Energy storage system health state scheduling method based on dynamic equivalent modeling

Abstract

The invention discloses an energy storage system health state scheduling method based on dynamic equivalent modeling. The method comprises the steps of obtaining load prediction data of a power grid and power prediction data of renewable energy sources in a future predetermined time window; calculating original grid-connected power prediction data of each period in a future predetermined time window according to the two data; generating a plurality of candidate operation schemes according to the original grid-connected power prediction data and a power grid safe operation rule; calculating the power grid service income which can be obtained by each candidate operation scheme according to a power grid service income pricing rule; acquiring real-time state parameters of the energy storage system; inputting the real-time state parameters and each candidate operation scheme into a pre-established dynamic equivalent aging model, and outputting the battery aging cost of each candidate operation scheme; calculating the net income of each candidate operation scheme according to the power grid service income and the battery aging cost, and determining a target operation scheme with the highest net income; and controlling the energy storage system to operate according to the target operation scheme.

Description

Technical Field

[0001] This application relates to the field of power management technology, and in particular to a health state scheduling method for energy storage systems based on dynamic equivalent modeling. Background Technology

[0002] As renewable energy sources such as wind and solar power are increasingly integrated into the power system, their impact is growing. Specifically, the power generation capacity of renewable energy fluctuates with changes in wind and solar power, easily causing grid frequency fluctuations and short-term supply-demand imbalances. To address this issue, a common approach is to enhance the grid's regulation capabilities through electrochemical energy storage systems, such as high-energy-density lithium-ion battery energy storage. By rationally scheduling and controlling these energy storage systems, the fluctuations in renewable energy can be mitigated, ensuring the safe, stable, and efficient operation of the power grid.

[0003] In the traditional scheduling mode, it is necessary to formulate charging and discharging scheduling instructions for the energy storage system. The traditional mode mainly formulates charging and discharging scheduling instructions based on the real-time status parameters of the battery (voltage, current, SOC (State of Charge) etc.) and instructions from the superior scheduling agency.

[0004] However, this scheduling method has the following problems: On the one hand, the scheduling instructions only focus on the current SOC and power capacity, without fully considering that different operating strategies (such as charge / discharge rate and depth) will accelerate battery aging and adversely affect battery life, thus lacking effective protection for battery life. On the other hand, the scheduling instructions only calculate the energy revenue (such as peak-valley price difference) and ignore the lifespan degradation cost caused by battery aging. The revenue assessment is incomplete, which can easily lead to overuse or overly conservative scheduling behavior, ultimately resulting in poor overall economic benefits of the energy storage system.

[0005] Based on this, this application proposes a health status scheduling method for energy storage systems based on dynamic equivalent modeling to solve the above-mentioned technical problems. Summary of the Invention

[0006] To address the aforementioned technical issues, this application provides a health status scheduling method for energy storage systems based on dynamic equivalent modeling. This method can comprehensively assess the economic benefits of energy storage systems and reduce the rapid aging of batteries caused by excessive focus on returns.

[0007] The technical solution provided in this application is described below: This application provides a health state scheduling method for energy storage systems based on dynamic equivalent modeling, including... Obtain load forecast data for the power grid and power forecast data for renewable energy within a future predetermined time window; Based on the load forecast data and the power forecast data, calculate the original grid-connected power forecast data for each cycle within the future predetermined time window; Based on the original grid-connected power prediction data and the preset grid safety operation rules, multiple candidate operation schemes for the energy storage system are generated. The candidate operation schemes include different charging and discharging power command sequences that satisfy the grid power balance conditions. According to the preset grid service revenue pricing rules, calculate the grid service revenue that each candidate operation scheme can obtain within the future predetermined time window; Acquire real-time status parameters of the energy storage system, wherein the real-time status parameters include at least state of charge, health status, and temperature; The real-time status parameters are input into a pre-established dynamic equivalent aging model along with each of the candidate operating schemes, and the battery aging cost corresponding to each candidate operating scheme is output. The net revenue of each candidate operation scheme is calculated based on the grid service revenue and battery aging cost, and the candidate operation scheme with the highest net revenue is determined as the target operation scheme. Control commands are generated and sent to the energy storage system to control the energy storage system to operate according to the target operation plan.

[0008] Optionally, the real-time state parameters are input into a pre-established dynamic equivalent aging model along with each of the candidate operating schemes, and the battery aging cost corresponding to each candidate operating scheme is output, including: Based on the candidate operation scheme and the real-time status parameters, the cyclic capacity decay and calendar capacity decay of the energy storage system are calculated respectively. The cyclic capacity decay is caused by the cyclic charging and discharging of the energy storage system, and the calendar capacity decay is caused by the energy storage system over time and in a static state. By aggregating the cyclic capacity decay and the calendar capacity decay, the predicted total capacity decay is obtained; The battery aging cost is calculated based on the predicted total capacity decay and the initial investment cost of the battery in the energy storage system.

[0009] Optionally, the calculation of cycle capacity decay includes: The charge and discharge power command sequence is analyzed to identify and extract multiple complete charge and discharge stress cycles of the energy storage system within the predetermined future time window; For each complete charge-discharge stress cycle, the percentage of capacity decay caused by the complete charge-discharge stress cycle is obtained by referring to a preset cycle life curve table based on the discharge depth, average rate and average temperature of the complete charge-discharge stress cycle.

[0010] Optionally, the calculation of calendar capacity decay includes: The future predetermined time window is divided into multiple time steps; For each time step, based on the average state of charge and average temperature within the time step, the percentage of capacity decay within the time step is calculated using a preset static decay quantization model.

[0011] Optionally, the formula used to calculate the battery aging cost is: The battery aging cost = (initial total investment cost / total available capacity degradation) × predicted total capacity degradation.

[0012] Optionally, obtaining power prediction data for renewable energy includes: Obtain weather forecast data within the predetermined future time window, wherein the weather forecast data includes at least wind speed, wind direction data, solar irradiance, and cloud cover data; The weather forecast data is input into a pre-established renewable energy prediction model to obtain the power prediction data of the renewable energy source. The renewable energy prediction model includes a wind power prediction model and / or a photovoltaic power prediction model.

[0013] Optionally, the multiple candidate operating schemes for the generated energy storage system include: The first candidate operating scheme is generated based on the strategy of smoothing grid-connected power fluctuations; A second candidate operating scheme is generated based on a strategy that tracks the output curve; A third candidate operating scheme is generated based on a strategy that maximizes grid service revenue.

[0014] Optionally, the power grid safety operation rules include operation constraint rules based on real-time health status, wherein: The maximum allowable charge and discharge power limit of the energy storage system is dynamically adjusted based on the real-time health status. And / or, dynamically scale the available state of charge range of the energy storage system based on the real-time health status.

