A coordinated control method and system of a super-charging station hybrid energy storage system

Abstract

The application discloses a kind of coordination control method and system of super charging station hybrid energy storage system, belong to super charging station control technical field, comprising: the real-time data and historical data of auxiliary operation data of super charging station photovoltaic side, energy storage side, super charging side, grid side are collected;Based on real-time data and historical data, future photovoltaic output, super charging load, rolling maximum demand and time-of-use electricity price are predicted using a prediction model according to a predetermined scheduling period;An electricity and demand cost collaborative optimization model with the goal of minimizing total electricity cost is constructed, and the optimal charging and discharging power instructions of solid-state lithium battery and sodium battery are solved by combining the prediction results and energy storage operation constraints;Based on the optimal charging and discharging power instructions, charging and discharging and charging scheduling operations are performed.This method can realize the coordinated regulation of super charging station, and ensure the economic and efficient operation of super charging station.

Description

Technical Field

[0001] This application belongs to the field of charging station control technology, and specifically relates to a coordinated control method and system for a supercharging station hybrid energy storage system. Background Technology

[0002] A supercharging station for new energy vehicles is a public infrastructure that uses high-power charging technology as its core to provide efficient and rapid energy replenishment services for new energy vehicles. Currently, new energy vehicle supercharging stations have significant shortcomings in core technology areas such as source-storage-charging coordinated control, energy storage system selection and configuration, coordinated optimization of dual electricity costs (electricity cost + demand cost), and short-term photovoltaic forecasting and integrated application. These shortcomings are as follows: 1. Overcharging often occurs during peak grid hours, resulting in high electricity costs. Furthermore, the load is characterized by short-term high power consumption, drastic fluctuations, and extreme peak-to-valley differences, which can easily lead to frequent over-limits of maximum demand and high demand-related electricity costs.

[0003] Existing technologies can only achieve peak shaving or peak-valley arbitrage, but cannot simultaneously reduce electricity costs for both electricity consumption and demand.

[0004] 2. Solid-state lithium-ion batteries alone cannot withstand frequent high-power charging and discharging, resulting in rapid degradation, short lifespan, and high cost. Sodium batteries have advantages such as high current tolerance, wide SOC, low cost, and long cycle life, making them suitable for peak load support, but current technologies have not yet achieved a reasonable division of labor with solid-state lithium-ion batteries.

[0005] 3. Without combining photovoltaic forecasting to achieve advanced synergy between power generation, energy storage, and charging, photovoltaic waste or sudden surges in demand are likely to occur.

[0006] Patent CN117498386A addresses the issues of large load fluctuations and high electricity costs at supercharging stations by constructing a hybrid energy storage collaborative architecture of "photovoltaic-solid-state lithium battery-sodium battery". This solution effectively alleviates the problems of load fluctuations and high electricity costs at supercharging stations, and achieves preliminary synergy between photovoltaic and hybrid energy storage. It provides basic technical support for the photovoltaic-energy storage collaborative control of supercharging stations, but it has not formed a systematic dual-electricity-cost collaborative optimization system, making it difficult to adapt to the actual needs of rolling scheduling of supercharging stations. Summary of the Invention

[0007] To address the aforementioned issues, this application provides a coordinated control method and system for a hybrid energy storage system in a supercharging station. By utilizing precise photovoltaic forecasting, dual-cost collaborative optimization, and differentiated division of labor control technology between solid-state lithium batteries and sodium batteries, the coordinated control and economical and efficient operation of the hybrid energy storage system in a supercharging station can be achieved.

[0008] The following is the technical content of this disclosure: A coordinated control method for a hybrid energy storage system in a supercharging station includes: Collect real-time and historical data from the photovoltaic side, energy storage side, supercharging side, grid side, and auxiliary operation data of the supercharging station; Based on the collected real-time and historical data, the predictive model is used to predict future photovoltaic output, overcharging load, rolling maximum demand and time-of-use electricity price according to the preset scheduling cycle. A collaborative optimization model for power consumption and demand consumption with the goal of minimizing total electricity costs is constructed. The optimal charging and discharging power commands for solid-state lithium batteries and sodium batteries are solved by combining prediction results and energy storage operation constraints. Execute charging / discharging and charging scheduling operations based on the optimal charging / discharging power command; The system updates the photovoltaic, energy storage, supercharging, grid, and auxiliary operation data, prediction model parameters, and collaborative optimization model parameters of the supercharging station on a rolling basis according to a preset cycle, so as to achieve coordinated control of the supercharging station's hybrid energy storage system.

[0009] Furthermore, the real-time data includes: Photovoltaic side: Photovoltaic output data, irradiance, ambient temperature, ambient humidity, wind speed; Energy storage side: Solid-state lithium battery state of charge and charge / discharge power; Sodium battery state of charge and charge / discharge power; Supercharging side: Real-time charging power, total charging load, and charging time for each charging station; Grid side: Real-time time-of-use electricity price, grid power exchange, maximum contracted demand, grid voltage, and grid frequency.

[0010] Furthermore, the prediction model predicts future photovoltaic output according to a preset scheduling cycle, including: Acquire historical data on irradiance, ambient temperature, and photovoltaic power output; Historical irradiance, ambient temperature, and photovoltaic power output data are input into an LSTM neural network model for training to obtain future photovoltaic power output predictions. At regular intervals, the predicted future photovoltaic output is revised using real-time data.

