A method for setting control parameters of an energy storage converter for frequency modulation

Abstract

The application discloses a kind of energy storage inverter control parameter setting method for frequency modulation, it is related to power system frequency modulation technical field, it can solve the problems of static, poor cooperation, real-time deficiency and not considering energy storage constraint in existing energy storage inverter control parameter setting.The present application comprises: the basic range of virtual inertia coefficient, damping coefficient and droop coefficient is calibrated based on engineering constraint;With minimizing frequency deviation peak value, minimizing power oscillation amplitude and maximizing energy storage utilization efficiency as optimization goal, output power model is constructed and initial parameters are solved;Real-time data is collected to identify disturbance type and intensity, and control parameters are dynamically adjusted according to the difference correction strategy of load mutation disturbance or new energy fluctuation disturbance;Establish a verification index system, and ensure the effectiveness of parameters through closed-loop feedback adjustment.The present application realizes the dynamic matching of control parameters, system frequency modulation demand and energy storage operating state.

Description

Technical Field

[0001] This invention relates to the field of power system frequency regulation technology, and specifically to a method for setting control parameters of an energy storage converter for frequency regulation. Background Technology

[0002] With the advancement of the carbon peaking and carbon neutrality strategy, the proportion of intermittent renewable energy sources such as wind power and photovoltaics in the power system continues to rise, while the installed capacity of traditional synchronous generators is declining. This leads to a significant reduction in the system's physical inertia and a weakening of its frequency regulation capability. When the power grid experiences disturbances such as sudden load changes or fluctuations in renewable energy power, the frequency deviation is prone to exceed the safety threshold and may even trigger low-frequency oscillations, threatening the safe and stable operation of the power grid.

[0003] As a flexible frequency regulation resource, energy storage converters can respond to system frequency changes and compensate for insufficient system inertia by configuring virtual inertia control, damping control, and droop control. However, existing methods for setting control parameters for energy storage converters have the following problems.

[0004] First, the parameter settings are static. Most methods use fixed parameters or are based on trial and error based on experience, without considering the differences in the dynamic characteristics of the system frequency, the state of energy storage (SOC), and the type of disturbance. This leads to a mismatch between the frequency regulation response and the actual demand, resulting in problems such as large overshoot and long recovery time.

[0005] Second, the parameters have poor coordination. Existing technologies mostly optimize for a single parameter, ignoring the coupling relationship between virtual inertia, damping, and droop parameters, which can easily lead to imbalance in the dynamic response of the system.

[0006] Third, the modeling is complex and lacks real-time performance. Some optimization methods rely on high-precision global system models, require a large amount of measured data, and involve cumbersome parameter calculations, making it difficult to meet the real-time requirements of millisecond-level frequency regulation after grid disturbances.

[0007] Fourth, the constraints of energy storage operation were not considered. Failure to incorporate operational constraints such as the energy storage's state of charge (SOC) and charging / discharging power limits into parameter optimization may lead to overcharging and discharging of the energy storage, shortening equipment lifespan or causing secondary frequency fluctuations.

[0008] Therefore, it is urgent to construct a four-stage dynamic parameter setting method, which includes basic boundary calibration, multi-objective collaborative optimization, real-time operating condition correction, and closed-loop verification feedback, to achieve dynamic matching between control parameters and system frequency regulation requirements and energy storage operating status, improve frequency regulation accuracy and response speed, and ensure the safety of energy storage equipment. Summary of the Invention

[0009] The present invention aims to solve at least one of the technical problems existing in the prior art, and provides a method for setting control parameters of energy storage converters for frequency regulation.

[0010] To achieve the above objectives, the present invention provides a method for setting control parameters of an energy storage converter for frequency regulation, comprising: Obtain the hardware parameters of the energy storage converter and the system safety standard parameters. Calculate the basic range upper and lower limits of the virtual inertia coefficient, virtual damping coefficient, and droop coefficient based on engineering constraints. The hardware parameters of the energy storage converter include rated active power, inherent response time, and maximum charge / discharge power. The system safety standard parameters include the maximum allowable frequency change rate, allowable frequency deviation, frequency dead zone, and allowable frequency overshoot. With the optimization objectives of minimizing the peak frequency deviation, minimizing the power oscillation amplitude, and maximizing the energy storage utilization efficiency, an SOC adaptive coefficient is introduced to construct an output power model of the energy storage converter that integrates virtual inertia control, damping control, and droop control. The initial control parameters are solved by an optimization algorithm within the stated range. Real-time grid frequency, frequency change rate, and energy storage SOC data are collected and filtered and smoothed. Based on the preprocessed data, the disturbance type and disturbance intensity are identified by frequency deviation, frequency change rate, and disturbance duration. According to the disturbance type, a differentiated correction strategy is adopted to dynamically correct the real-time control parameters based on the initial control parameters. A verification index system is established to perform closed-loop verification of the real-time control parameters. If all verification indicators meet the preset threshold, the real-time control parameters are used as the final effective parameters. Otherwise, targeted feedback adjustments are made based on the type of indicator that does not meet the threshold until the preset threshold is met.

[0011] Furthermore, the formula for calculating the upper limit of the basic range of the virtual inertia coefficient is as follows: ; in, This represents the upper limit of the virtual inertia coefficient. The rated active power of the energy storage converter, This is the inherent response time of the converter. The system's reference angular frequency, This represents the maximum permissible rate of change of the system's frequency. The lower limit of the basic range of the virtual inertia coefficient The upper limit The preset first proportional coefficient is times.

[0012] Furthermore, the formula for calculating the upper limit of the basic range of the virtual damping coefficient is as follows: ; in, This represents the upper limit of the virtual damping coefficient. This represents the maximum charge and discharge power of the energy storage converter. To allow for frequency overshoot, This is a frequency dead zone; The lower limit of the basic range of the virtual damping coefficient The upper limit The preset second proportional coefficient multiple; The formula for calculating the upper limit of the basic range of the droop coefficient is as follows: ; in, This is the upper limit of the droop coefficient. This represents the maximum value of the SOC adaptive coefficient. This refers to the allowable value for frequency deviation. The lower limit of the basic range of the droop coefficient The upper limit The preset third proportional coefficient multiple.