[0015] Optionally, controlling the energy storage system to operate according to the target operating scheme includes: Real-time monitoring of the operating data of the energy storage system, the operating data including at least voltage and temperature; The operational data is compared with the high-stress aging interval threshold defined by the dynamic equivalent aging model; If the current operating state is about to enter or has already entered the high-stress aging range, the current target operating scheme will be immediately interrupted, and a backup safe operating scheme aimed at minimizing battery stress will be switched.

[0016] Optionally, the step of calculating the net revenue of each candidate operating scheme based on the grid service revenue and battery aging cost, and determining the candidate operating scheme with the highest net revenue as the target operating scheme, includes: Based on the real-time health status of the energy storage system, a health status adaptive weighting factor is dynamically determined, wherein the lower the value of the real-time health status, the higher the value of the health status adaptive weighting factor. Based on the health status adaptive weighting factor, the battery aging cost of each candidate operating scheme is weighted and calculated to obtain the weighted aging cost; The net revenue of each candidate operation scheme is calculated by subtracting its corresponding weighted aging cost from the power grid service revenue of each candidate operation scheme. By comparing the net benefits of all candidate operating schemes, the candidate operating scheme with the highest net benefit is determined as the target operating scheme.

[0017] As can be seen from the above technical solutions, this application has the following beneficial effects: In this embodiment, load forecast data of the power grid and power forecast data of renewable energy within a predetermined future time window are obtained. Based on these two data sets, the original grid-connected power forecast data for each cycle within the predetermined future time window is calculated. Then, based on this original grid-connected power forecast data and power grid safety operation rules, multiple candidate operation schemes are generated. Next, the power grid service revenue for each candidate operation scheme is calculated according to the power grid service revenue pricing rules. Then, the real-time state parameters of the energy storage system are obtained, and these real-time state parameters, along with each candidate operation scheme, are input into a pre-established dynamic equivalent aging model. The battery aging cost corresponding to each candidate operation scheme is output. Finally, the power grid service revenue and battery aging cost are calculated for each... The net benefit of candidate operating schemes is calculated, and the candidate operating scheme with the highest net benefit is selected as the target operating scheme. Finally, control commands are generated and sent to the energy storage system to control the energy storage system to operate according to the target operating scheme. By introducing a dynamic equivalent aging model, the cumulative damage to battery health caused by different charge and discharge strategies (such as high rate and deep discharge) can be accurately quantified and converted into intuitive battery aging costs. This can avoid harmful operating schemes that accelerate battery aging, significantly extend the service life of the energy storage system, and effectively protect battery life. It solves the defect of traditional scheduling that ignores the impact of aging. It can also comprehensively evaluate the economic benefits of the energy storage system and reduce the situation where rapid battery aging is caused by excessive focus on benefits. Attached Figure Description

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

[0019] Figure 1 This is a schematic diagram of an embodiment of the energy storage system health state scheduling method based on dynamic equivalent modeling according to this application; Figure 2 A schematic diagram illustrating an embodiment of the battery aging cost corresponding to each candidate operating scheme for this application; Figure 3 This is a schematic diagram of one embodiment of calculating cycle capacity decay in this application; Figure 4 A schematic diagram illustrating one embodiment of calculating calendar capacity decay in this application; Figure 5 This is a schematic diagram of one embodiment of the energy storage system controlled by this application to operate according to a target operating scheme; Figure 6 This is a schematic diagram of an embodiment of determining the candidate operating scheme with the highest net benefit as the target operating scheme for this application. Detailed Implementation

[0020] To address the problem that traditional scheduling methods only focus on real-time SOC, power capacity, and energy gains, leading to accelerated battery aging and poor overall returns of the energy storage system, this application proposes a health state scheduling method for energy storage systems based on dynamic equivalent modeling. This method can comprehensively evaluate the economic benefits of the energy storage system and reduce the situation where excessive focus leads to rapid battery aging.

[0021] It should be noted that the energy storage system health state scheduling method based on dynamic equivalent modeling provided in this application can be applied to terminals, systems, and servers. For ease of explanation, this application uses a system as the execution subject for illustration.

[0022] Please see Figure 1 This application first provides an embodiment of a health state scheduling method for energy storage systems based on dynamic equivalent modeling, the embodiment including: 101. Obtain load forecast data for the power grid and power forecast data for renewable energy within a future predetermined time window.

[0023] The future scheduled time window can be from several hours to 24 hours in the future, for example, 24 hours in the future, with a time interval of 15 minutes or 1 hour.

[0024] Load forecast data refers to the total power demand forecast of all users (industrial, commercial, and residential) in the power grid. This load forecast data is obtained from the power grid dispatch center. In this embodiment, the load forecast data for a predetermined future time window is predicted, such as the total power demand forecast for the next 24 hours. Load forecast data can be based on historical electricity consumption patterns, weather, weekdays / holidays, and other factors.

[0025] In this embodiment, renewable energy includes wind power and photovoltaic power generation. Renewable energy power forecast data refers to the predicted output power of wind farms and photovoltaic power plants. The power generation of renewable energy directly depends on meteorological conditions and is generated by combining weather forecast data within a predetermined future time window. For wind farms, weather forecast data includes wind speed, wind direction, air density, temperature, etc. For photovoltaic power plants, weather forecast data includes solar irradiance (total irradiance, direct irradiance, diffuse irradiance), cloud cover (cloud type, coverage, and movement speed), temperature, humidity, aerosol concentration (affecting atmospheric transmittance), etc.

[0026] Power forecast data for renewable energy can be obtained from data released by the power grid dispatch center, data provided by third-party commercial power forecasting services, or calculated using statistical or machine learning models based on weather forecast data and historical power data.

[0027] In an optional embodiment, the specific steps for obtaining power forecast data for renewable energy include: Obtain weather forecast data for a future scheduled time window. The weather forecast data should include at least wind speed, wind direction, solar irradiance, and cloud cover.

[0028] Weather forecast data is input into a pre-established renewable energy prediction model to obtain renewable energy power prediction data. The renewable energy prediction model includes a wind power prediction model and / or a photovoltaic power prediction model.

[0029] Specifically, the wind power prediction model is based on the power curve characteristics of wind turbine generators and considers factors such as wake effects and array losses in wind farms, mapping the input wind speed and direction data sequences to the total output active power sequence of the wind farm. The photovoltaic power prediction model is based on the photoelectric conversion principle of photovoltaic modules and considers factors such as the installation tilt angle, azimuth angle, efficiency degradation, shading losses, and inverter efficiency of photovoltaic modules, mapping the input irradiance and other data sequences to the total output active power sequence of the photovoltaic power station.

[0030] If the renewable energy sources connected to the grid include both wind power and photovoltaic power, then wind power prediction sequences and photovoltaic power prediction sequences are obtained separately. The wind power prediction sequences and photovoltaic power prediction sequences are then algebraically added together at time points to obtain the power prediction data of renewable energy.

