The actual execution data includes the charging power measurement field, the discharging power measurement field, the state of charge measurement field, the mode switching event field, and the device alarm field.

[0169] Understandably, the system first verifies the validity of telemetry data, eliminating records with missing timestamps, failed verification, or with severe alarm status. The system then uses a time alignment mechanism to align the telemetry data with the timestamps in the issued instructions. Once aligned, the system uses the instruction number as a key to compare the telemetry power measurement field with the power instruction segment, the telemetry mode switch event field with the mode instruction segment, and the telemetry state-of-charge measurement field with the predicted state-of-charge field, one by one, creating a temporary cache of differences.

[0170] In the temporary difference cache, the system records three types of difference values ​​for each instruction: power difference, mode difference, and state difference. Power difference corresponds to charge and discharge power offset; mode difference corresponds to switching type error; and state difference corresponds to state of charge prediction error. For records with warning fields above the general level, the system adds an exception flag and highlights it when generating the subsequent deviation structure.

[0171] S530: Structurally process the comparative analysis results to generate an energy consumption deviation feedback structure.

[0172] After the temporary buffer of differences is frozen, the system enters the energy consumption deviation calculation process. First, the system calculates the power deviation by sliding according to the unified scheduling time to generate a charge and discharge deviation curve. Then, the system aggregates the mode and state differences by task identifier to generate a mode consistency table and a state error table.

[0173] Specifically, the system extracts the peak deviation, average deviation, and cumulative deviation fields from the charge-discharge deviation curve; counts the number of successful switching, failed switching, and delayed switching in the mode consistency table; and counts the maximum state-of-charge deviation, average state-of-charge deviation, and state recovery duration in the state error table. These three statistical results and alarm flags are packaged into the energy consumption deviation feedback structure.

[0174] The energy consumption deviation feedback structure consists of a header, timestamp, task identifier, power deviation segment, mode deviation segment, state deviation segment, alarm segment, and a footer. The system writes the energy consumption deviation feedback structure to the shared cache and broadcasts a deviation generation event on the system bus. After monitoring the deviation generation event, the model update module calls S610 to perform deviation analysis on the multi-energy power supply optimization model, forming an adaptive closed loop.

[0175] Through the implementation of this S500 module, the system realizes a real-time closed loop of command issuance, execution monitoring and deviation feedback. It can accurately capture the power deviation, mode deviation and state deviation between the energy storage execution layer and the scheduling model, and provide high-precision data support for subsequent model adaptive correction, thereby continuously improving the reliability and economy of the multi-energy collaborative optimization model.

[0176] Step S600 at least includes steps S610-S630:

[0177] S610: Obtain an energy consumption deviation feedback structure and perform deviation analysis on constraint parameters in a multi-energy power supply optimization model.

[0178] When the model update module starts, the system captures the deviation generation event broadcast by S530 through the bus event listener and reads the energy consumption deviation feedback structure. The energy consumption deviation feedback structure includes a structure header, a timestamp, a task identifier, a power deviation segment, a mode deviation segment, a state deviation segment, an alarm segment, and a structure footer. The system first performs an integrity check on the timestamp field to ensure that the time window corresponding to the deviation data is consistent with the current optimization cycle. If the time window is inconsistent, the record is marked as expired and archived, and is not included in the current analysis round.

[0179] Specifically, the system analyzes the power deviation segment to obtain three indicators: peak deviation, average deviation and cumulative deviation; analyzes the mode deviation segment to obtain three indicators: number of successful switching, number of failed switching and number of switching delays; analyzes the state deviation segment to obtain three indicators: maximum state of charge deviation, average state of charge deviation and state recovery time; analyzes the alarm segment to obtain two types of information: general alarm items and serious alarm items.

[0180] The system then uses a deviation mapping table to map nine deviation indicators to model constraint parameter dimensions. The mapping rules are as follows: peak deviation and average deviation are mapped to the maximum energy storage charge and discharge power constraint; cumulative deviation is mapped to the energy storage cycle count constraint; switching failures and switching delays are mapped to the mode switching delay margin; maximum state of charge deviation is mapped to the energy storage efficiency coefficient; average state of charge deviation and state recovery time are mapped to the state of charge continuity constraint; and severe alarm entries trigger the model safety threshold check process.