[0011] Furthermore, the step of using real-time data to correct the predicted future photovoltaic output includes: Real-time irradiance and ambient temperature data are collected periodically and compared with the predicted photovoltaic output to calculate the deviation rate. When the deviation rate exceeds the threshold, the future photovoltaic power output prediction value is corrected by the deviation ratio correction coefficient, and the corrected photovoltaic power output prediction value is output to the collaborative optimization model at regular intervals.

[0012] Furthermore, the collaborative optimization model is as follows: minC total ( k )= Celec ( k )+ C demand ( k ) in, C elec ( k The electricity cost (in yuan) for the k-th scheduling cycle is calculated using the following formula: C elec ( k )= P grid ( k )×Δ t × C ( k ) in, P grid ( k ) represents the power purchased from the grid in the k-th scheduling cycle (kW), Δt represents the duration of the scheduling cycle, and C(k) represents the time-of-use electricity price of the grid in the k-th scheduling cycle; C demand ( k Let be the demand charge for the k-th scheduling cycle, calculated using the following formula: C demand ( k )= max ( P demand (1), P demand (2), ..., P demand (k))× C d in, max ( P demand (1), P demand (2), ..., P demand (k) represents the rolling maximum power up to the k-th scheduling cycle; C d This is the unit price for electricity based on demand.

[0013] Furthermore, the collaborative optimization model is solved using a swarm optimization algorithm; The following constraints must be satisfied during the solution process: power balance constraint, grid demand constraint, state of charge constraint of solid-state lithium battery and sodium battery, energy storage power constraint, and photovoltaic absorption constraint. The power balance constraint is as follows: Pload ( k )= P pv ( k )+ P grid ( k )+ P b ( k )+ Pn ( k ) in, P load ( k () represents the total overcharging load in the k-th scheduling cycle; P pv ( k ) represents the predicted photovoltaic output for the k-th scheduling cycle; P grid ( k ) represents the power purchased from the grid during the k-th scheduling cycle; P b ( k ) represents the solid-state lithium battery power command for the kth scheduling cycle, with positive for discharging and negative for charging; Pn ( k ) represents the sodium power command for the k-th scheduling cycle, with positive indicating discharge and negative indicating charging; The grid demand constraint is: P grid ( k )≤ P th in, P grid ( k () represents the power purchased from the grid during the k-th scheduling cycle. P th This is the maximum demand threshold; The state of charge constraint of the solid-state lithium battery is as follows: SOC b,min ≤SOC b ( k )≤SOC b,max in, SOC b,min and SOC b,max These are the upper and lower limits of the state of charge of solid-state lithium batteries, respectively. The charge state constraint of the sodium battery is as follows: SOC n,min ≤SOC n ( k ) ≤SOC n,max in, SOCn,min and SOC n,max These are the upper and lower limits of the state of charge of sodium-ion batteries, respectively. The photovoltaic absorption constraint is: P pv ( k ) ≤P load ( k ) +|P b ( k,charge ) |+|P n ( k,charge ) | in, P pv ( k ) represents the predicted photovoltaic output for the k-th scheduling cycle. P load ( k Let be the total overcharging load in the k-th scheduling cycle. P b ( k,charge () represents the charging power of the solid-state lithium battery in the k-th scheduling cycle. P n ( k,charge ) represents the charging power of sodium batteries during the k-th scheduling cycle.

[0014] Furthermore, the execution of charge / discharge and charging scheduling operations based on the optimal charge / discharge power command includes: When the demand warning is triggered, the peak control mode is entered. Sodium batteries are discharged first to reduce the peak power purchase of the grid, and solid lithium batteries suspend arbitrage and help to smooth out power fluctuations. When demand warnings are not triggered, the charging and discharging power of solid-state lithium batteries and sodium batteries is allocated according to the time-of-use electricity price period and the photovoltaic output.

[0015] Furthermore, the peak control mode is as follows: Sodium-ion batteries preferentially discharge, suppressing the grid's power purchase capacity to within the demand warning threshold; When the sodium battery's SOC falls below the set lower limit, the discharge power is reduced, and the solid-state lithium battery supplements the discharge power. Solid-state lithium batteries suspend peak-valley arbitrage to help smooth out sudden fluctuations in supercharging load; photovoltaic power output prioritizes supplying supercharging load and does not charge energy storage.

[0016] A coordinated control system for a hybrid energy storage system in a supercharging station includes: The data acquisition module is used to collect real-time and historical data from the photovoltaic side, energy storage side, supercharging side, grid side, and auxiliary operation data of the supercharging station. The prediction module is used to predict future photovoltaic output, overcharging load, rolling maximum demand and time-of-use electricity price based on the collected real-time data and historical data, using a prediction model according to a preset scheduling cycle. The instruction generation module is used to construct a collaborative optimization model of power consumption and demand consumption with the goal of minimizing total electricity costs. It combines the prediction results and energy storage operation constraints to solve for the optimal charging and discharging power instructions for solid-state lithium batteries and sodium batteries. The scheduling module is used to perform charging and discharging scheduling operations based on the optimal charging and discharging power command; The feedback optimization module is used to update the photovoltaic side, energy storage side, supercharging side, grid side and auxiliary operation data, prediction model parameters and collaborative optimization model parameters of the supercharging station on a rolling basis according to a preset cycle, so as to realize the coordinated control of the supercharging station's hybrid energy storage system.