[0013] Furthermore, the calculation formula for the output power model of the energy storage converter is as follows: ; in, To output active power to the energy storage converter, For virtual inertia coefficient, The rate of change of frequency, This is the virtual damping coefficient. For frequency deviation, The droop coefficient is... These are the SOC adaptive coefficients; The formula for calculating the SOC adaptive coefficient is as follows: ; in, For the energy storage state of charge, the SOC adaptive coefficient ranges from 0.4 to 1.2.

[0014] Furthermore, the process of solving for the initial control parameters using an optimization algorithm includes: Construct a fitness function, the formula for which the fitness function is calculated is: ; in, For the fitness function value, , , These are the peak frequency deviation weight, the power oscillation amplitude weight, and the energy storage utilization efficiency weight, respectively, and the weights satisfy... , This represents the peak value of the frequency deviation. This represents the amplitude of the power oscillation. The particle swarm optimization algorithm is used to iteratively solve the problem within the basic range. When the change in the fitness function value of consecutive preset algebras is less than a preset convergence threshold or the preset maximum number of iterations is reached, the initial control parameters are output.

[0015] Furthermore, the filtering and smoothing preprocessing includes: A first-order low-pass filter is used to filter noise from the acquired data. The filtering formula is as follows: ; in, , The sampling period is The filtering time constant is This is the original data at the current moment. This is the filtered data. This is the filtered data from the previous time step; The sliding window method is used to smooth the filtered data. The smoothing formula is as follows: ; in, For window length, This is the smoothed data.

[0016] Furthermore, the identification of disturbance type and disturbance intensity includes: Calculate the duration of the disturbance The formula for calculating the duration of the disturbance is: ; in, This is the current sampling point number. The sampling period is The time of disturbance occurrence is determined by frequency mutation detection; When the absolute value of the smoothed frequency deviation is greater than the first frequency threshold, the absolute value of the smoothed frequency change rate is greater than the first change rate threshold, and the duration of the disturbance is less than the preset time threshold, it is determined to be a load change disturbance. When the absolute value of the smoothed frequency deviation is within the range of the second frequency threshold, the absolute value of the smoothed frequency change rate is less than or equal to the second change rate threshold, and the duration of the disturbance is greater than or equal to the preset time threshold, it is determined to be a new energy fluctuation disturbance. Calculate the disturbance intensity coefficient The formula for calculating the disturbance intensity coefficient is as follows: ; in, This represents the peak frequency deviation up to the current moment. When the disturbance intensity coefficient is greater than the preset intensity threshold, it is determined to be a strong disturbance; otherwise, it is determined to be a weak disturbance.

[0017] Furthermore, when the disturbance type is the aforementioned load abrupt disturbance, the correction formula for the differentiated correction strategy is: ; ; ; in, , , These are the corrected virtual inertia coefficient, virtual damping coefficient, and droop coefficient, respectively. , , These are the initial virtual inertia coefficient, the initial virtual damping coefficient, and the initial droop coefficient, respectively. The smoothed rate of change of frequency. The smoothed frequency deviation, For real-time SOC adaptive coefficients.

[0018] Furthermore, when the disturbance type is the aforementioned new energy fluctuation disturbance, the correction formula for the differentiated correction strategy is: ; ; ; in, This is the lower limit of the virtual inertia coefficient. This represents the lower limit of the virtual damping coefficient.

[0019] Furthermore, the verification index system includes frequency deviation correction time. Power oscillation attenuation rate SOC change rate And parameter adjustment smoothness ; The formula for calculating the power oscillation attenuation rate is as follows: ; in, This represents the peak value of the first power oscillation after the disturbance. This is the peak value of the second power oscillation; The formula for calculating the rate of change of SOC is: ; in, For a complete frequency modulation cycle, The SOC value at the moment the disturbance occurs. The SOC value at the end of the frequency modulation cycle; The feedback adjustments include: When the frequency deviation correction time is greater than the preset correction time threshold, the droop coefficient is increased to a first adjustment factor multiple of the original value and the virtual inertia coefficient is increased to a second adjustment factor multiple of the original value. When the power oscillation attenuation rate is less than the preset attenuation threshold, the virtual damping coefficient is increased to the third adjustment factor multiple of the original value; When the absolute value of the SOC change rate is greater than the preset rate threshold, the virtual inertia coefficient and the droop coefficient are reduced to a fourth adjustment factor multiple of their original values. When the smoothness of the parameter adjustment is greater than the preset smoothing threshold, the parameter is smoothed by low-pass filtering.

[0020] The beneficial effects of this invention are as follows: This invention deeply integrates virtual inertia, damping, droop parameters with SOC adaptive adjustment, and achieves synergistic matching of the three through a multi-objective optimization model. This solves the response imbalance problem caused by parameter coupling in the prior art, significantly improves frequency modulation dynamic performance, and can control the frequency deviation correction time within 1.5s, while achieving a power oscillation attenuation rate of over 0.75.

[0021] This invention is based on a differentiated correction strategy for disturbance type and intensity, which can adapt to complex operating conditions such as load changes and new energy fluctuations. Compared with fixed parameter methods, the frequency regulation response speed is improved by more than 30%, which can meet the diverse frequency regulation needs of grids with a high proportion of new energy connected to the grid.

[0022] This invention introduces an adaptive SOC coefficient and a charge / discharge rate constraint, strictly controlling the SOC change rate to within 0.08 / s, avoiding overcharging and discharging of energy storage, extending equipment life by more than 30%, and ensuring that energy storage can continuously participate in frequency regulation.

[0023] The parameter calculation of this invention is based on measurable physical quantities in engineering, without relying on complex global system models. The algorithm has low complexity and can be embedded in the energy storage converter controller without additional hardware investment. It is compatible with various energy storage carriers such as lithium batteries, flow batteries, and flywheels, and is easy to promote in engineering.