[0181] Based on the mapping results, the system generates a "Constraint Parameter Deviation List." This list records each constraint parameter's current value, deviation source, deviation direction, deviation magnitude, and recommended adjustment range. Constraint parameters with severe warnings are marked "Mandatory Adjustment"; those with only minor deviations are marked "Optional Adjustment." The Constraint Parameter Deviation List is written to the parameter buffer as the sole output of S610, and an analysis complete event is generated.

[0182] S620: Adjust the scheduling parameters of the multi-energy power supply optimization model according to the deviation analysis results.

[0183] After capturing the analysis completion event, the system reads the constraint parameter deviation list and begins the parameter adjustment process. To ensure the security of the update operation, the system first performs an adjustable range check on each constraint parameter in the deviation list. Checking items include the upper and lower limits of the parameter and the maximum step size for a single adjustment. If the recommended adjustment range exceeds the adjustable range, the system truncates it to the boundary value. If the single adjustment step size exceeds the limit, the system performs recursive step-by-step adjustments based on the maximum step size.

[0184] Specifically, the system executes the following procedures for the maximum power constraint of energy storage charging and discharging: if the peak deviation and the average deviation exceed the set threshold at the same time, the maximum power of energy storage charging and discharging will be increased or decreased by one step; if only the average deviation exceeds the threshold, only the average power limit will be adjusted; if only the peak deviation exceeds the threshold but the average deviation does not exceed the threshold, the peak power clipping factor will be adjusted.

[0185] The system executes the following procedures for the mode switching delay margin: if the number of switching failures or switching delays exceeds the threshold, the mode switching window is expanded and the switching lead time is synchronously updated in the instruction template library; if the switching success rate remains high and the number of delays is small, the switching window is shortened to improve time utilization.

[0186] Furthermore, the system executes the following procedures for the energy storage efficiency coefficient and the state of charge continuity constraint: if the maximum state of charge offset and the average state of charge offset are both positive, the energy storage efficiency coefficient is lowered; if the state recovery time is too long, the state of charge continuity constraint is tightened to reduce the tolerance; if the offset is negative and the state of charge recovery is fast, the constraint is appropriately relaxed to improve flexibility.

[0187] Furthermore, the system encapsulates the above adjustment actions as "scheduling parameter adjustment results," which include the parameter name, original value, adjusted value, adjustment reason, adjustment step, adjustment timestamp, and version number. The scheduling parameter adjustment results are written to the adjustment result buffer and trigger a parameter write event.

[0188] S630: Writing the adjustment processing results into the multi-energy power supply optimization model to complete the model's adaptive closed-loop correction.

[0189] After capturing a parameter write event, the system enters the model write-back process. First, the system obtains the scheduling parameter adjustment results and compares them with the current model parameter version numbers. If the version numbers match, the system enters the write lock phase. If the version numbers don't match, the model has been updated in a parallel process. The system determines the latest version based on the timestamp, remaps the adjustment results to the latest model instance, and then continues writing back.

[0190] Specifically, during the write lock phase, the system adds a write lock to the multi-energy power supply optimization model to prevent other threads from writing at the same time. The system reads the scheduling parameter adjustment results one by one, writes the adjusted values ​​into the corresponding parameter fields within the model, and records the adjustment reasons and adjustment timestamps in the model parameter table. After writing is completed, the system performs a fast consistency check, including supply and demand balance constraint check, energy storage power constraint check, and charge state continuity check. If all checks pass, the model is submitted and the write lock is released; if the check fails, the write is rolled back, and the reason for the failure is written into the exception log and sent back to the model update module for the next round of analysis.

[0191] After successful submission, the system generates a "model parameter version change event" and registers the new model version number in the optimization modeling module. The next round S210 will use the new version parameters when obtaining model parameters, thus forming a complete adaptive closed loop.

[0192] Through the implementation of this S600 module, the system can perform fine-grained adaptive corrections to model parameters based on real-time energy consumption deviations, so that the multi-energy power supply optimization model continues to maintain high consistency with on-site equipment, load characteristics, and electricity price changes; at the same time, the write lock and version control mechanism ensures that the parameter update process is safe and reliable, effectively preventing model oscillation and constraint failure, and providing core technical support for the economic operation and long-term stability of the entire system.

[0193] The key innovations of the present invention include:

[0194] (1) A unified mapping and normalized structural integration mechanism for multi-source heterogeneous data enables the integrated input of photovoltaic output, energy storage status, electricity price and load information, effectively solving the problems of asynchrony and incompatibility of multi-dimensional data.