[0031] 102. Calculate the original grid-connected power forecast data for each cycle within the future predetermined time window based on the load forecast data and power forecast data.

[0032] Raw grid-connected power represents the net power required by traditional controllable power sources (such as thermal power and nuclear power) in the power grid without considering active intervention from energy storage systems. The formula for calculating raw grid-connected power is: Raw grid-connected power equals total load forecast data minus renewable energy power forecast data. Based on this, the power output required by traditional controllable power sources (such as thermal power and nuclear power) to meet grid load demand can be calculated without the intervention of energy storage systems.

[0033] For example, in a 15-minute period, the total load is predicted to be 1000MW, and the renewable energy forecast is 350MW (200MW wind power and 150MW photovoltaic). From this, it can be calculated that the power required by traditional controllable power sources (such as thermal power and nuclear power) is 650MW (1000MW-350MW). 650MW is the power generation plan target that the power grid dispatch center needs to issue to the power plants (thermal power, nuclear power, etc.). Assuming that the power plants can provide a stable power of 400MW, in order to achieve the target of 650MW, the power plant units need to adjust their output (increase / decrease the generator unit power).

[0034] However, due to the variable nature of weather, the power output of wind and solar power fluctuates. Therefore, the power grid dispatch center needs to issue power generation targets of varying power outputs to power plants. But because power plants have slow ramp-up speeds and cannot respond quickly, this leads to fluctuations in grid power. Therefore, energy storage systems are needed to balance these grid power fluctuations.

[0035] When actual renewable energy output is lower than the power forecast, causing a sudden increase in the original grid-connected power demand, the energy storage system discharges to make up for the shortfall, so that power plants do not need to drastically increase their output. When actual renewable energy output is higher than the power forecast, causing a sudden decrease in the original grid-connected power demand, the energy storage system charges to absorb the excess power, so that power plants do not need to drastically reduce their output.

[0036] A future predetermined time window represents the total length of a future period that needs to be predicted. A single time unit is formed by evenly dividing the future predetermined time window; each time unit constitutes a cycle, and each cycle is the smallest time granularity for scheduling decisions. Within each cycle, the state of the power grid (load, renewable energy output) and the operation of the energy storage system (charging and discharging power) remain unchanged.

[0037] 103. Based on the original grid-connected power prediction data and the preset grid safety operation rules, generate multiple candidate operation schemes for the energy storage system. The candidate operation schemes include different charging and discharging power command sequences that meet the grid power balance conditions.

[0038] The power grid power balance condition is the operating rule of the power system, which requires that the total power output of all power generation units (including thermal power, renewable energy, and energy storage system discharge) be equal to the total power consumed by all loads (user consumption) and grid losses. However, in reality, due to the volatility of renewable energy, there are fluctuations between the total power and the total load, and this fluctuation needs to be reduced as much as possible.

[0039] The charge / discharge power command sequence is a set of commands arranged in chronological order. It defines the actions that the energy storage system should perform at each specific time period (e.g., every 15 minutes) within a complete future time window (e.g., the next 24 hours), such as how much power to charge, how much power to discharge, or to idle.

[0040] Each candidate operating scheme is a complete sequence of charging and discharging power commands that covers the entire future predetermined time window. These schemes have different strategies, provided that they meet the rules of grid balance and safety.

[0041] For example, given a power fluctuation curve for the next 24 hours, the system generates the following three strategy options: Option A: When there is a significant power deficit (e.g., during peak electricity prices), the batteries in the energy storage system discharge at a higher rate (e.g., 0.5C); when there is a power surplus (e.g., during off-peak electricity prices), the batteries in the energy storage system charge rapidly. This option provides strong support to the power grid and may yield high returns, but it results in significant battery losses.

[0042] Option B: The charge / discharge rate of the battery in the energy storage system is always controlled below 0.3C, allowing for smooth power regulation. This option is battery-friendly but may miss some high-yield opportunities.

[0043] Option C: Maintain the battery's state of charge at a high level at specific times (such as when photovoltaic power generation stops in the evening) to prepare for use during the evening peak. That is, charge the battery during the daytime when there is excess photovoltaic power to store energy, and discharge it during the evening peak when photovoltaic power is absent and the load is high.

[0044] It should be noted that when generating candidate operation schemes, the power change rate (ramp rate) is constrained by the power grid safety operation rules to prevent drastic changes in energy storage power from becoming a new source of disturbance.

[0045] In another optional embodiment, when generating multiple candidate operating schemes for the energy storage system, a first candidate operating scheme is generated based on a strategy to smooth grid-connected power fluctuations. In this scheme, the goal is to make the total grid-connected power after combining renewable energy sources such as wind and solar power smooth and stable. When the power fluctuates drastically, the energy storage system is charged (absorbing excess power) and discharged (supplementing the power deficit) to output a power curve with a gradual change, thereby improving the power quality of the grid and mitigating the impact of fluctuations.

[0046] The strategy of generating a second candidate operation plan based on tracking the output curve aims to strictly follow the power generation plan curve issued by the grid dispatching agency by combining the output of renewable energy and energy storage systems. The plan curve is compared with the renewable energy forecast data, the difference is calculated, and the energy storage system is instructed to accurately make up for the difference, thereby fulfilling the contractual obligations with the grid and ensuring the rigid execution of the grid dispatching plan.

[0047] A third candidate operating scheme is generated based on a strategy to maximize grid service revenue. The goal of this scheme is to maximize the grid service revenue of the energy storage system by participating in the electricity market. Based on the electricity price signal, the system charges during off-peak hours and discharges during peak hours to earn the maximum peak-valley price difference, thereby improving the economic efficiency of the energy storage system operation and achieving the highest profit.

[0048] In addition, when generating multiple candidate operation schemes, constraints are first defined, including power balance constraints (the charging and discharging power of the energy storage system matches the power regulation requirements of the grid), SOC path constraints (the state of charge of the energy storage system is always between its minimum and maximum allowable values ​​throughout the entire scheduling cycle), and power limit constraints (the charging and discharging power of the energy storage system in each cycle cannot exceed its current maximum allowable power).

[0049] Then, with the optimization objective of minimizing the sum of the absolute values ​​of the changes in grid-connected power between adjacent cycles, and under the condition of satisfying power balance constraints, a smooth charging and discharging power command sequence (first candidate operation scheme) is obtained. With the optimization objective of minimizing the sum of the absolute values ​​of the deviations between the actual grid-connected power and the planned power, and under the condition of satisfying SOC path constraints, a charging and discharging power command sequence that can accurately track the plan is obtained (second candidate operation scheme). With the optimization objective of maximizing grid service revenue (discharging revenue - charging cost), where revenue and cost are determined by time-of-use pricing, and under the condition of satisfying power limit constraints, a charging and discharging power command sequence that maximizes economic benefits is obtained (third candidate operation scheme).