[0017] Furthermore, the collaborative optimization model is as follows: minC total ( k )= C elec ( k )+ C demand ( k ) in, C elec ( k The electricity cost (in yuan) for the k-th scheduling cycle is calculated using the following formula: C elec ( k )= P grid ( k )×Δ t × C ( k ) in, P grid ( k Let Δt be the power (kW) purchased from the grid in the k-th scheduling cycle, Δt be the duration of the scheduling cycle, and C(k) be the time-of-use electricity price of the grid in the k-th scheduling cycle. C demand ( k Let be the demand charge for the k-th scheduling cycle, calculated using the following formula: C demand ( k )= max ( P demand (1), P demand (2), ...,P demand (k))× C d in, max ( P demand (1), P demand (2), ..., P demand (k) represents the rolling maximum power up to the k-th scheduling cycle; C d This refers to the unit price of electricity based on demand. The collaborative optimization model is solved using a swarm optimization algorithm. The following constraints must be satisfied during the solution process: power balance constraint, grid demand constraint, state of charge constraint of solid-state lithium battery and sodium battery, energy storage power constraint, and photovoltaic absorption constraint. The power balance constraint is as follows: P load ( k )= P pv ( k )+ P grid ( k )+ P b ( k )+ Pn ( k ) in, P load ( k () represents the total overcharging load in the k-th scheduling cycle; P pv ( k ) represents the predicted photovoltaic output for the k-th scheduling cycle; P grid ( k ) represents the power purchased from the grid during the k-th scheduling cycle; P b ( k ) represents the solid-state lithium battery power command for the kth scheduling cycle, with positive for discharging and negative for charging; Pn ( k ) represents the sodium power command for the k-th scheduling cycle, with positive indicating discharge and negative indicating charging; The grid demand constraint is: P grid ( k )≤ P th in, P grid (k () represents the power purchased from the grid during the k-th scheduling cycle. P th This is the maximum demand threshold; The state of charge constraint of the solid-state lithium battery is as follows: SOC b,min ≤SOC b ( k )≤SOC b,max in, SOC b,min and SOC b,max These are the upper and lower limits of the state of charge of solid-state lithium batteries, respectively. The charge state constraint of the sodium battery is as follows: SOC n,min ≤SOC n ( k ) ≤SOC n,max in, SOC n,min and SOC n,max These are the upper and lower limits of the state of charge of sodium-ion batteries, respectively. The photovoltaic absorption constraint is: P pv ( k ) ≤P load ( k ) +|P b ( k,charge ) |+|P n ( k,charge ) | in, P pv ( k ) represents the predicted photovoltaic output for the k-th scheduling cycle. P load ( k Let be the total overcharging load in the k-th scheduling cycle. P b ( k,charge () represents the charging power of the solid-state lithium battery in the k-th scheduling cycle. P n ( k,charge ) represents the charging power of sodium batteries during the k-th scheduling cycle.

[0018] Compared with the prior art, this application has the following advantages: By collecting real-time and historical data from the photovoltaic, energy storage, supercharging, grid, and auxiliary operation data of supercharging stations, this technology provides comprehensive, multi-time-series basic data support for photovoltaic output, supercharging load, demand and electricity price forecasting, and dual-feed collaborative optimization. Since the data covers the entire process of source, storage, charging, and grid and takes into account both real-time and historical dimensions, the comprehensiveness and accuracy of the data can support accurate prediction and optimization. Therefore, it can solve the problems mentioned in the background technology, such as the lack of integrated photovoltaic forecasting to achieve advanced source-storage-charging synergy, and the easy occurrence of photovoltaic waste or sudden demand surges. This application achieves simultaneous joint optimization of electricity cost and demand cost by constructing a collaborative optimization model for electricity cost and demand cost with the goal of minimizing total electricity cost. It combines prediction results with energy storage operation constraints to solve for the optimal charging and discharging power commands of solid-state lithium batteries and sodium batteries. It also matches the characteristics of solid-state lithium batteries and sodium batteries to carry out differentiated power allocation. Because it breaks the limitations of single electricity cost optimization and gives full play to the advantages of peak-valley arbitrage of solid-state lithium batteries and the ability of sodium batteries to withstand high current and voltage peaks, it can solve the problems of existing technologies that only achieve peak shaving or peak-valley arbitrage, cannot simultaneously reduce electricity cost and demand cost, and solid-state lithium battery energy storage suffers from rapid degradation, short lifespan, and high cost due to high power frequent charging and discharging, and sodium batteries do not form a reasonable division of labor with solid-state lithium batteries.

[0019] Other features and advantages of this application will be set forth in the description which follows, and will be apparent in part from the description, or may be learned by practicing the application. The objectives and other advantages of this application may be realized and obtained by means of the structures pointed out in the description, claims and drawings. Attached Figure Description

[0020] To more clearly illustrate the technical solutions in the embodiments of this application or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are some embodiments of this application. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0021] Figure 1 A schematic diagram of the method of this disclosure is shown. Detailed Implementation

[0022] To make the objectives, technical solutions, and advantages of the embodiments of this application clearer, the technical solutions of the embodiments of this application will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of this application, not all embodiments. Based on the embodiments of this application, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of this application.

[0023] likeFigure 1 This is a schematic diagram of the method disclosed herein, which specifically includes: The overall architecture of this solution is divided into five layers: perception layer, prediction layer, optimization layer, control layer, and execution layer. These layers work together to achieve deep integration of the power source (photovoltaics), energy storage (solid-state lithium batteries + sodium batteries), load (supercharging), and grid. The core architecture is as follows: 1) Sensing Layer: Responsible for real-time collection of various operational data from the supercharging station, providing basic support for prediction, optimization, and control. The collection frequency is once every 1 minute. Core collected data includes: - Photovoltaic side: Photovoltaic output data, irradiance, ambient temperature (used for photovoltaic output correction); - Energy storage side: Solid-state lithium battery SOC (State of Charge), charge / discharge power; Sodium battery SOC, charge / discharge power; - Supercharging side: Real-time charging power, total charging load, and charging time for each charging station; - Grid side: Real-time time-of-use electricity price, grid interconnection power, maximum contracted demand, grid voltage / frequency; - Auxiliary data: ambient humidity, wind speed (used for photovoltaic power output prediction correction).