[0195] (2) Construct a hybrid integer programming scheduling framework with photovoltaic priority and energy storage participation, support dynamic configuration of scheduling constraints, peak and valley identification and arbitrage label generation, and adapt to the multi-variable and strong coupling characteristics of multi-energy regulation.

[0196] (3) A scheduling closed-loop mechanism with execution data feedback and model parameter self-correction capabilities is proposed, so that the scheduling model is no longer static and rigid, but has dynamic adaptability and continuous optimization potential.

[0197] The following are its main beneficial effects:

[0198] (1) Significantly reduce the comprehensive power supply cost of charging stations. This invention obtains photovoltaic output data, energy storage status information, charging load data and real-time electricity price series, and for the first time proposes an input mechanism based on unified time mapping and normalized structure integration, breaking through the interaction bottleneck of multi-source asynchronous data. By constructing a mixed integer programming model with scheduling constraints, the system can dynamically generate photovoltaic-priority collaborative energy supply scheduling results, effectively improving the utilization rate of green electricity, reducing dependence on high-priced grid electricity, achieving peak-valley arbitrage and significantly reducing power supply costs.

[0199] (2) Improve the real-time performance and stability of the multi-energy scheduling model. Traditional charging pile systems mostly adopt static strategies or coarse-grained load forecasting methods, lacking a dynamic scheduling mechanism that adapts to the volatility of renewable energy. This invention introduces the power supply scheduling output structure and electricity price identification mechanism, constructs an arbitrage label and energy storage charging and discharging strategy generation model, and realizes the flexible scheduling of photovoltaic, energy storage and city power by charging stations while ensuring the continuity of energy supply. This method fully utilizes the buffering role of the energy storage system in load shifting and peak shaving, so that the scheduling system has stronger dynamic response capabilities and strategy robustness.

[0200] (3) Implementing a closed-loop adaptive correction model to improve scheduling accuracy and economy. In actual operation, due to the volatility of renewable energy power generation and the uncertainty of load demand, it is difficult for the optimization model to maintain the optimal operating state for a long time. To this end, the present invention proposes an energy consumption deviation feedback mechanism. By collecting the execution data of the energy control module, an energy consumption deviation feedback structure is constructed, and the parameters in the scheduling model are periodically adaptively corrected to achieve continuous dynamic optimization of the scheduling strategy. This closed-loop control design not only improves the stability and self-learning ability of the overall model, but also significantly improves the accuracy and profitability of the scheduling output.

[0201] Example 2: Figure 2FIG. 1 shows a structural block diagram of a charging pile background management system according to an embodiment of the present invention. Figure 2 As shown, the structure may include:

[0202] The data acquisition module 10 is used to obtain real-time photovoltaic output data, energy storage status information, charging load data, and electricity price series from the photovoltaic array controller, energy storage management system, smart charging pile controller, and billing platform; it uses a unified timestamp generation mechanism to achieve millisecond-level alignment and writes the original data cache area through a buffer queue.

[0203] The data preprocessing module 20 is used to perform time alignment, normalization, sliding average and structural integration processing on the original data, and output a unified data input structure; at the same time, it interpolates missing data, denoises noisy data, and marks abnormal data to provide standardized, complete and high-quality data for subsequent optimization modeling.

[0204] The optimization modeling module 30 is used to extract the available capacity and power range based on the unified data input structure, configure the photovoltaic priority strategy weight, energy storage scheduling constraints and charge state continuity constraints; build a multi-energy power supply optimization model, and the objective function comprehensively considers the power purchase cost of the power grid, the cost of energy storage use and the photovoltaic priority weight; and call the mixed integer programming solver to output the power supply scheduling output results.

[0205] The arbitrage analysis module 40 is used to receive the power supply scheduling output and the electricity price sequence, mark the peak and valley distribution of electricity prices, identify peak and valley periods, and calculate derived indicators; under the premise of considering energy storage capacity and power constraints, generate charge-discharge pairs and form arbitrage labels, and output an arbitrage reference structure.

[0206] Strategy generation module 50 is used to obtain the arbitrage reference structure and power supply scheduling output results, and determine the energy storage charge and discharge switching point based on the load power demand in the unified data input structure; calculate the charge and discharge power values ​​and generate the energy storage charge and discharge control strategy; encapsulate the message according to the energy storage device communication protocol and construct the energy storage control instruction structure.