[0024] This invention ensures that every effective parameter undergoes rigorous performance testing through closed-loop verification and feedback adjustment of four types of quantitative indicators, thereby avoiding risks to the power grid or energy storage equipment caused by unqualified parameters and improving the reliability and stability of the entire method. Attached Figure Description

[0025] Figure 1 This is a flowchart illustrating a method for setting control parameters of a frequency-modulated energy storage converter according to a specific embodiment of the present invention. Detailed Implementation

[0026] To make the objectives, technical solutions, and beneficial effects of this application clearer, the following detailed description, in conjunction with the accompanying drawings and specific embodiments, further illustrates this application. It should be understood that the specific embodiments described in this specification are merely for explaining this application and are not intended to limit it.

[0027] like Figure 1 As shown in the illustration, this invention provides a method for setting control parameters of an energy storage converter for frequency regulation. This method is applied to an energy storage frequency regulation system in a high-proportion renewable energy grid-connected scenario. The system is equipped with one energy storage converter, and the energy storage carrier is a lithium battery energy storage system. This invention achieves power system frequency regulation by dynamically optimizing the virtual inertia coefficient, damping coefficient, and droop coefficient of the energy storage converter.

[0028] The core parameter of this invention is defined as follows: virtual inertia coefficient. The virtual damping coefficient characterizes the ability of the energy storage converter to simulate the inertia support of a synchronous machine, reflecting the converter's sensitivity to frequency changes. A higher value indicates a stronger ability to suppress frequency abrupt changes, thus mitigating sudden frequency rises or falls. Used to suppress frequency and power oscillations, attenuate dynamic oscillations during frequency modulation, and shorten the system's recovery time. A larger value results in faster oscillation attenuation, preventing the frequency from fluctuating repeatedly around a stable value; droop coefficient. This is the ratio of frequency deviation to the active power output of the energy storage unit, used for static frequency deviation correction. A larger value results in higher static frequency recovery accuracy, allowing the frequency to return to near its rated value; SOC adaptive coefficient. The control parameter weights are dynamically adjusted based on the remaining energy storage capacity, with a value range of 0.4 to 1.2. The core function is to avoid overcharging and discharging of the energy storage. The value is the largest when the SOC is 0.5, and the converter participates in frequency regulation at full load capacity. The value is the smallest when the SOC is close to 0.2 or 0.8, and the converter derated for frequency regulation.

[0029] The core parameters of this invention are configured as follows: Regarding the hardware parameters of the energy storage converter, the rated active power... For 40MW, the inherent response time The maximum charge / discharge power is 0.1s. The capacity is 44MW, calculated based on the manufacturer's stated overload factor of 1.1. Regarding system safety standard parameters, the maximum permissible frequency variation rate is determined according to the "Guidelines for the Safety and Stability of Power Systems". The allowable frequency deviation is 0.5 Hz / s. The frequency dead zone is 0.2Hz. The allowable frequency overshoot is 0.03Hz. The SOC is 8%; in terms of energy storage operation constraints, the SOC range is 0.2 to 0.8, which is determined based on the cycle life test results of the energy storage battery to avoid battery capacity decay caused by deep charging and discharging. The optimal SOC range is 0.4 to 0.6, in which the energy storage battery has the highest charging and discharging efficiency and the least life loss. The initial SOC of energy storage is 0.55, and the charging and discharging power is limited to ±40MW.

[0030] The method for setting control parameters of a frequency-modulated energy storage converter according to the present invention includes the following steps: Step 1: Determine the basic range of calibration parameters based on engineering constraints.

[0031] This step is fundamental to the entire parameter setting method. Its core purpose is to provide a legitimate search boundary for subsequent parameter optimization, preventing parameters from exceeding hardware capabilities or system safety limits, and ensuring that all subsequent calculation results are engineering-feasible. Its inputs consist of three types of basic data: energy storage converter hardware parameters, system safety standard parameters, and energy storage operation constraint parameters. The output is the virtual inertia coefficient. Virtual damping coefficient and droop coefficient The clearly defined upper and lower limits are directly used as the optimization boundary constraints in step two.

[0032] First, a comprehensive collection of basic parameter data is conducted. The hardware parameters collected for the energy storage converter include: rated active power. The capacity is 40MW, determined by the rated capacity of the converter power modules; please refer to the manufacturer's technical manual. Inherent response time... The value is 0.1s, determined by the processing speed of the converter controller chip and the switching frequency of the power devices, and obtained from the dynamic response test report provided by the manufacturer, with a range of 0.05s to 0.2s; maximum charge / discharge power. It is 44MW, which is usually equal to the rated power. If the converter supports short-term overload, the overload factor can be calculated according to the manufacturer's specifications. This implementation uses an overload factor of 1.1. System safety standard parameters are determined based on the "Guidelines for the Safety and Stability of Power Systems": the maximum permissible frequency variation rate of the system. 0.5 Hz / s; permissible frequency deviation value The frequency is 0.2Hz, set according to power grid dispatch requirements, with a range of 0.15Hz to 0.25Hz; frequency dead zone. The set frequency is 0.03Hz. To avoid equipment damage caused by frequent converter operation, the converter will not respond when the frequency deviation is less than this value. The value range is 0.02Hz to 0.04Hz; allowable frequency overshoot. The frequency deviation during frequency modulation must not exceed 8%. The value ranges from 5% to 10%. Energy storage operation constraints include: the energy storage SOC operating range is 0.2 to 0.8, which is determined based on the energy storage battery cycle life test results to avoid battery capacity decay caused by deep charge and discharge; the optimal SOC range is 0.4 to 0.6, within which the energy storage battery has the highest charge and discharge efficiency and the least life loss.

[0033] Then, the basic range of the parameters is calculated. Based on the above basic data, the upper and lower limits of each parameter are calculated item by item using engineering formulas. During the calculation process, the principle of unit consistency must be strictly followed to ensure the accuracy of the results.