[0024] 2) Prediction Layer: With a scheduling cycle of 15 minutes, it proactively predicts various parameters, providing input for the optimization layer. Prediction accuracy is ≥90%. Core prediction content includes: - Photovoltaic output prediction: Based on historical irradiance data and real-time environmental data, the "LSTM + photovoltaic power correction" algorithm is used to predict the average photovoltaic output Ppv(k) for the next 15 minutes (k is the kth 15-minute scheduling cycle). - Supercharging load prediction: Based on historical charging data and time period characteristics (such as peak hours and holidays), predict the total load Pload(k) of the supercharging station in the next 15 minutes; - Rolling maximum demand forecast: Based on the grid interaction power of the previous scheduling cycle and the current overcharging load change trend, predict the rolling maximum demand Pdemand(k) for the next 15 minutes. - Time-of-use pricing confirmation: Simultaneously obtain the current and next 15 minutes' time-of-use electricity price C(k) (peak / off-peak / valley price). 3) Optimization layer: The core is to build a dual-electricity-cost collaborative optimization model. With the goal of "minimizing total electricity cost", it combines prediction data, energy storage characteristics and constraints to solve for the optimal charging and discharging power commands of solid-state lithium batteries and sodium batteries. The solution is rolled once every 15 minutes. 4) Control layer: Receives power commands output from the optimization layer, combines them with real-time operating status (such as energy storage SOC, sudden changes in overcharge load), corrects the commands, and outputs executable control signals to ensure control accuracy and operational safety. 5) Execution layer: including photovoltaic inverters, lithium battery energy storage converters (PCS), sodium battery energy storage converters (PCS), and supercharging pile controllers, receiving signals from the control layer and performing operations such as charging and discharging and charging scheduling to achieve optimization goals.

[0025] The following are embodiments based on the above architecture: Module 1: Construct a photovoltaic forecasting module with the next 15 minutes as the forecast node. This module is the prerequisite for the entire solution, addressing the problem of "the randomness of photovoltaic power output leading to lag in energy storage charging and discharging and excessive demand." It adopts a two-stage prediction method of "historical data training + real-time data correction," and the specific design is as follows: (1) Data preprocessing: Collect photovoltaic power output data, irradiance data and ambient temperature data for the past year, remove outliers, use linear interpolation to supplement missing data, and divide the data into 15-minute segments to form a training dataset.

[0026] (2) Model training: The basic prediction model is constructed using LSTM neural network. The input parameters are "historical 15-minute irradiance, ambient temperature, and photovoltaic power output", and the output parameter is "average photovoltaic power output in the next 15 minutes". The parameters of the model are optimized by gradient descent algorithm to ensure that the prediction error is ≤10%.

[0027] (3) Real-time correction: The prediction layer collects real-time irradiance and ambient temperature data every minute and compares them with the predicted values. If the deviation exceeds 5%, the photovoltaic output prediction value for the next 15 minutes will be corrected to ensure that the prediction accuracy matches the actual operating scenario.

[0028] The specific correction logic is as follows: First, calculate the deviation rate between the real-time data and the predicted value. After the correction condition is triggered, adjust the basic predicted value through the deviation ratio correction coefficient. Combine the weather change trend in the past 3 minutes to make a secondary prediction correction. At the same time, follow the physical boundary constraints (the predicted output is not less than 0 and does not exceed the rated output of the photovoltaic module) and dynamically adapt to weather fluctuations. The correction follows the principle of "real-time monitoring, deviation triggering, and continuous correction" and has no fixed number of times. It stops 3 minutes before the end of the scheduling cycle.

[0029] (4) Output results: 3 minutes before the end of each scheduling cycle, output the predicted photovoltaic output value for the next scheduling cycle. P pv ( k This data is synchronously transmitted to the optimization layer as the core input for power allocation.

[0030] 2. Module 2: Hybrid Energy Storage Division of Labor and Characteristic Adaptation Module This module addresses the pain point of "unclear division of labor between solid-state lithium batteries and sodium batteries, and failure to fully leverage the advantages of hybrid energy storage." By combining the characteristics of both energy storage methods, it achieves "each performing its own function and synergistic efficiency." The specific design is as follows: (1) Energy storage characteristics adaptation Solid-state lithium battery energy storage is responsible for peak-valley arbitrage, reducing electricity consumption and electricity costs; Sodium-ion battery storage suppresses high-power peaks in supercharging, reducing demand-based electricity costs.

[0031] (2) Demand control priority: Regardless of the operating scenario, as long as the demand warning threshold is approached, sodium batteries will be started first, and solid-state lithium batteries will suspend arbitrage tasks. In conjunction with sodium batteries, fluctuations will be suppressed to ensure that the demand does not exceed the limit. (3) Optimal energy utilization: During off-peak hours and peak photovoltaic power generation periods, priority is given to charging sodium batteries (to ensure peak suppression capability the next day), and the remaining electricity is then supplied to charge solid-state lithium batteries (for peak-valley arbitrage). (4) Division of labor for fluctuation suppression: Within a 15-minute scheduling cycle, high-frequency fluctuations are suppressed by sodium batteries with rapid response, while low-frequency fluctuations are gradually supported by solid-state lithium batteries; (5) Fault redundancy: If sodium batteries fail, solid-state lithium batteries can temporarily take on part of the demand control task; if solid-state lithium batteries fail, sodium batteries can temporarily take on part of the arbitrage task.

