A hybrid energy storage platform capacity configuration system in a microgrid

Abstract

The application discloses a micro-grid hybrid energy storage platform capacity configuration system, and belongs to the field of power system planning and design.The system comprises a power demand profiling module, which is used for decomposing the obtained net load power time sequence into a high-frequency power component and a low-frequency energy component based on a preset energy storage type demarcation frequency; a cost modeling module, which is used for determining the annual aging cost of a super capacitor based on the high-frequency power component, and determining the annual aging cost of a battery based on the low-frequency energy component; the annual aging cost of the battery comprises a stress additional cost related to the power change rate of the low-frequency energy component; and a collaborative optimization configuration module, which is used for combining the annual aging cost of the super capacitor and the annual aging cost of the battery determined by the cost modeling module, and a preset initial investment cost, so that the optimization algorithm can evaluate and measure the specific influence of different operation strategies on the long-term health of the battery.

Description

Technical Field

[0001] This invention relates to the field of power system planning and design, specifically to a capacity configuration system for a hybrid energy storage platform in a microgrid. Background Technology

[0002] The stable operation of microgrids is highly dependent on energy storage systems. Hybrid energy storage systems aim to combine the advantages of different energy storage technologies. However, existing capacity configuration methods typically treat this process as a static optimization problem. These methods meet peak demand by setting sufficient capacity redundancy, but ignore the nonlinear and asymmetric impact of load power dynamics on the lifespan of energy storage units. Specifically, energy storage units, such as lithium batteries, will suffer irreversible internal damage if frequently subjected to high-frequency, high-amplitude power surges outside their design range, leading to accelerated capacity and lifespan degradation. This phenomenon forms a vicious cycle of power surges, accelerated aging, and capacity degradation. Existing static configuration methods cannot identify and avoid this cross-timescale degradation coupling effect, resulting in the total lifespan cost of the configured system far exceeding expectations and decreased reliability.

[0003] The information disclosed in the background section above is only intended to enhance the understanding of the background of this disclosure, and therefore may include information that does not constitute prior art known to those skilled in the art. Summary of the Invention

[0004] The purpose of this invention is to provide a capacity configuration system for a hybrid energy storage platform in a microgrid, so as to solve the problems mentioned in the background art.

[0005] The technical solution of the present invention includes: a power demand segmentation module, used to decompose the acquired net load power time series into high-frequency power components and low-frequency energy components based on a preset energy storage type boundary frequency;

[0006] The cost modeling module is used to determine the annual aging cost of supercapacitors based on high-frequency power components and the annual aging cost of batteries based on low-frequency energy components. The annual aging cost of batteries includes stress-related additional costs associated with the power change rate of the low-frequency energy components.

[0007] The collaborative optimization configuration module is used to combine the annual aging cost of the supercapacitor and the battery determined by the cost modeling module, as well as the preset initial investment cost, to calculate the optimal capacity configuration of the supercapacitor and battery with the goal of minimizing the total cost of the system's entire life cycle.

[0008] The parameter feedback correction module is used to dynamically correct the power stress time constant in the cost modeling module based on the deviation between the actual health state and the predicted health state of the battery.

[0009] Preferably, the power demand profiling module is specifically used for:

[0010] Wavelet packet transform is used to process the net load power time series;

[0011] Based on the energy storage type boundary frequency, high-frequency power components and low-frequency energy components are separated from the net load power time series.

[0012] Preferably, the energy storage type boundary frequency is determined by physical experimental calibration of the target battery; the energy storage type boundary frequency is the power change frequency corresponding to the inflection point of nonlinear growth in the battery's life decay rate during cyclic charge-discharge experiments.

[0013] Preferably, the cost modeling module is used specifically to determine the annual aging cost of the battery:

[0014] The annual aging cost is determined by accumulating the instantaneous aging cost over an annual period.

[0015] The instantaneous aging cost is proportional to the absolute value of the low-frequency energy component, and the unit energy throughput cost is corrected by a dynamic stress factor that is related to the square of the power change rate of the low-frequency energy component.