[0034] Virtual inertia coefficient The calculation logic for the basic range is that the magnitude of the inertia must match the converter's response capability and the system's frequency surge suppression requirements. This avoids situations where the inertia is too small, leading to ineffective suppression of frequency surges, or too large, causing converter response delays. Upper limit value. The calculation formula is: ; in, The system reference angular frequency is set to a value of [value missing]. rad / s, corresponding to a power frequency of 50Hz, is a fixed physical quantity and requires no adjustment. Lower limit value. Take 20% of the upper limit, that is, set the first proportional coefficient to 0.2, and the formula is: This ensures sufficient optimization space while avoiding parameters that are too small and thus lose their inertia support. In this embodiment, the calculated values ​​are obtained by substituting the parameters: ; ; Virtual damping coefficient The calculation logic for the basic range is that the damping magnitude needs to balance the oscillation suppression effect and response speed, avoiding excessively small damping that leads to excessively long oscillation duration, or excessively large damping that leads to slow converter response. Upper limit value. The calculation formula is: ; in, This is the overshoot correction factor, ensuring that the damping effect effectively limits frequency overshoot. Lower limit value. Take 30% of the upper limit, that is, preset the second proportional coefficient to be 0.3, the formula is as follows: This ensures that the damping effect is not too weak.

[0035] In this embodiment, the following is calculated by substituting the parameters: ; ; Sag coefficient The calculation logic for the basic range is that the droop factor must ensure the static frequency deviation correction capability while avoiding power output overflow. Upper limit value. The calculation formula is: ; in, This is the maximum value of the SOC adaptive coefficient, which is 1.2. When the SOC is equal to 0.5, it ensures that the converter has the strongest static frequency regulation capability under the optimal SOC state. The effective frequency deviation range refers to the frequency deviation interval in which the converter actually participates in the correction. Lower limit value. Take 40% of the upper limit, that is, preset the third proportional coefficient to be 0.4, the formula is as follows: This ensures that the sag adjustment will not fail. In this embodiment, the following calculation is obtained by substituting the parameters: ; ; By strictly limiting the feasible domain of parameters through engineering constraints, it is ensured that the subsequent optimization process in step two will not produce invalid solutions that exceed hardware capabilities or system security requirements, thus laying the foundation for the engineering feasibility of the entire method.

[0036] Step 2: Solve for the initial control parameters based on the multi-objective collaborative optimization model.

[0037] This step uses the baseline defined in step one as a rigid boundary to achieve the virtual inertia coefficient through multi-objective optimization. Virtual damping coefficient and droop coefficient This method utilizes coordinated matching to address the poor parameter coordination issues in existing technologies. Its inputs are the parameter range from step one, the system frequency regulation target, and energy storage operation constraints; the output is the initial parameter set. , , This parameter set serves as the benchmark for real-time correction in step three and possesses good general frequency modulation capabilities.

[0038] First, the core objectives of the multi-objective optimization are clearly defined. Combining the core needs of grid frequency regulation with the constraints of energy storage safe operation, three indivisible optimization objectives are set. The priority of each objective is balanced using weighting coefficients, with the default condition being met. It can be dynamically adjusted according to the actual scenario: Objective 1 is to minimize the peak value of the frequency deviation, and the weights are... The value is 0.45, and the core requirement is that the peak frequency deviation during frequency modulation does not exceed [a certain value]. To ensure grid frequency stability and prevent protection actions from being triggered by frequencies exceeding safety thresholds; objective 2 is to minimize power oscillation amplitude, with weights... The weight is 0.3. The core requirement is that the oscillation amplitude of the energy storage converter's output power does not exceed 10% of the rated power, reducing the secondary impact of power fluctuations on the power grid and avoiding the initiation of new frequency fluctuations; Objective 3 is to maximize the utilization efficiency of energy storage, with a weight of 0.3. The core requirement is to maintain the energy storage SOC in the optimal range of 0.4 to 0.6 through adaptive SOC adjustment, thereby avoiding overcharging and discharging, extending the lifespan of the energy storage equipment, and ensuring that the energy storage can continuously participate in frequency regulation.

[0039] Then, strict constraints are set. These constraints define the rule boundaries of the optimization process, directly relating to the basic scope of step one and actual engineering limitations, ensuring that the optimization result is not only optimal but also feasible. The parameter boundary constraints strictly adhere to the calculation results of step one, i.e. , , No parameter is allowed to exceed this range. Power constraints require the energy storage converter to output active power... Must meet The negative sign indicates charging, and the positive sign indicates discharging, to prevent power overflow that could damage power devices. The SOC constraint requires the energy storage SOC to always be within the range of 0.2 to 0.8, achieved through an adaptive SOC coefficient. Implementing dynamic constraints, i.e. When the SOC approaches its limit, Reducing the frequency converter output power forces the State of Charge (SOC) to remain within a safe range. Response speed constraints require a specific frequency modulation response time. s, where Defined as the time from the occurrence of a disturbance to the energy storage output power reaching 50% of the rated power, this constraint ensures that the converter can respond to the grid frequency regulation demand and meet the millisecond-level frequency regulation requirements.

[0040] Next, an output power model for the energy storage converter is established. To accurately describe the mapping relationship between parameters and frequency regulation performance, an output power model integrating virtual inertia control, damping control, droop control, and SOC adaptive adjustment is established. This model is the core basis for the optimization process and directly relates the control parameters to the actual frequency regulation performance. ; in, Hz, 50Hz is the real-time frequency of the power grid, and 50Hz is the rated power frequency. , i.e., the rate of change of frequency, is obtained by performing first-order difference calculation on real-time frequency data; Its value changes dynamically with SOC, achieving a balance between energy storage safety and frequency regulation capability. The core innovation of this model lies in its deep integration of three types of control strategies with SOC constraints, rather than simple superposition, ensuring that the coupling relationship of various factors is fully considered during parameter optimization, and avoiding system imbalance caused by single parameter optimization.