[0050] In another optional embodiment, the power grid safety operation rules include operation constraint rules based on real-time health status, wherein: The maximum allowable charging and discharging power limit of the energy storage system is dynamically adjusted based on the real-time health status. And / or, dynamically scale the available state of charge range of the energy storage system based on real-time health status.

[0051] Dynamically adjusting the power limit means that when the battery is new, it is allowed to charge and discharge at a higher power. As the battery ages, the system will actively reduce the maximum charge and discharge power, thereby effectively reducing internal battery losses and slowing down its aging process.

[0052] Dynamically adjusting the power usage range means that when the battery is new, almost all of its power can be used (e.g., from 10% to 90%). As the battery ages, the system narrows the allowed power usage range (e.g., only from 30% to 70%), significantly extending its remaining lifespan by allowing the battery to operate in a more comfortable middle power range.

[0053] 104. Based on the preset grid service revenue pricing rules, calculate the grid service revenue that each candidate operation scheme can obtain within a future predetermined time window.

[0054] Grid service revenue refers to the direct economic return that an energy storage system can obtain in the electricity market or under a contract with the grid by executing a specific sequence of charging and discharging commands. The pricing rules for grid service revenue are a set of standardized formulas that convert the physical services (such as discharging, charging, and power regulation) provided by the energy storage system into monetary value.

[0055] The pricing rules for grid service revenue define time-of-use electricity prices for different time periods. During calculation, the discharge revenue is obtained by multiplying the discharge volume for each cycle by the electricity sales price for that cycle; the charging cost is obtained by multiplying the charging volume for each cycle by the electricity purchase price for that cycle.

[0056] The grid service revenue for each candidate operation scheme = the total discharge revenue of the energy storage battery in the candidate operation scheme - the total charging cost of the energy storage battery in the candidate operation scheme.

[0057] For example, the peak-valley electricity pricing rules are: peak electricity price is 1.0 yuan / kWh, and off-peak electricity price is 0.3 yuan / kWh.

[0058] For Scheme A in step 103 above, it discharges 100 MWh during a peak period and charges 120 MWh during a previous off-peak period (considering charging losses). Its grid service revenue during this time window = (100 MWh × 1.0 yuan / kWh) - (120 MWh × 0.3 yuan / kWh) = 100,000 yuan - 36,000 yuan = 64,000 yuan.

[0059] The calculation of power grid service revenue for other candidate operation schemes is similar to that of Scheme A, and will not be repeated here.

[0060] 105. Obtain the real-time status parameters of the energy storage system. The real-time status parameters shall include at least the state of charge, health status, and temperature.

[0061] State of charge (SOC) indicates the percentage of remaining charge in a battery. The current SOC determines how much charge the battery can still discharge or recharge.

[0062] State of Health (SOH) indicates the current health status of a battery, usually expressed as a percentage of its rated capacity relative to its current maximum capacity. A battery with an SOH of 80% will have completely different charge / discharge capabilities and aging rates compared to a new battery.

[0063] Temperature indicates the current operating temperature of a battery pack or cell, and temperature affects battery performance, safety, and aging rate.

[0064] The real-time status of the energy storage system is obtained from its internal battery management system.

[0065] 106. Input the real-time status parameters and each candidate operating scheme into the pre-established dynamic equivalent aging model, and output the battery aging cost corresponding to each candidate operating scheme.

[0066] The dynamic equivalent aging model is a mathematical model that can simulate the lifespan degradation process of a battery under specific operating stresses. It can transform the abstract concept of "loss" into a quantifiable "aging index" (such as the percentage of capacity degradation).

[0067] Battery aging cost refers to the immediate economic cost of battery life degradation caused by implementing a specific operating program, calculated based on its total lifespan value.

[0068] Specifically, the real-time state parameters of the energy storage system (including the current state of charge, health status, and temperature) and the complete charge / discharge power command sequence defined by the candidate operating scheme (this sequence implicitly contains the operating stresses the battery will experience in each future operating cycle, such as key stress factors like charge / discharge rate and cycle depth) are input into the dynamic equivalent aging model. Based on its built-in electrochemical mechanisms, empirical degradation formulas, or data-driven algorithms, the model simulates the battery's state evolution under specified operating stresses on a time-by-time basis. After the simulation is complete, the model outputs one or more quantifiable physical aging indicators, such as the percentage of battery capacity degradation over the entire prediction time window.

[0069] In the above empirical degradation formula, the battery capacity degradation rate is first correlated with multiple stress factors such as charge / discharge rate, ambient temperature, state of charge range, and cumulative throughput. Then, the following empirical formula is used for quantitative calculation: .

[0070] In the formula, Q represents the percentage of capacity decay; B is the pre-factor, obtained by fitting battery cycle life test data; Ea is the activation energy, used to characterize the effect of temperature T (in Kelvin) on the aging rate; R is the ideal gas constant; Ah is the cumulative charge-discharge ampere-hours under the current stress conditions; and Z is the power law exponent, determined by fitting experimental data. Finally, the power sequence and real-time temperature from the candidate operating scheme are received, the entire time window is discretized into multiple hourly segments, and the decay amount within each segment is calculated based on the current rate, state of charge (SOC), and temperature. These are then accumulated into the total decay amount, and the predicted capacity decay value after the entire scheme is executed is finally output.

[0071] Then, a depreciation method based on the battery's full life-cycle value is adopted, where the initial total asset value of the battery system and the threshold for determining the end of its lifespan are known. The aging cost is calculated using the following formula: Battery aging cost = (Initial total investment cost / Total usable capacity degradation) × Predicted total capacity degradation.

[0072] For example, if we input the aforementioned schemes A and B into a dynamic equivalent aging model, the model will output that the total capacity degradation caused by implementing scheme A is 0.012%, while the total capacity degradation caused by scheme B is 0.003%. Assuming the total value of the battery pack is 5 million yuan, and the total degradeable capacity is 20% of the rated capacity (i.e., SOH from 100% to 80%), then the cost of each 1% capacity degradation is 250,000 yuan. Therefore, the aging cost of scheme A = 250,000 yuan / % × 0.012% = 300 yuan. The aging cost of scheme B = 250,000 yuan / % × 0.003% = 75 yuan.

[0073] Each candidate operating scheme will receive a corresponding estimated battery aging cost, expressed in monetary terms, which will serve as the basis for subsequent economic decisions.

[0074] Specifically, the implementation of the dynamic equivalent aging model is described below: First, identify the key stress factors affecting battery cycle aging, including at least depth of discharge, average charge / discharge rate, and average temperature. For battery samples of the same type used in the energy storage system, conduct routine accelerated cycle aging tests under different combinations of depth of discharge, rate, and temperature. Record the percentage of capacity decay caused by each complete charge / discharge cycle under various stress combinations, and organize the recorded data into a cycle life curve table.