[0032] 3. Module 3: Co-optimization model for electricity consumption and demand-based electricity costs (1) Objective function: The core objective is to minimize the total electricity cost of the supercharging station in every 15-minute scheduling cycle. The objective function is as follows: minC total ( k )= C elec ( k )+ C demand ( k ) in, C elec ( k The electricity cost (in yuan) for the k-th scheduling cycle is calculated using the following formula: C elec ( k )= P grid ( k )×Δ t × C ( k ) in, P grid ( k) represents the power purchased from the grid in the k-th scheduling cycle (kW), Δt represents the duration of the scheduling cycle, and C(k) represents the time-of-use electricity price of the grid in the k-th scheduling cycle; C demand ( k Let be the demand charge for the k-th scheduling cycle, calculated using the following formula: C demand ( k )= max ( P demand (1), P demand (2), ..., P demand (k))× C d in, max ( P demand (1), P demand (2), ..., P demand (k) represents the rolling maximum power up to the k-th scheduling cycle; C d The unit price is based on demand and is executed according to the power grid contract.

[0033] (2) Constraints: To ensure the feasibility of the optimization results and the safety of operation, the following constraints are set. All constraints are designed in combination with the actual operation scenario and energy storage characteristics of the supercharging station.

[0034] -Power balance constraints: P load ( k )= P pv ( k )+ P grid ( k )+ P b ( k )+ Pn ( k ) in, P load ( k () represents the total overcharging load in the k-th scheduling cycle; P pv ( k ) represents the predicted photovoltaic output for the k-th scheduling cycle; P b ( k ) represents the solid-state lithium battery power command for the kth scheduling cycle, with positive for discharging and negative for charging;Pn ( k ) represents the sodium power command for the kth scheduling cycle, with positive for discharging and negative for charging.

[0035] -Demand constraint: P grid ( k )≤ P th The demand warning threshold is set to 95% of the historical maximum demand to prevent demand from exceeding the limit.

[0036] - Energy storage SOC constraint: Solid-state lithium batteries: SOC b,min ≤SOC b ( k )≤SOC b,max ,in SOC b,min =20%, SOC b,max =80%; Sodium electricity: SOC n,min ≤SOC n ( k ) ≤SOC n,max ,in SOC n,min =10%, SOC n,max =90%.

[0037] - Energy storage power constraints: Solid-state lithium batteries: - P b,rated ≤ P b ( k )≤ P b,rated Sodium electricity: - P n,rated ≤ P n ( k )≤ P n,rated ; P b,rated 、P n,rated These are the rated power (kW) of solid-state lithium batteries and sodium batteries, respectively.

[0038] -Constraints on photovoltaic power absorption: P pv ( k ) ≤P load ( k ) +|P b ( k,charge ) |+|Pn ( k,charge ) | in, P pv ( k ) represents the predicted photovoltaic output for the k-th scheduling cycle. P load ( k Let be the total overcharging load in the k-th scheduling cycle. P b ( k,charge () represents the charging power of the solid-state lithium battery in the k-th scheduling cycle. P n ( k,charge ) represents the charging power of sodium batteries during the k-th scheduling cycle.

[0039] Ensure that photovoltaic power output is prioritized for use by supercharging loads, and that surplus electricity is prioritized for absorption by energy storage, with a photovoltaic absorption rate of ≥95%.

[0040] Optimize the solution algorithm: The optimization model is solved using a "group optimization algorithm," which has low computational cost and fast convergence speed, making it suitable for 15-minute rolling optimization and compatible with the deployment of supercharging station field controllers. The specific optimization steps are as follows: - Initialization parameters: Set the number of algorithm iterations (100-150 times), population size (30-50), and the range of values ​​for optimization variables Pb(k) and Pn(k); - Input forecast data: Input 15-minute photovoltaic forecast, supercharging load forecast, time-of-use electricity price, and demand forecast data into the algorithm; - Fitness calculation: Using the minimization of total electricity cost as the fitness function, calculate the fitness value of each individual in the population, and at the same time check whether all constraints are satisfied; Iterative optimization: By iteratively updating the population, the individuals with the best fitness are retained until the upper limit of the number of iterations is reached; Output results: Output the solid-state lithium battery power command Pb(k) and sodium battery power command Pn(k) that satisfy all constraints and have the minimum total power cost, and transmit them to the control layer.

[0041] 4. Module 4: 15-minute rolling control module (1) Step 1: Periodic initialization The prediction layer outputs the photovoltaic output prediction value Ppv(k), the overcharging load prediction value Pload(k), the time-of-use electricity price C(k), and the rolling maximum demand prediction value Pdemand(k) for the current scheduling cycle; the control layer reads the solid-state lithium battery SOCb(k), sodium battery SOCn(k), and the grid interaction power Pgrid(k-1) of the previous cycle to complete the initialization.

[0042] Step 2: Calculation of purchased power and judgment of demand warning Calculate the net power in the current dispatching period: Pnet(k) = Pload(k) - Ppv(k); Judge whether to trigger the demand warning: If Pnet(k)≥Pth, trigger the demand warning and enter the peak control mode; if Pnet(k)<Pth, no warning is triggered and enter the conventional optimization mode.