[0016] Preferably, the power stress time constant in the dynamic stress factor is calibrated through physical experiments; the calibration methods include:

[0017] The target battery was subjected to baseline and experimental experiments. The baseline group used a smooth charge-discharge curve, while the experimental group used a charge-discharge curve with violent fluctuations.

[0018] Based on the cost difference between the two sets of experiments when the same degree of lifetime decay is achieved, the power stress time constant is solved inversely.

[0019] Preferably, the objective function of the collaborative optimization configuration module includes:

[0020] The initial investment cost is calculated based on the optimal capacity configuration to be determined and the preset unit capacity investment cost.

[0021] The total lifecycle operating cost is the sum of the present value of the annual aging costs for all future years, calculated based on a preset discount rate.

[0022] Preferably, the constraints of the collaborative optimization configuration module include:

[0023] The power balance constraint requires that the sum of the output power of the supercapacitor and the battery be no less than the instantaneous value of the net load power time series;

[0024] Physical constraints require that the operating power of supercapacitors and batteries does not exceed their respective rated power, and the rated power of batteries is related to the product of energy capacity and rated charge / discharge rate.

[0025] Preferably, the parameter feedback correction module is specifically used for:

[0026] The actual health status of the battery is collected, and the predicted health status is calculated based on the optimal capacity configuration output by the collaborative optimization configuration module and the actual operating power data.

[0027] Determine the deviation between actual health status and predicted health status;

[0028] A proportional-integral controller is used to update the power stress time constant using the deviation, which can be used for optimization or capacity expansion planning in the next cycle.

[0029] This invention provides an improved capacity configuration system for a hybrid energy storage platform in a microgrid, which has the following improvements and advantages compared to existing technologies:

[0030] 1. This invention achieves refined analysis and targeted allocation of energy storage demand. By employing wavelet packet transform and based on a physically calibrated energy storage type boundary frequency, it can accurately decompose the original net load power time series into high-frequency power components and low-frequency energy components. It can identify severe, high-frequency power fluctuations that significantly damage battery life and guide them to supercapacitors designed to handle such conditions, while allocating relatively gentle energy demand to the battery. This is fundamentally different from existing technologies that apply the entire power curve indiscriminately to the energy storage system, causing the battery to frequently experience power surges outside its design range.

[0031] 2. This invention constructs a nonlinear cost model capable of accurately quantifying the impact of dynamic power stress. The cost modeling module establishes a more precise annual battery aging cost model, which is not purely mathematical deduction but a quantitative model with clear physical meaning constructed based on a profound understanding of the electrochemical aging mechanism. The power stress time constant in the dynamic stress factor is calibrated through comparative experiments with specific benchmark and experimental groups, ensuring that the model parameters are directly linked to the physical characteristics of the target battery, possessing high feasibility and accuracy. A penalty term related to the square of the power change rate of the low-frequency energy component monetizes the additional loss to battery life caused by power shocks, forming a clear stress-related additional cost. This enables the optimization algorithm to assess and measure the specific impact of different operating strategies on the long-term health of the battery.

[0032] 3. This invention achieves global collaborative optimization with the goal of minimizing the total cost over the entire life cycle. The collaborative optimization configuration module combines the precisely quantified initial investment cost with the discounted value of the annual aging cost calculated based on the aforementioned nonlinear model, spanning all future years, to construct an optimization function with the sole objective of minimizing the total cost over the entire life cycle of the system. The decision-making process is no longer a simple addition of capacities, but rather a complex trade-off and optimization between the initial investment of the supercapacitor and the battery and their respective aging costs. The system may choose a scheme with a slightly higher initial investment, i.e., configuring a larger capacity supercapacitor, because the model clearly reveals that this can significantly reduce the stress-induced additional cost of the battery, thereby saving more replacement and maintenance costs throughout the entire life cycle and achieving globally optimal economic efficiency.