[0041] Finally, an improved particle swarm optimization (PSO) algorithm was used to solve for the initial parameters. The improved PSO algorithm was chosen for optimization because it features fast convergence, strong robustness, and ease of engineering implementation, making it suitable for frequency modulation scenarios with high real-time requirements. Algorithm parameter initialization: the population size was set to 30, which ensures population diversity to avoid getting trapped in local optima while controlling computational load to ensure real-time performance; the number of iterations was set to 50, as extensive simulations have verified that 50 iterations allow the algorithm to converge sufficiently, avoiding poor optimization results due to insufficient iterations; the inertia weight adopted a linear decreasing strategy from 0.5 to 0.9, i.e., an inertia weight of 0.9 in the early stages of iteration to enhance global search capabilities and avoid missing the optimal solution, and an inertia weight of 0.5 in the later stages of iteration to enhance local search capabilities and approach the optimal solution; the learning factor... and The value is set to 1.8 to balance the influence of individual experience (i.e., one's own historical best solution) and group experience (i.e., the population's historical best solution) on particle updates.

[0042] The fitness function is a core indicator for measuring optimization performance. It is calculated by weighting and summing the three optimization objectives. The smaller the fitness function value, the better the optimization performance. ; in, This represents the peak value of the frequency deviation. The amplitude of power oscillation. To account for the degree to which the SOC deviates from the optimal value, all three terms are normalized to ensure the effectiveness of the weighting coefficients.

[0043] The iterative solution process is as follows: The first step is to initialize the population. Within the parameter range of step one, 30 sets of parameter combinations, i.e., 30 particles, are randomly generated, and each set of parameters satisfies the constraints. The second step is to calculate the fitness by substituting each set of parameters into the power model, simulating the frequency modulation process, and calculating the corresponding fitness function value. The third step is to update the individual optimal and the population optimal. Record the historical optimal fitness and corresponding parameters of each particle, which is the individual optimal, and record the historical optimal fitness and corresponding parameters of the entire population, which is the population optimal. The fourth step is to update the particle positions. Based on the inertia weight, learning factor, individual optimality, and group optimality, the position of each particle, i.e., the parameter combination, is updated. The fifth step is constraint checking. The updated particle positions are checked for constraints. If the parameters exceed the basic range in step one, they are corrected to the range boundary. Step 6: Convergence check. Repeat steps 2 through 5 until the fitness function value changes by less than 1 / 2 for 8 consecutive generations. Or reach the maximum number of iterations, 50. Output initial parameters: After the iteration ends, use the parameter set corresponding to the population optimum as the initial control parameters. , , This parameter set can achieve good frequency modulation effect under normal operating conditions without specific disturbances, providing a benchmark for real-time correction in step three.

[0044] In this implementation, after 50 iterations, the convergence condition is met, i.e., the change in fitness is less than a certain value for eight consecutive generations. The initial parameters are obtained as follows: It is 0.018 kg·m². It is 800 MW·s / Hz. It is 200MW / Hz.

[0045] Step 3: Dynamically correct control parameters based on real-time operating conditions.

[0046] This step is the core of achieving dynamic parameter adaptation, solving the problems of static parameters and poor adaptability in existing technologies. Its core logic is to sense changes in operating conditions in real time, perform targeted corrections based on initial parameters, and take real-time data collected by the PMU and the initial parameters from step two as inputs. The output is real-time parameters that match the current operating conditions. , , It is directly sent to the converter controller for execution.

[0047] First, real-time data acquisition and preprocessing are performed. The accuracy and stability of the acquired data are prerequisites for parameter calibration and must be strictly followed according to the procedures below. Regarding the acquisition equipment and installation location, a PMU device is installed at the AC side outlet of the energy storage converter. This location allows for direct acquisition of real-time data on the converter's output power and grid frequency, avoiding data distortion caused by line losses. Regarding the acquisition content and frequency, the PMU sampling frequency is set to 100Hz to meet millisecond-level frequency regulation response requirements. The following four types of data are acquired in real-time: grid-side data, including the real-time grid frequency. Accuracy of ±0.001Hz, frequency change rate The accuracy is ±0.01 Hz / s obtained through first-order difference calculation of frequency data; converter-side data includes output active power. The accuracy is ±0.1MW, and the output reactive power is used to assist in judging the grid operating conditions and does not participate in parameter calculation; the energy storage side data includes real-time SOC value with an accuracy of ±0.01, and energy storage battery terminal voltage is used to monitor the energy storage status and avoid battery failure.

[0048] Data preprocessing includes two steps: noise filtering and data smoothing. Noise filtering uses a first-order low-pass filter to remove high-frequency noise; the filter transfer function is... The filtering time constant The time constant is 0.02s. Simulations verify that this time constant can effectively filter noise without affecting the dynamic characteristics of the data. The filtering formula is: ; in, , The sampling period is 0.01s. This is the original data at the current moment. This is the filtered data. This is the filtered data from the previous time step. Data smoothing uses a sliding window method with a window length of 0.1s (10 sampling points). This smooths the filtered data to avoid parameter misadjustment caused by sudden changes in single-point data. The smoothing formula is: ; in, The window length is 10. This is the smoothed data.

[0049] Then, disturbance type and intensity are identified. Based on preprocessed real-time data, multi-feature joint judgment is used to accurately identify the type and intensity of power grid disturbances, providing a basis for differentiated correction strategies and avoiding a one-size-fits-all approach. Disturbance duration. It is an indicator for determining the type of disturbance, and the calculation formula is: ; in, The time of the disturbance is determined by frequency mutation detection, i.e., when... At Hz, determine for , This is the current sampling point number. The sampling period is 0.01s.

[0050] Disturbance type identification uses binary classification logic. A sudden load change disturbance must simultaneously meet three conditions: First, If the frequency is greater than the first frequency threshold (in this embodiment, the first frequency threshold is 0.05Hz), it indicates a significant frequency change. second, The frequency change rate is greater than the first rate of change threshold. In this embodiment, the first rate of change threshold is 0.2 Hz / s, which indicates a large frequency change rate. third, If the duration is less than the preset time threshold, which is 0.5s in this embodiment, it indicates that the disturbance duration is short.

[0051] Typical scenarios for this type of disturbance include sudden factory startup and sudden application of large loads, requiring enhanced inertia support and damping suppression.