[0043] Step 3: Cooperative processing control of hybrid energy storage - Mode 1: Peak control mode (Pnet(k)≥Pth, demand priority) Sodium battery: Prioritize starting high-power discharge, and the discharge power Pn(k)=Pnet(k)-Pth, forcing the grid-purchased power within the demand warning threshold; if the SOC of the sodium battery is lower than 10%, reduce the discharge power, and the solid-state lithium battery supplements Pn(k)=Pnet(k)-Pth-Pn(k); Solid-state lithium battery: Pause the peak-valley arbitrage task, only used to supplement fluctuations. If there is a sudden mutation in the overcharge load, the solid-state lithium battery responds quickly to suppress fluctuations and ensure that the grid interaction power does not exceed Pth; Photovoltaic: Prioritize supplying the overcharge load and do not participate in energy storage charging to ensure maximum support for the load and reduce grid power purchase.

[0044] - Mode 2: Conventional optimization mode (P(k)<Pth, optimal for double electricity charges) Sub-mode 2.1: Peak electricity price period Solid-state lithium battery: Discharge for arbitrage, and the discharge power Pb(k) = min(Pb, rated, Pnet(k)), discharging to the maximum extent to reduce the electricity charge for electricity consumption; Sodium battery: Keep the high SOC on standby, neither discharging nor charging, and ready to respond to sudden peaks at any time; Sub-mode 2.2: Flat electricity price period Solid-state lithium battery: Charge and discharge in small amounts, adjust the charge and discharge power according to the photovoltaic output and the fluctuation of the overcharge load, maximize self-use, and reduce grid power purchase; Sodium battery: Maintain the SOC at 50%-70% in the standby state. If there is a small increase in the load, the sodium battery discharges slightly to avoid triggering the demand warning; Sub-mode 2.3: Valley electricity price period Sodium battery: Prioritize charging, and the charging power P n ( k ) = min ( P n,rated , P grid,max}- Pnet ( k Until SOC reaches 90%, ensuring peak suppression capability for the next day; Solid-state lithium battery: After the sodium battery is fully charged, it is then charged again, charging power P b ( k ) = min ( P b,rated , P grid,max - P net ( k ) - | Pn ( k )|), used for arbitrage during peak hours; Sub-mode 2.4: Peak photovoltaic power generation period ( P pv ( k )≥ P load ( k )) Solid-state lithium batteries: rapidly absorb excess photovoltaic power, charging power P b ( k )= min ( P b,rated , P pv ( k )- P load ( k Responding to photovoltaic fluctuations; Sodium-ion batteries: absorb the remaining photovoltaic power, providing charging power. P n ( k )= P pv ( k )- P load ( k )-| P b ( k until SOC reaches 90%.

[0045] Step 4: Instruction Modification and Execution The control layer receives the output of the optimization layer. P b ( k ), P n ( k Combined with real-time acquired SOC and power data, the command is corrected: - If the SOC of the solid-state lithium battery is below 20%, the solid-state lithium battery discharge will be forcibly stopped and the charging mode will be switched (prioritizing the absorption of photovoltaic power); if it is above 80%, the solid-state lithium battery charging will be forcibly stopped. - If the sodium battery SOC is below 10%, the sodium battery discharge will be forcibly stopped and the charging mode will be switched; if it is above 90%, the sodium battery charging will be forcibly stopped. - If the overcharging load suddenly changes (deviation from the predicted value exceeds 10%), the power commands for sodium batteries and solid-state lithium batteries will be adjusted in real time to ensure that the demand does not exceed the limit. - The corrected power command is transmitted to the execution layer, where it is executed by the solid-state lithium battery PCS, sodium battery PCS, and photovoltaic inverter. At the same time, the execution results are collected in real time and fed back to the sensing layer.

[0046] Step 5: Rolling Updates Record the actual operating data (PV output, overcharging load, grid interaction power, energy storage SOC, electricity cost) of the current scheduling cycle to correct the prediction model and optimization parameters for the next scheduling cycle, achieve 15-minute rolling optimization, and ensure that the control effect remains optimal.

[0047] 5. Module 5: Security Protection and Redundancy Module To ensure the safe operation of supercharging stations, energy storage systems, and the power grid, the following safety protection mechanisms are implemented to prevent safety hazards caused by equipment failures and sudden power surges: (1) Overcurrent / overvoltage protection: When the charging and discharging current and cell voltage of solid lithium battery or sodium battery exceed the threshold, the PCS will automatically shut down, the control layer will issue an alarm signal, and switch to redundant control mode at the same time. (2) Temperature protection: Real-time monitoring of the energy storage battery temperature. If the temperature exceeds 45°C, the charging and discharging power will be automatically reduced; if the temperature exceeds 55°C, the machine will be forced to shut down and the cooling system will be activated. (3) Demand overload protection: If the power of grid interaction exceeds the maximum contracted demand due to sudden load changes, sodium batteries will be forced to discharge at full power, solid lithium batteries will assist in the discharge, and unnecessary overcharging piles (such as idle charging piles) will be cut off to quickly reduce the power to within the threshold. (4) Redundancy control: If a certain energy storage system fails (such as a solid-state lithium battery PCS failure), the control layer automatically adjusts the power command of another energy storage system to ensure that the core demand control target does not fail; if the photovoltaic prediction fails, it switches to the "real-time power control mode" and performs power allocation based on real-time data.