[0075] Secondly, conventional high-temperature static aging experiments were conducted on the battery samples, which involved placing the batteries under multiple different constant high-temperature and constant SOC conditions. Capacity decay was measured periodically, and the decay data at different temperatures were fitted using the formula k=A·exp(-Ea / (R·T)) to determine the activation energy Ea. In the formula, k is the aging reaction rate, A is the pre-factor, R is the ideal gas constant (approximately 8.314 J / (mol·K), and T is the temperature. For most lithium-ion batteries, the typical value of Ea ranges from 30,000 J / mol to 60,000 J / mol. The static aging rate of the batteries under different initial SOCs was analyzed to determine the specific form and parameters of the SOC stress function f(SOC). The stress function f(SOC) is an exponential function or a piecewise linear function of SOC.

[0076] Finally, the constructed cycle life curve table and the quantified static decay model with calibrated parameters are integrated to form a complete dynamic equivalent aging model. In application, the real-time battery state parameters and candidate operating schemes are input into the model. The model queries the curve table and calculates static decay, outputting the predicted total capacity decay. Based on the formula Battery Aging Cost = (Initial Total Investment Cost / Total Usable Capacity Decay) × Predicted Total Capacity Decay, the decay amount is converted into economic cost.

[0077] 107. Calculate the net revenue of each candidate operation scheme based on the grid service revenue and battery aging cost, and determine the candidate operation scheme with the highest net revenue as the target operation scheme.

[0078] Net income refers to the net profit obtained after deducting battery aging costs from grid service revenue. The formula for calculating net income is: Net Income = Grid Service Revenue - Battery Aging Costs.

[0079] By calculating the net profit of each candidate running scheme and then comparing the net profits of each candidate running scheme, the candidate running scheme with the highest net profit is determined as the target running scheme.

[0080] For example, Option A: Revenue of 64,000 yuan - Aging cost of 300 yuan = Net revenue of 63,700 yuan.

[0081] Option B: Revenue of 50,000 yuan - Aging cost of 75 yuan = Net revenue of 49,925 yuan.

[0082] Option D: The profit is 65,000 yuan (higher than Option A), but its crude operating strategy results in aging costs as high as 2,000 yuan. Therefore, its net profit = 65,000 - 2,000 = 63,000 yuan. At this point, the system will reject Option D because its net profit is lower than Option A's 63,700 yuan.

[0083] Although Option A causes greater battery wear, its net benefit, after deducting the cost of wear, is still far higher than that of Option B. Therefore, the system will select Option A as the target operating option.

[0084] 108. Generate control commands and send them to the energy storage system to control the energy storage system to operate according to the target operating plan.

[0085] In this embodiment, the system parses the target operating plan into a set of specific instructions containing timestamps and power values. For example: at 12:00, discharge continuously at 50MW for 60 minutes; at 13:00, stop operation; at 14:00, charge at 30MW for 90 minutes... This instruction is sent to the energy storage system, which further decomposes the instruction and sends it to each PCS (converter). The PCS precisely controls the charging and discharging of the battery, thereby achieving power regulation of the power grid.

[0086] Specifically, the system divides the future scheduled time window into multiple rolling optimization cycles (e.g., optimization is performed every hour). At the beginning of each rolling optimization cycle, the system performs the following steps: Collect actual load data and actual renewable energy output data from the previous cycle to the current moment.

[0087] The predicted data from the previous period is compared with the actual data to calculate the prediction error.

[0088] The prediction error is used as feedback and input into the load forecasting model and the renewable energy forecasting model to dynamically correct the forecast data for the current cycle and subsequent cycles. For example, an error autoregressive model can be used to correct the prediction curve in real time.

[0089] Based on the corrected prediction data, steps 102 to 108 are re-executed to generate a new target operation plan, but only the instructions for the next rolling optimization cycle (e.g., 1 hour) are executed.

[0090] Therefore, by using rolling optimization, the robustness of the system can be improved to cope with prediction errors.

[0091] In this embodiment, load forecast data of the power grid and power forecast data of renewable energy within a predetermined future time window are obtained. Based on these two data, the original grid-connected power forecast data for each cycle within the predetermined future time window is calculated. Then, based on the original grid-connected power forecast data and the power grid safety operation rules, multiple candidate operation schemes are generated. Next, the power grid service revenue of each candidate operation scheme is calculated according to the power grid service revenue pricing rules. Then, the real-time state parameters of the energy storage system are obtained, and these real-time state parameters and each candidate operation scheme are input into a pre-established dynamic equivalent aging model. The battery aging cost corresponding to each candidate operation scheme is output. Then, the net revenue of each candidate operation scheme is calculated based on the power grid service revenue and the battery aging cost. The candidate operation scheme with the highest net revenue is taken as the target operation scheme. Finally, a control command is generated and sent to the energy storage system to control the energy storage system to operate according to the target operation scheme. Thus, by introducing the dynamic equivalent aging model, the cumulative damage to battery health caused by different charging and discharging strategies (such as high rate and deep discharge) can be accurately quantified and transformed into an intuitive battery aging cost. This enables scheduling decisions to proactively avoid harmful operating schemes that accelerate battery aging, significantly extending the lifespan of the energy storage system, effectively protecting battery life, and solving the shortcomings of traditional scheduling that ignores the effects of aging.

[0092] By avoiding high-stress operating schemes, the rate of battery capacity degradation can be effectively slowed down, the battery's lifespan can be significantly extended, thereby reducing the frequency of battery replacement and lowering the system's total lifespan operating costs.

[0093] Please see Figure 2 , Figure 2 An example of outputting the battery aging cost corresponding to each candidate operating scheme for this application, the example includes: 201. Based on the candidate operation scheme and real-time status parameters, calculate the cyclic capacity decay and calendar capacity decay of the energy storage system. The cyclic capacity decay is caused by the cyclic charging and discharging of the energy storage system, while the calendar capacity decay is caused by the energy storage system over time and under resting conditions.

[0094] Cyclic capacity decay refers to the irreversible capacity loss of a battery during active charge-discharge cycles.

[0095] Calendar capacity decay refers to the aging of a battery when it is in a static state (not charged or discharged) due to the passage of time and the influence of the environment (mainly temperature and state of charge).

[0096] The cycle capacity decay and calendar capacity decay in this embodiment will be described in detail in subsequent embodiments.

[0097] 202. Combine the cycle capacity decay and calendar capacity decay to obtain the predicted total capacity decay.

[0098] In this embodiment, the degradation caused by two different mechanisms, cycle capacity degradation and calendar capacity degradation, is combined to obtain the percentage decrease in battery health expected over the entire future time window after executing the candidate operating scheme. For example, the predicted total capacity degradation is 0.05%.