[0033] 4. This invention introduces an adaptive feedback correction mechanism for long-term operation. Its unique parameter feedback correction module periodically compares the actual and predicted health states of the battery, utilizing the deviation between the two to dynamically correct the core parameter in the cost modeling module—the power stress time constant—using a proportional-integral controller. This constructs a closed-loop feedback path from actual operation to model optimization, ensuring the long-term effectiveness of the system model. This enables the entire configuration system to have self-learning and evolution capabilities, adapting to changes in the operating environment and the uncertainties of equipment aging. It ensures that in future capacity expansion plans or operational strategy adjustments, the energy storage system always makes decisions based on the most realistic cost model, thus ensuring that the benefits of optimization are maintained throughout the system's service life. Attached Figure Description

[0034] The present invention will be further explained below with reference to the accompanying drawings and embodiments:

[0035] Figure 1 This is a flowchart of a capacity configuration system for a hybrid energy storage platform in a microgrid according to the present invention. Detailed Implementation

[0036] To make the objectives, technical solutions, and advantages of this invention clearer, the invention will be further described in detail below with reference to specific embodiments.

[0037] Example 1:

[0038] Please see Figure 1 The present invention provides a capacity configuration system for a hybrid energy storage platform in a microgrid, comprising: a power demand segmentation module, used to decompose the acquired net load power time series into high-frequency power components and low-frequency energy components based on a preset energy storage type boundary frequency;

[0039] The cost modeling module is used to determine the annual aging cost of supercapacitors based on high-frequency power components and the annual aging cost of batteries based on low-frequency energy components. The annual aging cost of batteries includes stress-related additional costs associated with the power change rate of the low-frequency energy components.

[0040] The collaborative optimization configuration module is used to combine the annual aging cost of the supercapacitor and the battery determined by the cost modeling module, as well as the preset initial investment cost, to calculate the optimal capacity configuration of the supercapacitor and battery with the goal of minimizing the total cost of the system's entire life cycle.

[0041] The parameter feedback correction module is used to dynamically correct the power stress time constant in the cost modeling module based on the deviation between the actual health state and the predicted health state of the battery.

[0042] This embodiment provides a hybrid energy storage platform capacity configuration system for microgrids. As a computer-aided design platform, it integrates a power demand decomposition module, a cost modeling module, a collaborative optimization configuration module, and a parameter feedback correction module to construct a dynamic optimization process with closed-loop adaptive capabilities. The system's operation logic begins with a deep analysis of the grid load, then accurately quantifies the nonlinear aging costs of different energy storage units due to different power characteristics, driving a collaborative optimization process aimed at minimizing the total lifecycle cost, and using actual operating data to provide feedback correction to the core model parameters. This multi-module, multi-timescale collaborative operation can proactively avoid accelerated battery aging caused by power surges, ensuring that the output hybrid energy storage capacity configuration scheme achieves optimal economic efficiency throughout the project lifecycle while meeting grid power requirements, significantly improving the long-term value and operational reliability of microgrid energy storage investment.

[0043] Example 2

[0044] The power demand profiling module is specifically used for:

[0045] Wavelet packet transform is used to process the net load power time series;

[0046] Based on the energy storage type boundary frequency, high-frequency power components and low-frequency energy components are separated from the net load power time series.

[0047] The energy storage type boundary frequency is determined by physical experimental calibration of the target battery; the energy storage type boundary frequency is the power change frequency corresponding to the inflection point of the nonlinear growth rate of the battery's life decay rate in the cyclic charge-discharge experiment.