[0048] Based on the method of this disclosure, embodiments of this disclosure also provide a system corresponding to the above method, which includes: The data acquisition module is used to collect real-time and historical data from the photovoltaic side, energy storage side, supercharging side, grid side, and auxiliary operation data of the supercharging station. The prediction module is used to predict future photovoltaic output, overcharging load, rolling maximum demand and time-of-use electricity price based on the collected real-time data and historical data, using a prediction model according to a preset scheduling cycle. The instruction generation module is used to construct a collaborative optimization model of power consumption and demand consumption with the goal of minimizing total electricity costs. It combines the prediction results and energy storage operation constraints to solve for the optimal charging and discharging power instructions for solid-state lithium batteries and sodium batteries. The scheduling module is used to perform charging and discharging scheduling operations based on the optimal charging and discharging power command; The feedback optimization module is used to update the photovoltaic side, energy storage side, supercharging side, grid side and auxiliary operation data, prediction model parameters and collaborative optimization model parameters of the supercharging station on a rolling basis according to a preset cycle, so as to realize the coordinated control of the supercharging station's hybrid energy storage system.

[0049] Although this application has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some of the technical features; and these modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the embodiments of this application.

Claims

1. A coordinated control method for a hybrid energy storage system in a supercharging station, characterized in that, include: Collect real-time and historical data from the photovoltaic side, energy storage side, supercharging side, grid side, and auxiliary operation data of the supercharging station; Based on the collected real-time and historical data, the predictive model is used to predict future photovoltaic output, overcharging load, rolling maximum demand and time-of-use electricity price according to the preset scheduling cycle. A collaborative optimization model for power consumption and demand consumption with the goal of minimizing total electricity costs is constructed. The optimal charging and discharging power commands for solid-state lithium batteries and sodium batteries are solved by combining prediction results and energy storage operation constraints. Execute charging / discharging and charging scheduling operations based on the optimal charging / discharging power command; The system updates the photovoltaic, energy storage, supercharging, grid, and auxiliary operation data, prediction model parameters, and collaborative optimization model parameters of the supercharging station on a rolling basis according to a preset cycle, so as to achieve coordinated control of the supercharging station's hybrid energy storage system.

2. The method according to claim 1, characterized in that, The real-time data includes: Photovoltaic side: Photovoltaic output data, irradiance, ambient temperature; Energy storage side: Solid-state lithium battery state of charge and charge / discharge power; Sodium battery state of charge and charge / discharge power; Supercharging side: Real-time charging power, total charging load, and charging time for each charging station; Grid side: Real-time time-of-use electricity price, grid power exchange, maximum contracted demand, grid voltage, and grid frequency.

3. The method according to claim 2, characterized in that, The prediction model predicts future photovoltaic output according to a preset scheduling cycle, including: Acquire historical data on irradiance, ambient temperature, and photovoltaic power output; Historical irradiance, ambient temperature, and photovoltaic power output data are input into an LSTM neural network model for training to obtain future photovoltaic power output predictions. At regular intervals, the predicted future photovoltaic output is revised using real-time data.

4. The method according to claim 3, characterized in that, The method of correcting future photovoltaic power output forecasts using real-time data includes: Real-time irradiance and ambient temperature data are collected periodically and compared with the predicted photovoltaic output to calculate the deviation rate. When the deviation rate exceeds the threshold, the future photovoltaic power output prediction value is corrected by the deviation ratio correction coefficient, and the corrected photovoltaic power output prediction value is output to the collaborative optimization model at regular intervals.

5. The method according to claim 1, characterized in that, The collaborative optimization model is as follows: minC total ( k )= C elec ( k )+ C demand ( k ) in, C elec ( k The electricity cost (in yuan) for the k-th scheduling cycle is calculated using the following formula: C elec ( k )= P grid ( k )×Δ t × C ( k ) in, P grid ( k ) represents the power purchased from the grid in the k-th scheduling cycle (kW), Δt represents the duration of the scheduling cycle, and C(k) represents the time-of-use electricity price of the grid in the k-th scheduling cycle; C demand ( k Let be the demand charge for the k-th scheduling cycle, calculated using the following formula: C demand ( k )= max ( P demand (1)， P demand (2)，...， P demand (k))× C d in, max ( P demand (1), P demand (2), ..., P demand (k) represents the rolling maximum power up to the k-th scheduling cycle; C d This is the unit price for electricity based on demand.

6. The method according to claim 5, characterized in that, The collaborative optimization model is solved using a swarm optimization algorithm. The following constraints must be satisfied during the solution process: power balance constraint, grid demand constraint, state of charge constraint of solid-state lithium battery and sodium battery, energy storage power constraint, and photovoltaic absorption constraint. The power balance constraint is as follows: P load ( k )= P pv ( k )+ P grid ( k )+ P b ( k )+ Pn ( k ) in, P load ( k () represents the total overcharging load in the k-th scheduling cycle; P pv ( k ) represents the predicted photovoltaic output for the k-th scheduling cycle; P grid ( k ) represents the power purchased from the grid during the k-th scheduling cycle; P b ( k ) represents the solid-state lithium battery power command for the kth scheduling cycle, with positive for discharging and negative for charging; Pn ( k ) represents the sodium power command for the k-th scheduling cycle, with positive indicating discharge and negative indicating charging; The grid demand constraint is: P grid ( k )≤ P th in, P grid ( k () represents the power purchased from the grid during the k-th scheduling cycle. P th This is the maximum demand threshold; The state of charge constraint of the solid-state lithium battery is as follows: SOC b,min ≤SOC b ( k )≤SOC b,max in, SOC b,min and SOC b,max These are the upper and lower limits of the state of charge of solid-state lithium batteries, respectively. The charge state constraint of the sodium battery is as follows: SOC n,min ≤SOC n ( k ) ≤SOC n,max in, SOC n,min and SOC n,max These are the upper and lower limits of the state of charge of sodium-ion batteries, respectively. The photovoltaic absorption constraint is: P pv ( k ) ≤P load ( k ) +|P b ( k,charge ) |+|P n ( k,charge ) | in, P pv ( k ) represents the predicted photovoltaic output for the k-th scheduling cycle. P load ( k Let be the total overcharging load in the k-th scheduling cycle. P b ( k,charge () represents the charging power of the solid-state lithium battery in the k-th scheduling cycle. P n ( k,charge ) represents the charging power of sodium batteries during the k-th scheduling cycle.