[0099] Specifically, the cycle capacity decay calculated in step 201 (denoted as Q) 循环 ) and calendar capacity decay (denoted as Q) 日历 The two attenuation mechanisms are algebraically added together, and the linear superposition value is used as the predicted total capacity attenuation (denoted as Q). 预测 ).

[0100] The formula used to calculate the predicted total capacity decay is: Q 预测 =Q 循环 +Q 日历 .

[0101] Among them: Q 循环 With Q 日历 All figures are expressed as a percentage of the battery's rated capacity.

[0102] 203. Calculate the battery aging cost based on the predicted total capacity decay and the initial investment cost of the battery in the energy storage system.

[0103] The initial total investment cost is the total capital expenditure for purchasing and installing the energy storage system; The formula used in this embodiment to calculate the battery aging cost is: Battery aging cost = (Initial total investment cost / Total usable capacity degradation) × Predicted total capacity degradation.

[0104] The total usable capacity degradation value refers to the total capacity degradation experienced by the battery from its initial healthy state to its end-of-life state. For example, if the state of harmlessness (SOH) of a brand new battery is 100%, and the end of its life is defined as when the SOH drops to 80%, then the total degradeable capacity = rated capacity × (100% - 80%).

[0105] For example, the rated capacity of the energy storage system is 100MWh, the initial investment cost is 50 million yuan, and the state of equilibrium (SOH) at the end of its life is 80%. Therefore, the total decayable capacity = 100MWh × (100% - 80%) = 20MWh, meaning that the cost of 50 million yuan corresponds to a 20MWh consumable lifetime. Thus, the cost per 1MWh of decayed capacity = 50 million yuan / 20MWh = 2.5 million yuan / MWh.

[0106] If the predicted total capacity degradation is 0.01 MWh after implementing a certain candidate operating scheme, then the battery aging cost of this candidate operating scheme = 0.01 MWh × 2.5 million yuan / MWh = 25,000 yuan.

[0107] Please see Figure 3 , Figure 3 One embodiment for calculating cycle capacity decay in this application includes: 301. Analyze the charge and discharge power command sequence to identify and extract multiple complete charge and discharge stress cycles of the energy storage system within a predetermined time window in the future.

[0108] The charging and discharging power command sequence of the candidate operating scheme is a dataset arranged in chronological order. Each data point in the dataset contains a timestamp and the corresponding power value (positive value for discharging, negative value for charging).

[0109] In this embodiment, firstly, the charge / discharge power command sequence from the candidate operating schemes is obtained, and based on this sequence, the predicted change curve of the state of charge (SOC) of the energy storage system within a predetermined time window is calculated. Then, the predicted SOC change curve is traversed to identify all complete charge / discharge cycle intervals. Within a complete charge / discharge cycle interval, the cycle start point is a local high point of the SOC, and the cycle end point is when the SOC returns to a point near the start point after a drop (discharge) and subsequent rise (charge) from the local high point. Here, "near" is a preset tolerance; for example, a cycle is considered complete when the SOC returns to within ±2% of the initial value. Finally, for each identified complete charge / discharge cycle interval, the cycle stress parameters are calculated. These parameters include at least: the depth of discharge within the interval, the average charge / discharge rate within the interval, and the average operating temperature within the interval.

[0110] Depth of discharge (DOC) represents the difference between the initial state of charge (SOC) and the lowest SOC within a complete charge-discharge cycle. For example, if the initial SOC is 80% and the lowest SOC is 30%, then the DOC is 50%.

[0111] The average rate represents the percentage of the absolute power relative to the battery's rated capacity during the discharge and charge phases of the cycle, and then a representative value is taken over the entire cycle (such as the arithmetic mean of the discharge and charge rates).

[0112] Average temperature represents the average operating temperature of the energy storage system during the duration of the cycle.

[0113] For example, in one candidate operating scenario, the energy storage system's State of Charge (SOC) begins to decrease from 90% at 9:00 AM, discharges to 40% at 10:00 AM, then begins charging, reaching 88% by 11:30 AM. The system recognizes this as a complete stress cycle. Its parameters are: Depth of discharge = 90% - 40% = 50%; Average rate = (Average rate during discharge + Average rate during charging) / 2 = 0.4C; Average temperature = 25°C.

[0114] 302. For each complete charge-discharge stress cycle, the percentage of capacity decay caused by the complete charge-discharge stress cycle is obtained by referring to the preset cycle life curve table based on the discharge depth, average rate and average temperature of the complete charge-discharge stress cycle.

[0115] The pre-set cycle life curve table is a database that has been pre-established and stored in the system through a large number of battery cycle aging experiments. The database stores data on the percentage of capacity decay caused by a single complete cycle of the battery under different combinations of depth of discharge, average rate, and average temperature conditions.

[0116] For each complete charge-discharge cycle interval, the corresponding cycle stress parameter is used as a query condition to match and query a preset cycle life database to obtain the single-cycle capacity decay percentage caused by that cycle. Then, the single-cycle capacity decay percentages obtained from all identified cycles are summed to obtain the total cycle capacity decay percentage caused by cycle aging after executing the entire candidate operating scheme.

[0117] For example, in the entire 24-hour operation plan, the system identified 5 similar complete charge-discharge stress cycles, but their stress parameters were slightly different. The single-cycle attenuation values ​​obtained from the table were 0.005%, 0.007%, 0.004%, 0.006%, and 0.005%, respectively. Therefore, the total cycle capacity attenuation = 0.005% + 0.007% + 0.004% + 0.006% + 0.005% = 0.027%.

[0118] Please see Figure 4 , Figure 4 One embodiment for calculating calendar capacity decay in this application includes: 401. Divide the future scheduled time window into multiple time steps.

[0119] The system divides a predetermined future time window (e.g., the next 24 hours) into continuous, fixed time intervals, known as time steps. The duration of each time step is a preset system parameter, which can be selected as 1 hour, 30 minutes, or 15 minutes, etc.

[0120] 402. For each time step, based on the average state of charge and average temperature within the time step, calculate the percentage of capacity decay within the time step using a preset static decay quantization model.

[0121] For each time step, the system obtains the average state of charge and average temperature within that step from the candidate operating schemes and the battery state prediction (the predicted state of charge and predicted temperature for each time step within the future predetermined time window, derived from the candidate operating schemes).

[0122] The average state of charge (SOC) represents the arithmetic mean of the predicted SOC values ​​at the beginning and end of the time step. The average temperature represents the predicted average operating temperature of the energy storage system within the time step.

[0123] The preset static aging degradation quantification model is a mathematical formula based on battery static aging experimental data, which quantifies the rate of battery capacity degradation under specific SOC and temperature conditions per unit time.