[0048] To achieve refined management of hybrid energy storage units, the power demand segmentation module in this embodiment receives the net load power time series P from the microgrid monitoring and forecasting system.load (t), and call the wavelet packet transform algorithm to process it; the function is defined by the following formula:

[0049] (P high (t),P low (t))=WPT(P load (t),f c )

[0050] The underlying logic lies in the original net load power P load (t) contains complex frequency components that have vastly different impacts on the lifespan of different energy storage technologies; to achieve effective protection of energy storage units and accurate cost accounting, these mixed power demands must first be effectively separated; among them, P load (t) represents the input net load power time series, in kW; WPT(·) represents the wavelet packet transform function; f c It is the energy storage type boundary frequency, measured in Hz, and serves as a key control parameter in the WPT decomposition process; output P high (t) is the high-frequency power component specified by the response of power-type units such as supercapacitors, while P low (t) represents the low-frequency energy component that should be supplied by energy-type units such as batteries;

[0051] To further clarify, the energy storage type dividing frequency f c The determination method has a clear physical basis and is not arbitrarily set; it involves conducting multiple sets of charge-discharge experiments on the target battery model, monitoring its capacity decay curve, and calibrating the power change frequency corresponding to the inflection point where the rate of life decay begins to increase nonlinearly and sharply; this ensures the physical validity of the frequency division, and the calibration process includes the following steps:

[0052] Sample testing: Select battery samples of the target model from the same batch and divide them into several groups.

[0053] Variable frequency cycling: For each group of battery samples, under the same ambient temperature and average state of charge, different main frequencies f are applied, for example, from 0.001Hz to 1Hz, but with the same total energy throughput, to perform cyclic charge and discharge tests on sinusoidal or square wave power curves.

[0054] Data acquisition: After reaching the preset number of cycles, the capacity decay rate ΔC(f) of each sample is measured using a standard capacity testing method;

[0055] Inflection point determination: By fitting the capacity decay rate ΔC(f) as a function of the power change frequency f, the curve ΔC = g(f) is obtained; the second derivative of this curve is then calculated. The point where the second derivative exhibits a significant positive peak is identified as the inflection point of the nonlinear increase in the decay rate; the frequency corresponding to this point is the energy storage type boundary frequency f. c This makes P load The high-frequency power component P separated from (t) high (t) is precisely the main component that causes additional battery loss; this module will not contain the drastically fluctuating low-frequency energy component P. low (t) is passed to the battery cost model, and the high-frequency power component P is passed to it. high (t) is passed to the cost model of the supercapacitor; this approach effectively protects the battery from the source and lays a precise data foundation for subsequent optimization of the system's life cycle cost.

[0056] Example 3

[0057] The cost modeling module is specifically used to determine the annual aging cost of a battery, specifically for:

[0058] The annual aging cost is determined by accumulating the instantaneous aging cost over an annual period.

[0059] The instantaneous aging cost is proportional to the absolute value of the low-frequency energy component, and the unit energy throughput cost is corrected by a dynamic stress factor that is related to the square of the power change rate of the low-frequency energy component.

[0060] The power stress time constant in the dynamic stress factor is calibrated through physical experiments; the calibration methods include:

[0061] The target battery was subjected to baseline and experimental experiments. The baseline group used a smooth charge-discharge curve, while the experimental group used a charge-discharge curve with violent fluctuations.

[0062] Based on the cost difference between the two sets of experiments when the same degree of lifetime decay is achieved, the power stress time constant is solved inversely.

[0063] The inverse solution process is as follows:

[0064] Defined curves: The benchmark group uses smooth charge-discharge curves with constant current and constant voltage. The experimental group, based on the same average charge-discharge rate, superimposes a power fluctuation profile with known frequency and amplitude, such as a square wave, to introduce drastic power changes dP. low (t) / dt;

[0065] Cost accounting: Assume the battery replacement cost is C. replace The total aging cost at the end of its lifespan equals the replacement cost. When both the baseline and experimental groups reach the same degree of lifespan decline, for example, a 10% decrease in healthy state state oxygen saturation (SOH), the cumulative aging cost is the same, both being 0.1°C. replace ;

[0066] Establishing an equation: Based on the aging cost formula, the following equation can be established:

[0067]