7. The method according to claim 1, characterized in that, The process of executing charge / discharge and charging scheduling operations based on optimal charge / discharge power commands includes: When the demand warning is triggered, the peak control mode is entered. Sodium batteries are discharged first to reduce the peak power purchase of the grid, and solid lithium batteries suspend arbitrage and help to smooth out power fluctuations. When demand warnings are not triggered, the charging and discharging power of solid-state lithium batteries and sodium batteries is allocated according to the time-of-use electricity price period and the photovoltaic output.

8. The method according to claim 7, characterized in that, The peak control mode is as follows: Sodium-ion batteries preferentially discharge, suppressing the grid's power purchase capacity to within the demand warning threshold; When the sodium battery's SOC falls below the set lower limit, the discharge power is reduced, and the solid-state lithium battery supplements the discharge power. Solid-state lithium batteries suspend peak-valley arbitrage to help smooth out sudden fluctuations in supercharging load; photovoltaic power output prioritizes supplying supercharging load and does not charge energy storage.

9. A coordinated control system for a hybrid energy storage system in a supercharging station, characterized in that, include: The data acquisition module is used to collect real-time and historical data from the photovoltaic side, energy storage side, supercharging side, grid side, and auxiliary operation data of the supercharging station. The prediction module is used to predict future photovoltaic output, overcharging load, rolling maximum demand and time-of-use electricity price based on the collected real-time data and historical data, using a prediction model according to a preset scheduling cycle. The instruction generation module is used to construct a collaborative optimization model of power consumption and demand consumption with the goal of minimizing total electricity costs. It combines the prediction results and energy storage operation constraints to solve for the optimal charging and discharging power instructions for solid-state lithium batteries and sodium batteries. The scheduling module is used to perform charging and discharging scheduling operations based on the optimal charging and discharging power command; The feedback optimization module is used to update the photovoltaic side, energy storage side, supercharging side, grid side and auxiliary operation data, prediction model parameters and collaborative optimization model parameters of the supercharging station on a rolling basis according to a preset cycle, so as to realize the coordinated control of the supercharging station's hybrid energy storage system.

10. The system according to claim 9, characterized in that, The collaborative optimization model is as follows: minC total ( k )= C elec ( k )+ C demand ( k ) in, C elec ( k The electricity cost (in yuan) for the k-th scheduling cycle is calculated using the following formula: C elec ( k )= P grid ( k )×Δ t × C ( k ) in, P grid ( k ) represents the power purchased from the grid in the k-th scheduling cycle (kW), Δt represents the duration of the scheduling cycle, and C(k) represents the time-of-use electricity price of the grid in the k-th scheduling cycle; C demand ( k Let be the demand charge for the k-th scheduling cycle, calculated using the following formula: C demand ( k )= max ( P demand (1)， P demand (2)，...， P demand (k))× C d in, max ( P demand (1), P demand (2), ..., P demand (k) represents the rolling maximum power up to the k-th scheduling cycle; C d This refers to the unit price of electricity based on demand. The collaborative optimization model is solved using a swarm optimization algorithm. The following constraints must be satisfied during the solution process: power balance constraint, grid demand constraint, state of charge constraint of solid-state lithium battery and sodium battery, energy storage power constraint, and photovoltaic absorption constraint. The power balance constraint is as follows: P load ( k )= P pv ( k )+ P grid ( k )+ P b ( k )+ Pn ( k ) in, P load ( k () represents the total overcharging load in the k-th scheduling cycle; P pv ( k ) represents the predicted photovoltaic output for the k-th scheduling cycle; P grid ( k ) represents the power purchased from the grid during the k-th scheduling cycle; P b ( k ) represents the solid-state lithium battery power command for the kth scheduling cycle, with positive for discharging and negative for charging; Pn ( k ) represents the sodium power command for the k-th scheduling cycle, with positive indicating discharge and negative indicating charging; The grid demand constraint is: P grid ( k )≤ P th in, P grid ( k () represents the power purchased from the grid during the k-th scheduling cycle. P th This is the maximum demand threshold; The state of charge constraint of the solid-state lithium battery is as follows: SOC b,min ≤SOC b ( k )≤SOC b,max in, SOC b,min and SOC b,max These are the upper and lower limits of the state of charge of solid-state lithium batteries, respectively. The charge state constraint of the sodium battery is as follows: SOC n,min ≤SOC n ( k ) ≤SOC n,max in, SOC n,min and SOC n,max These are the upper and lower limits of the state of charge of sodium-ion batteries, respectively. The photovoltaic absorption constraint is: P pv ( k ) ≤P load ( k ) +|P b ( k,charge ) |+|P n ( k,charge ) | in, P pv ( k ) represents the predicted photovoltaic output for the k-th scheduling cycle. P load ( k Let be the total overcharging load in the k-th scheduling cycle. P b ( k,charge () represents the charging power of the solid-state lithium battery in the k-th scheduling cycle. P n ( k,charge ) represents the charging power of sodium batteries during the k-th scheduling cycle.

Images