[0124] The formula used in this static decay quantization model is as follows: .

[0125] in Δt is the percentage of capacity decay caused within this time step; A is the model preconditioner, determined by fitting experimental data; exp(-Ea / (R*T)) is the Arrhenius term, used to characterize the exponential effect of temperature on the chemical reaction rate, where Ea is the activation energy and R is the ideal gas constant; f(SOC) is a function of the average state of charge (SOC), used to characterize the difference in aging rate at different SOC levels, and is usually larger when SOC is extremely high; Δt is the duration of this time step.

[0126] For each time step, the acquired average temperature and average state of charge are substituted into the model above to calculate the calendar capacity decay ΔQ within that step. This process is repeated for all N time steps within the predetermined future time window, and the ΔQ calculated for each step is summed. The final sum is the predicted total calendar capacity decay caused by executing this candidate running scheme.

[0127] Please see Figure 5 , Figure 5 This application provides an embodiment of controlling the energy storage system to operate according to a target operating scheme, which includes: 501. Monitor the operating data of the energy storage system in real time. The operating data should include at least voltage and temperature.

[0128] In this embodiment, the system continuously collects the real-time voltage and temperature of representative battery cells or battery modules in the energy storage system at an extremely high frequency (e.g., several times per second) through the sensor network of the battery management system (BMS). This is to monitor the battery status in real time when executing the target operating scheme.

[0129] In actual systems, the monitored operating data also include current and internal resistance.

[0130] 502. Compare the running data with the high-stress aging interval threshold defined by the dynamic equivalent aging model.

[0131] The high-stress aging range threshold includes temperature and voltage thresholds. Specifically, when the battery temperature is >45°C, it is considered to have entered the high-stress range; when the external voltage of a single battery cell is >the upper limit (e.g., 3.65V) or <the lower limit (e.g., 2.5V), it is considered to have entered the high-stress range.

[0132] The high-stress aging range threshold can also include the SOC-power joint threshold, which is based on the battery's peak power capability curve. When the system's requested charge and discharge power causes the battery to operate in the region of SOC < 10% or SOC > 90%, it is considered to have entered the high-stress range because the internal resistance increases and heat generation intensifies at this time.

[0133] The high-stress aging range threshold is a warning line derived from the dynamic equivalent aging model and is directly related to the long-term health of the battery. This threshold is dynamic and can be adjusted according to factors such as the battery's current SOH and initial SOC. For example, when the battery's health status is 90%, the temperature threshold is 45°C; when the SOH drops to 85%, the temperature threshold will be adjusted to 42.5°C.

[0134] The operating data is compared with the high-stress aging range threshold to determine whether the operating data exceeds the threshold. If it does, it means that the battery's current operating state is about to enter or has already entered the high-stress aging range, and step 503 is executed at this time.

[0135] If the battery's operating data does not exceed the threshold, it means that the target operating plan can continue to be executed.

[0136] 503. If the current operating state is about to enter or has already entered the high-stress aging range, the current target operating scheme will be interrupted immediately, and a backup safe operating scheme aimed at minimizing battery stress will be switched.

[0137] The target operating plan is the economic benefit plan. If the plan causes the battery to enter a high stress range during operation, the economic benefit plan will be interrupted and a backup safe operating plan will be sought.

[0138] The backup safe operation scheme is one of many candidate operation schemes. When generating many candidate operation schemes, it is necessary not only to generate the target operation scheme with the highest net benefit, but also to generate the alternative safe operation scheme with the lowest battery stress. Although the alternative safe operation scheme has a lower net benefit, it ensures that the basic requirements of grid power balance can be met under any circumstances, while minimizing battery loss.

[0139] In the aforementioned process of determining the target operating scheme, the system will identify scheme a with the lowest battery stress and mark scheme a. When the current operating state of the battery in the target operating scheme is about to enter or has already entered the high-stress aging range, the system will directly replace the target operating scheme with scheme a.

[0140] Please see Figure 6 , Figure 6 One embodiment of determining the candidate operating scheme with the highest net benefit as the target operating scheme for this application includes: 601. Based on the real-time health status of the energy storage system, dynamically determine the adaptive weighting factor of the health status. The lower the value of the real-time health status, the higher the value of the adaptive weighting factor of the health status.

[0141] In this embodiment, the system obtains the current State of Health (SOH) value from the Battery Management System (BMS). SOH is usually expressed as a percentage; for example, 100% represents a brand new battery, and 80% is often defined as the end of its lifespan.

[0142] When the State of Harmony (SOH) is high (e.g., >95%), it indicates that the battery has a strong ability to withstand aging. In this case, the system can focus more on pursuing grid service benefits. Therefore, assigning a small weighting factor (e.g., close to 1) makes the impact of aging costs on decision-making relatively low.

[0143] As the State of Harmony (SOH) gradually decreases (e.g., to 90%, 85%), it indicates that the battery is more sensitive to stress, resulting in greater absolute capacity loss and higher residual value with each high-stress operation. Therefore, a gradually increasing weighting factor is assigned (e.g., 1.2, 1.5).

[0144] The weighting factor can be calculated using the following formula: Weighting factor = 1 + (1 - SOH) * K, where K is the amplification factor, for example, K is 5. When SOH = 100%, the weighting factor = 1; when SOH = 90%, the factor = 1.5.

[0145] 602. Based on the health status adaptive weighting factor, the battery aging cost of each candidate operation scheme is weighted and calculated to obtain the weighted aging cost.

[0146] Read the battery aging cost value for each candidate operating scheme calculated using the dynamic equivalent aging model; multiply the battery aging cost of each candidate operating scheme by the health status weight factor determined in the previous steps to obtain the weighted aging cost. Record the weighted aging cost value for each candidate operating scheme as input data for subsequent net profit calculation.

[0147] 603. Subtract the corresponding weighted aging cost from the power grid service revenue of each candidate operation scheme to calculate the net revenue of each candidate operation scheme.

[0148] Read the grid service revenue value for each candidate operation scheme calculated according to the grid service revenue pricing rules. For each candidate operation scheme, subtract the weighted aging cost value obtained in step 602 from its grid service revenue value to obtain the net revenue value of the scheme.

[0149] 604. Compare the net benefits of all candidate operating schemes, and determine the candidate operating scheme with the highest net benefit as the target operating scheme.

[0150] Compare and analyze the net profit values ​​of all candidate operating schemes obtained in step 603, and find the maximum value. The candidate operating scheme with the highest net profit value is determined as the target operating scheme. If multiple schemes have the same net profit, the scheme with the lower weighted aging cost is selected first.

[0151] Those skilled in the art will clearly understand that, for the sake of convenience and brevity, the specific working processes of the systems, devices, and units described above can be referred to the corresponding processes in the foregoing method embodiments, and will not be repeated here.