[0068] Solve for the parameter: the charge / discharge power curve P corresponding to the same degree of lifetime degradation for the reference group and the experimental group in this calibration experiment. base (t) and P exp (t), and their respective total runtime T base and T exp All values ​​are measurements; therefore, the total energy throughput of the two groups is... and It is known, and they are usually not equal; since the cost equation includes ω stress All other parameters are known or measurable, therefore the power stress time constant ω can be uniquely solved from this equation. stress The value;

[0069] The cost modeling module is used to determine the annual aging cost of the battery. It abandons the traditional linear cost model and establishes a nonlinear aging cost model that can quantify the additional damage caused by power fluctuations; the mathematical expression is as follows:

[0070]

[0071] This formula aims to address the problem that existing models cannot quantify the nonlinear damage to battery life caused by dynamic power characteristics. The underlying logic is to introduce a dynamic stress term that is directly related to the rate of power change to punish drastic power fluctuations, thereby enabling the cost model to more realistically reflect the electrochemical aging mechanism of the battery.

[0072] In the formula, C aging,BAT The total aging cost of the battery over one year; T is the year; k E,BAT This is the basic cost per unit of energy throughput, expressed in yuan / kWh; P low (t) is the low-frequency energy component output by the power demand profiler module, in kW; |P low (t)| represents its absolute value, ensuring that both charging and discharging are included in the cost; dP low (t) / dt is the rate of change of power, in kW / h; P rated,BAT The rated power of the battery to be optimized, in kW, serves as a standardized benchmark for the rate of power change; ω stress It is the power stress time constant, with units of time squared (h). 2 ω is a key physical quantity characterizing the battery's sensitivity to power fluctuations; to ensure the mathematical rigor of the model, ω stress The dimension is set to h 2 , so that ωstress ·((dP low (t) / dt) / P rated,BAT ) 2 Becoming a dimensionless term ensures that the dimension of the entire integral term is yuan / hour, and the dimension of the final integral result is yuan; T base The duration of the baseline experiment refers to the time it takes for the baseline batteries to reach a preset lifespan degradation level, such as a 10% decrease in state of health (SOH); P base (t): The smoothed charge-discharge power curve used in the benchmark experiment; P exp (t): The charge-discharge power curve with drastic fluctuations used in the experimental group; T exp The duration of the experimental group experiment refers to the time it takes for the experimental group batteries to reach the same level of lifespan degradation as the baseline group.

[0073] To ensure that it can be implemented by those skilled in the art, the power stress time constant ω stress The method for obtaining ω is determined through physical experiments: two sets of comparative experiments are conducted on the target battery. The benchmark group is subjected to a smooth charge-discharge curve, while the experimental group is subjected to a power curve containing drastic fluctuations. Based on the difference in operating costs when the two groups of batteries reach the same lifespan degradation endpoint, such as the difference in operating costs when the state of health (SOH) drops to 80%, ω can be accurately solved. stress The value of this cost formula means that in subsequent optimizations, any configuration scheme that causes drastic changes in battery power will be subject to a quadratic level cost penalty, thereby guiding the optimization algorithm to select a better hybrid energy storage ratio, protecting the battery from the root and avoiding accelerated aging caused by power surges.

[0074] Meanwhile, the cost modeling module is based on the high-frequency power component P. high (t) determines the annual aging cost of the supercapacitor; unlike batteries, the physical aging mechanism of supercapacitors is mainly related to voltage and temperature, but in system-level cost optimization, the aging cost can be simplified to a linear model proportional to the energy throughput; this model assumes that the performance of the supercapacitor will degrade to a certain extent for every certain amount of energy processed; the annual aging cost C aging,SC The mathematical expression is as follows:

[0075]

[0076] Where, k E,SC This is the basic cost per unit energy throughput of a supercapacitor, expressed in yuan / kWh. It is an empirical constant calibrated using lifetime cycle data or experimental data provided by the manufacturer; C aging,SC,y : Annual aging cost of the supercapacitor in year y; |P high(t)| represents the absolute value of the high-frequency power component that the supercapacitor needs to respond to, in kW; T is the annual period. This model provides the cost term of the supercapacitor for subsequent co-optimization, making the objective function of minimizing the total cost complete, where C aging,BAT,y and C aging,SC,y These represent the aging costs of the battery and supercapacitor in the y-th operating year, calculated using the formula.