[0207] The instruction execution module 60 is used to synchronously send the energy storage control instruction structure and the power supply scheduling output results to the energy control module; establish a high-priority communication channel to complete clock synchronization and message verification; and collect on-site charging and discharging execution data, mode switching logs and equipment alarm information in real time to provide basic data for deviation analysis.

[0208] The model update module 70 is used to perform deviation analysis on model parameters such as energy storage power constraints, mode switching margin, state of charge continuity constraints and photovoltaic priority weights based on the energy consumption deviation feedback structure generated by the instruction execution module; complete parameter adjustment according to the recursive step-by-step principle; write the adjustment results into the multi-energy power supply optimization model after adding a write lock and trigger a version update event to achieve model adaptive closed-loop correction.

[0209] The system monitoring and visualization module 80 is used to monitor the operating status of each functional module in real time, displaying the photovoltaic-energy storage-grid power flow, peak-valley identification results, strategy execution trajectory, and model parameter evolution curve. It supports operation and maintenance personnel to review historical data online, download message logs, and manually trigger rollback or retraining operations to ensure the long-term stable operation of the system under complex working conditions.

[0210] Beneficial effects of the embodiment:

[0211] (1) Reduce costs and increase efficiency. By solving the multi-energy power supply optimization model in real time through the optimization modeling module and superimposing the peak-valley arbitrage strategy of the arbitrage analysis module, the comprehensive operating cost of the charging station is reduced by about 25 percentage points compared with the model that relies solely on the power grid.

[0212] (2) Green electricity is given priority, and the weight of the photovoltaic priority strategy is explicitly reflected in the model, maximizing the utilization rate of local renewable energy, reducing peak electricity purchases and reducing carbon emissions.

[0213] (3) Fast closed loop: The energy consumption deviation feedback structure drives the model update module to adaptively adjust constraints and weights within a minute-level cycle, realizing a high-speed closed loop of strategy-execution-feedback-correction, significantly improving scheduling accuracy and energy storage life.

[0214] (4) Safe and reliable. The instruction execution module adopts clock synchronization, write lock, version control and alarm linkage mechanism to ensure that the control instructions are implemented reliably and avoid the risks of energy storage overcharging, over-discharging or incorrect mode switching.

[0215] (5) Easy to expand. The system monitoring and visualization module provides an open interface that can be connected to new energy equipment such as wind power and hydrogen energy storage on demand, and has good horizontal expansion and industry migration capabilities.

[0216] Obviously, the embodiments described above are only some of the embodiments of the present application, rather than all of the embodiments. The preferred embodiments of the present application are given in the accompanying drawings, but they do not limit the patent scope of the present application. The present application can be implemented in many different forms. On the contrary, the purpose of providing these embodiments is to make the understanding of the disclosure of the present application more thorough and comprehensive. Although the present application has been described in detail with reference to the aforementioned embodiments, for those skilled in the art, it is still possible to modify the technical solutions described in the aforementioned specific embodiments, or to make equivalent replacements for some of the technical features therein. Any equivalent structure made using the contents of the present application specification and the accompanying drawings, directly or indirectly used in other related technical fields, is also within the scope of patent protection of the present application.

Claims

1. A charging pile background management method, characterized in that: The following steps are involved: Acquire photovoltaic output data, energy storage status information, and charging load data, complete time alignment, normalization, and structural integration, and generate a unified data input structure; Obtain a unified data input structure, configure photovoltaic priority strategies and energy storage scheduling constraint parameters, build and solve a multi-energy power supply optimization model, and generate power supply scheduling output results; The expression of the power supply scheduling output result is: in, Output structure for power supply scheduling; is the optimal charge and discharge power; is the optimal power purchase power; is the energy storage state trajectory at each moment; is the unit energy consumption scheduling cost; K is the time index variable; N is the total number of time steps in the scheduling period; Obtain power supply scheduling output and electricity price series, identify peak and valley periods, generate arbitrage tags, and build an arbitrage reference structure; Obtain the arbitrage reference structure and power supply dispatch output results, combine them with the load data in the unified data input structure, generate the energy storage charging and discharging control strategy and build the energy storage control instruction structure; Issue energy storage control instruction structure and power supply scheduling output results, collect execution data of energy control module, and generate energy consumption deviation feedback structure; Obtain energy consumption deviation feedback structure, analyze and adjust scheduling parameters in the multi-energy power supply optimization model, and complete model adaptive closed-loop correction; The analysis and adjustment of the scheduling parameters in the multi-energy power supply optimization model include: The peak deviation and average deviation are mapped to the maximum power constraint of energy storage charging and discharging; the cumulative deviation is mapped to the energy storage cycle number constraint; the number of switching failures and switching delays are mapped to the mode switching delay margin; the maximum state of charge offset is mapped to the energy storage efficiency coefficient; the average state of charge offset and state recovery time are mapped to the state of charge continuity constraint; The model adaptive closed-loop correction includes: The system performs fast consistency checks, including supply and demand balance constraint checks, energy storage power constraint checks, and charge state continuity checks.