[0152] In the several embodiments provided in this application, it should be understood that the disclosed systems, apparatuses, and methods can be implemented in other ways. For example, the apparatus embodiments described above are merely illustrative; for instance, the division of units is only a logical functional division, and in actual implementation, there may be other division methods. For example, multiple units or components may be combined or integrated into another system, or some features may be ignored or not executed. Furthermore, the coupling or direct coupling or communication connection shown or discussed may be an indirect coupling or communication connection between apparatuses or units through some interfaces, and may be electrical, mechanical, or other forms.

[0153] The units described as separate components may or may not be physically separate. The components shown as units may or may not be physical units; that is, they may be located in one place or distributed across multiple network units. Some or all of the units can be selected to achieve the purpose of this embodiment according to actual needs.

[0154] Furthermore, the functional units in the various embodiments of this application can be integrated into one processing unit, or each unit can exist physically separately, or two or more units can be integrated into one unit. The integrated unit can be implemented in hardware or as a software functional unit.

[0155] If the integrated unit is implemented as a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of this application, in essence, or the part that contributes to the prior art, or all or part of the technical solution, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute all or part of the steps of the methods described in the various embodiments of this application. The aforementioned storage medium includes various media capable of storing program code, such as USB flash drives, portable hard drives, read-only memory (ROM), random access memory (RAM), magnetic disks, or optical disks.

Claims

1. A health state scheduling method for energy storage system based on dynamic equivalent modeling, characterized in that, The method comprises the following steps: obtaining load prediction data of a power grid and power prediction data of a renewable energy source within a future scheduled time window; calculating original grid-connected power prediction data of each period within the future scheduled time window according to the load prediction data and the power prediction data; generating a plurality of candidate operation schemes of an energy storage system according to the original grid-connected power prediction data and preset grid safety operation rules, wherein the candidate operation schemes include different charge and discharge power instruction sequences that meet the grid power balance condition; calculating grid service benefits that can be obtained by each candidate operation scheme within the future scheduled time window according to a preset grid service benefit pricing rule; obtaining real-time state parameters of the energy storage system, wherein the real-time state parameters at least include a state of charge, a state of health and a temperature; inputting the real-time state parameters and each candidate operation scheme into a pre-established dynamic equivalent aging model respectively, and outputting a battery aging cost corresponding to each candidate operation scheme; calculating a net benefit of each candidate operation scheme according to the grid service benefits and the battery aging cost, and determining a candidate operation scheme with the highest net benefit as a target operation scheme; generating a control instruction and delivering the control instruction to the energy storage system to control the energy storage system to operate according to the target operation scheme.

2. The energy storage system health scheduling method of claim 1, wherein, The method for calculating the battery aging cost corresponding to each candidate operation scheme comprises the following steps: calculating a cycle capacity attenuation and a calendar capacity attenuation of the energy storage system according to the candidate operation scheme and the real-time state parameters, wherein the cycle capacity attenuation is caused by the cycle charging and discharging of the energy storage system, and the calendar capacity attenuation is caused by the time and static state of the energy storage system; aggregating the cycle capacity attenuation and the calendar capacity attenuation to obtain a predicted total capacity attenuation; converting the battery aging cost according to the predicted total capacity attenuation and an initial investment cost of a battery in the energy storage system.

3. The energy storage system health scheduling method of claim 2, wherein, The method for calculating the cycle capacity attenuation comprises the following steps: analyzing the charge and discharge power instruction sequence, identifying and extracting a plurality of complete charge and discharge stress cycles of the energy storage system within the future scheduled time window; for each complete charge and discharge stress cycle, querying a preset cycle life curve table according to a discharge depth, an average rate and an average temperature of the complete charge and discharge stress cycle to obtain a capacity attenuation percentage caused by the complete charge and discharge stress cycle.

4. The energy storage system health scheduling method of claim 2, wherein, The method for calculating the calendar capacity attenuation comprises the following steps: dividing the future scheduled time window into a plurality of time steps; for each time step, calculating a capacity attenuation percentage in the time step through a preset static attenuation quantification model based on an average state of charge and an average temperature in the time step.

5. The energy storage system health scheduling method of claim 2, wherein, The formula used for converting the battery aging cost is: The battery aging cost = (the initial total investment cost / total available capacity attenuation value) x predicted total capacity attenuation.

6. The energy storage system health scheduling method of any one of claims 1-5, wherein, The method for obtaining the power prediction data of the renewable energy source comprises the following steps: obtaining weather forecast data in the future scheduled time window, the weather forecast data at least including wind speed, wind direction data, solar irradiance, cloud cover data; inputting the weather forecast data into a pre-established renewable energy prediction model to obtain power prediction data of the renewable energy, the renewable energy prediction model including a wind power prediction model and / or a photovoltaic power prediction model.

7. The energy storage system health scheduling method of any one of claims 1-5, wherein, The generating a plurality of candidate operation schemes of the energy storage system comprises: generating a first candidate operation scheme based on a strategy of smoothing grid-connected power fluctuation; generating a second candidate operation scheme based on a strategy of tracking output curve; generating a third candidate operation scheme based on a strategy of maximizing grid service benefit.

8. The energy storage system health scheduling method of any one of claims 1-5, wherein, The grid safe operation rule comprises an operation constraint rule based on real-time health state, wherein: dynamically adjusting the maximum allowed charge-discharge power limit of the energy storage system according to the real-time health state; and / or, dynamically scaling the available state of charge range of the energy storage system according to the real-time health state.

9. The energy storage system health scheduling method of any one of claims 1-5, wherein, The controlling the energy storage system to operate according to the target operation scheme comprises: real-time monitoring operation data of the energy storage system, the operation data at least including voltage and temperature; comparing the operation data with high stress aging interval threshold defined by the dynamic equivalent aging model; if the current operation state is about to enter or has entered the high stress aging interval, immediately triggering interrupting the current target operation scheme and switching to a backup safe operation scheme with the target of minimizing battery stress.

10. The energy storage system health scheduling method of any one of claims 1-5, wherein, The calculating net benefit of each candidate operation scheme according to the grid service benefit and battery aging cost and determining the candidate operation scheme with the highest net benefit as the target operation scheme comprises: dynamically determining a health state adaptive weight factor according to the real-time health state of the energy storage system, wherein the lower the value of the real-time health state, the greater the value of the health state adaptive weight factor; based on the health state adaptive weight factor, performing weighted calculation on the battery aging cost of each candidate operation scheme to obtain weighted aging cost; subtracting the weighted aging cost corresponding to each candidate operation scheme from the grid service benefit of the candidate operation scheme to calculate the net benefit of each candidate operation scheme; comparing the net benefits of all candidate operation schemes to determine the candidate operation scheme with the highest net benefit as the target operation scheme.

Images