[0052] New energy fluctuations and disturbances must simultaneously meet three conditions: First, Within the second frequency threshold range, which in this embodiment is 0.01Hz to 0.05Hz, it indicates that the frequency fluctuation is smooth; second, Less than or equal to the second rate of change threshold, which in this embodiment is 0.2 Hz / s, indicating a small rate of frequency change; third, A time value greater than or equal to the preset time threshold indicates a long duration of disturbance.

[0053] Typical scenarios for this type of disturbance include slow changes in photovoltaic irradiance and random fluctuations in wind power, requiring weakening of inertia and strengthening of droop adjustment.

[0054] Disturbance intensity coefficient The formula used to quantize the amplitude of the disturbance is as follows: ; in, As of the current moment The peak frequency deviation is 50Hz, which is the rated frequency. If the strength exceeds a preset threshold, it is determined to be a strong disturbance; otherwise, it is determined to be a weak disturbance. In this embodiment, the preset threshold is 0.2. The disturbance strength coefficient directly determines the magnitude of parameter correction. The correction magnitude is larger under strong disturbance and smaller under weak disturbance, thus avoiding overcorrection.

[0055] Next, the real-time SOC adaptive coefficient is calculated. This is based on the real-time SOC smoothing value. calculate To ensure the safe operation of energy storage, the formula is: ; The specific characteristics are as follows: When hour, The converter participates in frequency regulation at maximum capacity; when or hour, The converter maintains a high frequency regulation capability; when or hour, The converter is derated for frequency regulation to avoid overcharging or over-discharging. This coefficient is updated in real time to ensure dynamic matching between energy storage status and frequency regulation capability.

[0056] Finally, differentiated dynamic correction control parameters are applied. This is based on the disturbance type, disturbance intensity, and... Using the initial parameters from step two as a benchmark, a differentiated correction strategy is adopted to ensure that the parameters match the current working conditions. During the correction process, the parameters must be strictly limited to not exceed the basic range of step one.

[0057] When the disturbance type is a load change disturbance, the correction logic is that the load change causes a frequency change. Therefore, it is necessary to prioritize strengthening the virtual inertia to suppress the frequency change and the damping coefficient to accelerate the oscillation decay, while appropriately increasing the droop coefficient to ensure the accuracy of static frequency recovery. The correction formula is as follows: ; ; ; in, and This indicates that the correction magnitude is positively correlated with the disturbance intensity; the correction magnitude is larger under strong disturbances and smaller under weak disturbances. and It is used to dynamically adjust the correction amplitude according to the real-time frequency, ensuring that the parameters match the frequency changes in real time.

[0058] When the disturbance type is a new energy fluctuation disturbance, the correction logic is that the new energy fluctuation causes the frequency to be flat but continuously changing. Strong inertia support is not required to avoid over-response leading to new oscillations. Therefore, priority should be given to strengthening the droop adjustment to correct the static frequency deviation, while appropriately reducing the damping coefficient to ensure response flexibility. The correction formula is: ; ; ; in, This indicates that the droop coefficient correction amplitude is positively correlated with the real-time frequency deviation. The larger the frequency deviation, the stronger the droop adjustment capability, ensuring the accuracy of static frequency recovery.

[0059] In this embodiment, parameter correction is performed under a simulated load increase of 30MW. The PMU acquires the real-time frequency after the disturbance. The frequency deviation is 49.91Hz. The frequency is -0.09Hz, after filtering and smoothing. 0.088 Hz; frequency change rate -0.35Hz / s, after smoothing The frequency is 0.34 Hz / s; the real-time SOC is 0.55, after smoothing. It is 0.54.

[0060] The disturbance identification result is as follows: Hz, Hz / s, duration of disturbance s, determined to be a sudden load disturbance; disturbance intensity coefficient ,because It was determined to be a weak disturbance.

[0061] SOC adaptive coefficient .

[0062] The corrected parameters are calculated as follows:

[0063] ; ; ; ; ; ; ; ; The core value of this step lies in achieving dynamic matching between control parameters and grid conditions and energy storage status through a closed-loop process of real-time data acquisition, operating condition identification, and differentiated correction. This solves the problem of poor adaptability of fixed parameters and ensures that optimal frequency regulation can be achieved under different disturbance scenarios.

[0064] Step 4: Closed-loop verification and parameter feedback adjustment.

[0065] This step is the last line of defense to ensure the validity of the parameters. It verifies the frequency tuning effect of the parameters through quantitative indicators. If the requirements are not met, feedback is sent to step three for recalibration, forming a complete closed loop of calibration, verification, and recalibration, ensuring that the final output parameters have engineering applicability. Its inputs are the real-time parameters from step three and the operating data collected by the PMU; the output is the final effective control parameters.

[0066] First, a comprehensive verification indicator system is established. Four types of quantifiable and operable verification indicators are set, and qualified thresholds are defined to comprehensively evaluate the frequency regulation effect, energy storage safety, and system stability of the parameters.

[0067] Indicator 1 is the frequency deviation correction time. Defined as from the moment the disturbance occurs The time it takes for the absolute value of the real-time frequency deviation to remain no greater than 0.02Hz for three sampling periods (0.03s), measured in seconds, is considered acceptable. The time limit is no greater than the first time threshold. In this embodiment, the first time threshold is set to 1.5s. Its core significance is to measure the speed of frequency recovery, ensure that the power grid can recover stability in a short time, and avoid the frequency from deviating from the rated value for a long time.

[0068] Indicator 2 is the power oscillation attenuation rate. The formula is defined as follows ,in This represents the peak value of the first power oscillation after the disturbance. This is the second power oscillation peak, and the acceptable threshold is... The value should not be less than a preset attenuation threshold. In this embodiment, the preset attenuation threshold is 0.75. The core significance of this threshold is to measure the effectiveness of oscillation suppression, ensure power fluctuation attenuation, and avoid secondary impact on the power grid.