2. The charging pile background management method according to claim 1, characterized in that: The acquisition of photovoltaic output data, energy storage status information and charging load data, and the completion of time alignment, normalization and structural integration include: Time-align PV output data, energy storage status information, and charging load data; Normalize and perform sliding average processing on photovoltaic output data, energy storage status information, and charging load data; The normalization and sliding average processing results are structurally integrated to generate a unified data input structure.

3. The charging pile background management method according to claim 1, characterized in that: The configuration of the photovoltaic priority strategy and energy storage scheduling constraint parameters includes: Obtain a unified data input structure and configure photovoltaic priority strategy and energy storage scheduling constraint parameters.

4. The charging pile background management method according to claim 1, characterized in that: The construction and solution of the multi-energy power supply optimization model includes: A multi-energy power supply optimization model is constructed, and the objective function includes the power purchase cost of the power grid and the energy loss of energy storage scheduling.

5. The charging pile background management method according to claim 1, characterized in that: The obtaining of the power supply scheduling output result and the electricity price sequence, identifying the peak period and the valley period, and generating the arbitrage label includes: Obtain power supply scheduling output results and electricity price series, and mark the peak and valley distribution of electricity prices; Analyze and process the peak and valley distribution of electricity prices to identify peak and valley periods; Generate arbitrage labels based on peak and valley periods to build an arbitrage reference structure.

6. The charging pile background management method according to claim 1, characterized in that: The generating of energy storage charge and discharge control strategy and building of energy storage control instruction structure includes: Determine the energy storage charging and discharging periods based on the arbitrage reference structure and power supply scheduling output results; Generate energy storage charging and discharging control strategies based on load data in a unified data input structure; The energy storage charging and discharging control strategy is structured to generate the energy storage control instruction structure.

7. The charging pile background management method according to claim 1, characterized in that: The steps of issuing the energy storage control instruction structure and the power supply scheduling output result and collecting the energy control module execution data include: Synchronously send the energy storage control instruction structure and power supply dispatch output results to the energy control module; Collect actual execution data from the energy control module and compare and analyze it with the issued instructions.

8. The charging pile background management method according to claim 1, characterized in that: The step of generating the energy consumption deviation feedback structure includes: The comparative analysis results are structured to generate an energy consumption deviation feedback structure.

9. A charging pile background management system, applied to the method according to any one of claims 1 to 8, characterized in that: include: Data acquisition module, used to obtain real-time photovoltaic output data, energy storage status information, charging load data and electricity price series from the photovoltaic array controller, energy storage management system, smart charging pile controller and billing platform; The data preprocessing module is used to perform time alignment, normalization, sliding average and structure integration processing on the raw data and output a unified data input structure; An optimization modeling module, which extracts available capacity and power ranges based on a unified data input structure, configures PV priority strategy weights, energy storage scheduling constraints, and state-of-charge continuity constraints; Arbitrage analysis module, which receives the power supply scheduling output and electricity price series, marks the peak and valley distribution of electricity prices, identifies peak and valley periods, and calculates derived indicators; The strategy generation module is used to obtain the arbitrage reference structure and power supply scheduling output results, and determine the energy storage charging and discharging switching points based on the load power requirements in the unified data input structure; The instruction execution module is used to synchronously send the energy storage control instruction structure and the power supply scheduling output results to the energy control module; The model update module is used to analyze the deviation of energy storage power constraints, mode switching margins, state of charge continuity constraints, and photovoltaic priority weight model parameters based on the energy consumption deviation feedback structure generated by the instruction execution module; The system monitoring and visualization module is used to monitor the operating status of each functional module in real time, displaying the photovoltaic-energy storage-grid power flow, peak and valley identification results, strategy execution trajectory and model parameter evolution curve.

Images