[0077] Example 4

[0078] The objective function of the collaborative optimization configuration module includes:

[0079] The initial investment cost is calculated based on the optimal capacity configuration to be determined and the preset unit capacity investment cost.

[0080] The total lifecycle operating cost is the sum of the present values ​​of the annual aging costs for all future years, calculated based on a preset discount rate.

[0081] The constraints of the collaborative optimization configuration module include:

[0082] The power balance constraint requires that the sum of the output power of the supercapacitor and the battery be no less than the instantaneous value of the net load power time series;

[0083] Physical constraints require that the operating power of supercapacitors and batteries does not exceed their respective rated power, and the rated power of batteries is related to the product of energy capacity and rated charge / discharge rate.

[0084] The collaborative optimization configuration module in this embodiment is the system's optimization decision module. It determines the most economical energy storage capacity ratio by constructing and solving a nonlinear programming problem. The objective function of this module follows standard engineering economics net present value analysis methods, aiming to minimize the total cost over the system's entire lifecycle, expressed as:

[0085]

[0086] This function aims to weigh one-time equipment procurement decisions against long-term, dynamically changing operating costs within the same framework, thereby making a globally optimal judgment; where C total It is the total cost of the system's entire lifecycle; the decision variable is E. BAT The rated energy capacity of the battery, kWh and P SC The rated power of the supercapacitor is kW; the first term is the initial investment cost, where k... cap,BAT and k cap,SC The first term represents the market unit capacity investment cost of batteries and supercapacitors, respectively; the second term represents the sum of the discounted present value of the total life-cycle operating costs, where C... aging,BAT,y and C aging,SC,yy represents the annual aging cost in year y, calculated by the aforementioned cost modeling module; L is the system's planned lifespan (in years), and r is the economic discount rate;

[0087] The collaborative optimization configuration module follows a series of constraints during the solution process; the power balance constraint P BAT (t)+P SC (t)≥P load (t) Ensure that the energy storage system always meets the grid demand; physical constraints |P BAT (t)|≤P rated,BAT and |P SC (t)|≤P SC To ensure the equipment operates within a safe range, a key intrinsic factor is the battery's rated power P. rated,BAT With decision variable E BAT Through the rated charge / discharge rate C rate Related, i.e., P rated,BAT =C rate ·E BAT This association tightly couples the decision variables with the penalty term in the battery aging cost model. Based on the above objective function and constraints, this module accurately finds the optimal solution in a complex decision space by calling a nonlinear programming solver. and C rate Rated charge / discharge rate of the battery; P BAT (t) represents the instantaneous output power of the battery; P SC (t) represents the instantaneous output power of the supercapacitor.

[0088] Example 5

[0089] The parameter feedback correction module is specifically used for:

[0090] The actual health status of the battery is collected, and the predicted health status is calculated based on the optimal capacity configuration output by the collaborative optimization configuration module and the actual operating power data.

[0091] Determine the deviation between actual health status and predicted health status;

[0092] A proportional-integral controller is used to update the power stress time constant using the deviation, which can be used for optimization or capacity expansion planning in the next cycle.

[0093] The parameter feedback correction module aims to endow the system with adaptive correction capabilities, thereby transforming the static design process into a dynamic closed-loop correction process. Its internal logic involves continuously calibrating and optimizing key parameters of its internal model using actual operating data to address the inevitable mismatch between the theoretical model and complex realities. Its correction rules are executed by a proportional-integral controller.