[0069] Indicator 3 is the rate of change of energy storage SOC. , defined as the change in SOC per unit time, is calculated using the following formula: The unit is 1 / s, where 3s is a complete frequency modulation cycle, and the acceptable threshold is... The rate should not exceed a preset rate threshold. In this embodiment, the preset rate threshold is 0.08 / s. The core purpose is to measure the smoothness of energy storage charging and discharging, and to avoid the degradation of energy storage battery life caused by changes in SOC.

[0070] Indicator 4 is the smoothness of parameter adjustment. , defined as the ratio of the change in a parameter at adjacent time points to the maximum value of that parameter, is calculated using the following formula: ,in For any control parameter, This is the upper limit of the basic range for this parameter, and the acceptable threshold is... The value should not exceed the preset smoothing threshold. In this embodiment, the preset smoothing threshold is set to 0.1. The core meaning is to measure the smoothness of parameter adjustment and avoid sudden parameter changes that could cause a surge in converter output power and trigger new frequency fluctuations.

[0071] Then, strictly implement the real-time verification process. The verification process must be carried out simultaneously with parameter calibration to ensure real-time performance.

[0072] The first step is parameter distribution and data acquisition, which involves sending the real-time parameters corrected in step three. , , The data is sent to the energy storage converter controller, which executes frequency regulation control according to the parameters. At the same time, the PMU continuously collects frequency, power, and SOC data during the frequency regulation process. The collection time is 3 seconds after the disturbance occurs, covering the complete frequency regulation process, including the inertial response, damping response, and droop response stages. The data storage format is timestamp, frequency, power, and SOC, which is convenient for subsequent index calculation.

[0073] The second step is indicator calculation and data processing. After data collection, the following procedures are followed to calculate four types of verification indicators: frequency deviation correction time. By iterating through the frequency data within 3 seconds, we find the first instance that satisfies three consecutive sampling periods. Hz time ,but Power oscillation attenuation rate The peak value of the first oscillation was identified by extracting power data within 3 seconds. That is, the first maximum or minimum power value reached after the disturbance and the second oscillation peak value. That is, the next peak after the first peak, which is then substituted into the formula for calculation. SOC change rate By extracting the time of disturbance occurrence SOC value and SOC value after 3 seconds Substitute into the formula to calculate; adjust parameters for smoothness. By iterating through the parameter data within 3 seconds, the ratio of the parameter change to the upper limit of the parameter at each adjacent time point is calculated, and the maximum value is taken as the result. .

[0074] The third step is threshold comparison and result judgment. The actual values ​​of the four types of indicators are compared with the qualified thresholds one by one. If all four types of indicators meet the qualified thresholds, it means that the current parameters match the operating conditions, and the frequency regulation effect, energy storage safety and system stability all meet the standards. The parameter is directly used as the final effective parameter and continues to be executed until the operating conditions change. If any indicator does not meet the qualified threshold, it means that there is room for optimization of the current parameters. The feedback adjustment stage is immediately entered to avoid the risk to the power grid or energy storage equipment caused by the continuous operation of unqualified parameters.

[0075] Finally, targeted feedback and adjustments are made. The core principle of feedback and adjustment is to correct the unmet indicators without affecting other compliant indicators. The adjusted parameters need to be re-submitted into the verification process until all indicators meet the standards.

[0076] The specific adjustment rules are as follows: If 's' indicates that the frequency recovery is too slow. The problem is caused by insufficient droop coefficient (i.e., weak static recovery capability) or insufficient inertia coefficient (i.e., weak dynamic suppression capability). The adjustment measure is to increase the droop coefficient... The static recovery capability is enhanced by increasing the first adjustment factor to the original value. In this embodiment, the first adjustment factor is taken as 1.2, and the virtual inertia coefficient is also increased. The second adjustment coefficient, increased to the original value, enhances the dynamic suppression capability. In this embodiment, the second adjustment coefficient is 1.1. After adjustment, it must be ensured that the parameters do not exceed the basic range of step one; if This indicates insufficient oscillation damping. The problem is that the damping coefficient is too small to effectively dampen the oscillation. The adjustment measure is to increase the virtual damping coefficient. Increase to a third adjustment factor multiple of the original value. In this embodiment, the third adjustment factor is 1.3, and it remains unchanged. and No change, adjustments must ensure ;like / s indicates that the SOC is changing too rapidly. The problem is caused by an excessively large inertia coefficient or droop coefficient, resulting in excessive energy storage charging and discharging power. The adjustment measure is to reduce... and The value is reduced to a fourth adjustment factor multiple of the original value. In this embodiment, the fourth adjustment factor is 0.7. This reduces the energy storage charging and discharging power. After adjustment, it is necessary to ensure that the parameters are not lower than the basic range in step one. The error indicates a sudden parameter change. The cause is excessive parameter correction, leading to drastic parameter variations between adjacent time points. The adjustment measure is to smooth the parameters using a first-order low-pass filter with a filtering time constant of 0.02s. The filtering formula is the same as in step three, ensuring smooth parameter changes while maintaining dynamic adaptability. The adjusted parameters are then reissued to the controller, and the PMU re-acquires 3s of data. The index calculations and threshold comparisons are repeated until all indices meet the acceptable thresholds. The parameters at this point are considered the final effective parameters and are used for stable execution.

[0077] In this embodiment, after the PMU acquires 3 seconds of data, the frequency deviation correction time is calculated. It is 1.2s, which satisfies S meets the standard; power oscillation attenuation rate ,satisfy Meets the standard; SOC change rate / s, satisfying / s meets the standard; parameter adjustment smoothness The value is 0.07, which satisfies the condition. Meets the standards. All indicators meet the acceptable thresholds, and the parameters take effect immediately and continue to be executed until the operating conditions change.

[0078] In summary, the embodiments disclosed herein have at least the following technical effects: This invention deeply integrates virtual inertia, damping, droop parameters with SOC adaptive adjustment, and achieves synergistic matching of the three through a multi-objective optimization model. This solves the response imbalance problem caused by parameter coupling in the prior art, significantly improves frequency modulation dynamic performance, and can control the frequency deviation correction time within 1.5s, while achieving a power oscillation attenuation rate of over 0.75.