[0094]

[0095] Where, ω stress (y+1) is the updated power stress time constant for the next optimization cycle; ω stress (y) is the value used in the current period; e(y) is the prediction deviation of the current period, defined as e(y) = SOH actual (y)-SOH predict (y), here SOH actual (y) Data actually collected by the battery management system, SOH predict (y) is the theoretical value calculated by this system based on the initial configuration and actual operating power data through an aging model;

[0096] The predicted health status SOH predict (y) is directly related to the cumulative aging cost of the battery, and is calculated as follows:

[0097]

[0098] in, This is the cumulative battery aging cost calculated by the cost modeling module from the start of operation to year y; C replace,BAT This represents the initial investment cost or equivalent replacement cost of the battery, signifying its total value over its entire lifecycle. This formula establishes a clear quantitative relationship between the cumulative economic costs and the physical degradation of its health.

[0099] K p and K i These are the proportional gain and integral gain, respectively, which serve as adjustable parameters of the controller. Wherein, K... p It is the proportional gain, with units of h. 2 ;K i It is the integral gain, in h. 2 / Year;

[0100] These gain parameters are not arbitrarily set. Their initial values ​​can be initially estimated using mature controller parameter tuning methods in engineering, such as the Ziegler-Nichols method. After the system is put into actual operation, it can be further optimized by empirical fine-tuning or online adaptive algorithms based on the convergence characteristics of the deviation e(y) to ensure the stability and speed of the correction process, thereby achieving the best adaptive correction effect; j: the year index from the start of operation to the y-th year; Δt: the execution time step of the PI controller; according to the context, the correction is performed on an annual cycle, so Δt represents one cycle, i.e., one year;

[0101] After the system is put into operation, this module performs periodic corrections; it continuously compares the actual battery health status obtained by the BMS with the prediction results of its own model; once a deviation e(y) is detected, the PI controller intervenes and adjusts the key parameter ω. stress Fine-tuning will be performed; if the actual aging is faster than predicted (e(y) is negative), the controller will increase ω. stress The value of this value is used to increase the penalty for power fluctuations in future optimizations. This mechanism ensures that the system's cost model can continuously track actual aging characteristics, so that the decision-making basis in future operation strategy optimization or capacity expansion planning always remains realistic and accurate, thereby achieving dynamic optimization throughout the entire life cycle.

[0102] It should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and are not intended to limit it. Although the present invention has been described in detail with reference to preferred embodiments, those skilled in the art should understand that modifications or equivalent substitutions can be made to the technical solutions of the present invention without departing from the spirit and scope of the technical solutions of the present invention, and all such modifications or substitutions should be covered within the scope of the claims of the present invention.

Claims

1. A capacity configuration system for a hybrid energy storage platform in a microgrid, characterized in that, include: The power demand segmentation module is used to decompose the acquired net load power time series into high-frequency power components and low-frequency energy components based on a preset energy storage type boundary frequency. The cost modeling module is used to determine the annual aging cost of supercapacitors based on high-frequency power components. The annual aging cost of the battery is determined based on the low-frequency energy component; the annual aging cost of the battery includes the stress-related additional costs associated with the power change rate of the low-frequency energy component. The collaborative optimization configuration module is used to combine the annual aging cost of the supercapacitor and the battery determined by the cost modeling module, as well as the preset initial investment cost, to calculate the optimal capacity configuration of the supercapacitor and battery with the goal of minimizing the total cost of the system's entire life cycle. The parameter feedback correction module is used to dynamically correct the power stress time constant in the cost modeling module based on the deviation between the actual health state and the predicted health state of the battery. The inverse process of solving the power stress time constant includes: Formula 1: ； In the calibration experiment, the charge-discharge power curves corresponding to the same degree of lifetime decay for the reference group and the experimental group were compared. and and their respective total runtime and All values ​​are measurements; therefore, the total energy throughput of the two groups is... and They are known and usually not equal; because the cost equation includes, except for All other parameters are known or measurable, therefore the power stress time constant can be uniquely solved from the equation. The value; The cost modeling module is used to determine the annual aging cost of the battery. It abandons the traditional linear cost model and establishes a nonlinear aging cost model that can quantify the additional damage caused by power fluctuations; Formula 2 is established: ； In Formula 1 and Formula 2, The total aging cost of a battery over a one-year cycle; It is an annual cycle; This is the basic cost per unit of energy throughput, expressed in yuan / kWh; It is the low-frequency energy component output by the power demand profiler module, in kW; This represents its absolute value, ensuring that both charging and discharging costs are factored into the calculation. It is the rate of change of power, in kW / h; It is the rated power of the battery to be optimized, in kW, which serves as a standardized benchmark for the rate of power change. It is the power stress time constant, with units of time squared ( ( ), is a key physical quantity characterizing the battery's sensitivity to power fluctuations; to ensure the mathematical rigor of the model, The dimensions are set as , making Becoming a dimensionless term ensures that the dimension of the entire integral term is yuan / hour, and the dimension of the final integral result is yuan. The duration of the baseline experiment refers to the time during which the baseline batteries reach the preset lifespan degradation level. The smooth charge-discharge power curves used in the benchmark experiments; The charge-discharge power curves used in the experimental group's experiments contained drastic fluctuations. The duration of the experimental group experiment refers to the time it takes for the experimental group batteries to reach the same level of lifespan degradation as the baseline group.

2. The capacity configuration system for a hybrid energy storage platform in a microgrid according to claim 1, characterized in that, The power demand profiling module is specifically used for: Wavelet packet transform is used to process the net load power time series; Based on the energy storage type boundary frequency, high-frequency power components and low-frequency energy components are separated from the net load power time series.

3. The capacity configuration system for a hybrid energy storage platform in a microgrid according to claim 2, characterized in that, The energy storage type boundary frequency is determined by physical experimental calibration of the target battery; the energy storage type boundary frequency is the power change frequency corresponding to the inflection point of the nonlinear growth rate of the battery's life decay rate in the cyclic charge-discharge experiment.

4. The capacity configuration system for a hybrid energy storage platform in a microgrid according to claim 1, characterized in that, The cost modeling module is specifically used to determine the annual aging cost of a battery, specifically for: The annual aging cost is determined by accumulating the instantaneous aging cost over an annual period. The instantaneous aging cost is proportional to the absolute value of the low-frequency energy component, and the unit energy throughput cost is corrected by a dynamic stress factor that is related to the square of the power change rate of the low-frequency energy component.

5. The capacity configuration system for a hybrid energy storage platform in a microgrid according to claim 4, characterized in that, The power stress time constant in the dynamic stress factor is calibrated through physical experiments; the calibration methods include: The target battery was subjected to baseline and experimental experiments. The baseline group used a smooth charge-discharge curve, while the experimental group used a charge-discharge curve with violent fluctuations. Based on the cost difference between the two sets of experiments when the same degree of lifetime decay is achieved, the power stress time constant is solved inversely.

6. The capacity configuration system for a hybrid energy storage platform in a microgrid according to claim 1, characterized in that, The objective function of the collaborative optimization configuration module includes: The initial investment cost is calculated based on the optimal capacity configuration to be determined and the preset unit capacity investment cost. The total lifecycle operating cost is the sum of the present value of the annual aging costs for all future years, calculated based on a preset discount rate.

7. The capacity configuration system for a hybrid energy storage platform in a microgrid according to claim 1, characterized in that, The constraints of the collaborative optimization configuration module include: The power balance constraint requires that the sum of the output power of the supercapacitor and the battery be no less than the instantaneous value of the net load power time series; Physical constraints require that the operating power of supercapacitors and batteries does not exceed their respective rated power, and the rated power of batteries is related to the product of energy capacity and rated charge / discharge rate.

8. The capacity configuration system for a hybrid energy storage platform in a microgrid according to claim 1, characterized in that, The parameter feedback correction module is specifically used for: The actual health status of the battery is collected, and the predicted health status is calculated based on the optimal capacity configuration output by the collaborative optimization configuration module and the actual operating power data. Determine the deviation between actual health status and predicted health status; A proportional-integral controller is used to update the power stress time constant using the deviation, which can be used for optimization or capacity expansion planning in the next cycle.

Images