[0079] This invention is based on a differentiated correction strategy for disturbance type and intensity, which can adapt to complex operating conditions such as load changes and new energy fluctuations. Compared with fixed parameter methods, the frequency regulation response speed is improved by more than 30%, which can meet the diverse frequency regulation needs of grids with a high proportion of new energy connected to the grid.

[0080] This invention introduces an adaptive SOC coefficient and a charge / discharge rate constraint, strictly controlling the SOC change rate to within 0.08 / s, avoiding overcharging and discharging of energy storage, extending equipment life by more than 30%, and ensuring that energy storage can continuously participate in frequency regulation.

[0081] The parameter calculation of this invention is based on measurable physical quantities in engineering, namely frequency, SOC, and power. It does not rely on complex global system models, has low algorithm complexity, can be embedded in the energy storage converter controller, requires no additional hardware investment, is compatible with various energy storage carriers such as lithium batteries, flow batteries, and flywheels, and is easy to promote in engineering.

[0082] This invention ensures that every effective parameter undergoes rigorous performance testing through closed-loop verification and feedback adjustment of four types of quantitative indicators, thereby avoiding risks to the power grid or energy storage equipment caused by unqualified parameters and improving the reliability and stability of the entire method.

[0083] It is understood that the above embodiments are merely exemplary implementations used to illustrate the principles of the present invention, and the present invention is not limited thereto. For those skilled in the art, various modifications and improvements can be made without departing from the spirit and essence of the present invention, and these modifications and improvements are also considered to be within the scope of protection of the present invention.

Claims

1. A method for setting control parameters of an energy storage converter for frequency regulation, characterized in that, include: Obtain the hardware parameters of the energy storage converter and the system safety standard parameters. Calculate the basic range upper and lower limits of the virtual inertia coefficient, virtual damping coefficient, and droop coefficient based on engineering constraints. The hardware parameters of the energy storage converter include rated active power, inherent response time, and maximum charge / discharge power. The system safety standard parameters include the maximum allowable frequency change rate, allowable frequency deviation, frequency dead zone, and allowable frequency overshoot. With the optimization objectives of minimizing the peak frequency deviation, minimizing the power oscillation amplitude, and maximizing the energy storage utilization efficiency, an SOC adaptive coefficient is introduced to construct an output power model of the energy storage converter that integrates virtual inertia control, damping control, and droop control. The initial control parameters are solved by an optimization algorithm within the stated range. Real-time grid frequency, frequency change rate, and energy storage SOC data are collected and filtered and smoothed. Based on the preprocessed data, the disturbance type and disturbance intensity are identified by frequency deviation, frequency change rate, and disturbance duration. According to the disturbance type, a differentiated correction strategy is adopted to dynamically correct the real-time control parameters based on the initial control parameters to obtain the real-time control parameters.

2. The method for setting control parameters of an energy storage converter for frequency regulation according to claim 1, characterized in that, The formula for calculating the upper limit of the basic range of the virtual inertia coefficient is as follows: ； in, This represents the upper limit of the virtual inertia coefficient. The rated active power of the energy storage converter, This is the inherent response time of the converter. The system's reference angular frequency, This represents the maximum permissible rate of change of the system's frequency. The lower limit of the basic range of the virtual inertia coefficient The upper limit The preset first proportional coefficient is times.

3. The method for setting control parameters of an energy storage converter for frequency regulation according to claim 1, characterized in that, The formula for calculating the upper limit of the basic range of the virtual damping coefficient is as follows: ； in, This represents the upper limit of the virtual damping coefficient. This represents the maximum charge and discharge power of the energy storage converter. To allow for frequency overshoot, This is a frequency dead zone; The lower limit of the basic range of the virtual damping coefficient The upper limit The preset second proportional coefficient multiple; The formula for calculating the upper limit of the basic range of the droop coefficient is as follows: ； in, This is the upper limit of the droop coefficient. This represents the maximum value of the SOC adaptive coefficient. This refers to the allowable value for frequency deviation. The lower limit of the basic range of the droop coefficient The upper limit The preset third proportional coefficient multiple.

4. The method for setting control parameters of an energy storage converter for frequency regulation according to claim 1, characterized in that, The calculation formula for the output power model of the energy storage converter is as follows: ； in, To output active power to the energy storage converter, For virtual inertia coefficient, The rate of change of frequency, This is the virtual damping coefficient. For frequency deviation, The droop coefficient is... These are the SOC adaptive coefficients; The formula for calculating the SOC adaptive coefficient is as follows: ； in, For the energy storage state of charge, the SOC adaptive coefficient ranges from 0.4 to 1.

2.

5. The method for setting control parameters of an energy storage converter for frequency regulation according to claim 1, characterized in that, The process of solving the initial control parameters using an optimization algorithm includes: Construct a fitness function, the formula for which the fitness function is calculated is: ； in, For the fitness function value, , , These are the peak frequency deviation weight, the power oscillation amplitude weight, and the energy storage utilization efficiency weight, respectively, and the weights satisfy... , This represents the peak value of the frequency deviation. This represents the amplitude of the power oscillation. The particle swarm optimization algorithm is used to iteratively solve the problem within the basic range. When the change in the fitness function value of consecutive preset algebras is less than a preset convergence threshold or the preset maximum number of iterations is reached, the initial control parameters are output.

6. The method for setting control parameters of an energy storage converter for frequency regulation according to claim 1, characterized in that, The filtering and smoothing preprocessing includes: A first-order low-pass filter is used to filter noise from the acquired data. The filtering formula is as follows: ； in, , The sampling period is The filtering time constant is This is the original data at the current moment. This is the filtered data. This is the filtered data from the previous time step; The sliding window method is used to smooth the filtered data. The smoothing formula is as follows: ； in, For window length, This is the smoothed data.

7. The method for setting control parameters of an energy storage converter for frequency regulation according to claim 1, characterized in that, The identification of disturbance type and disturbance intensity includes: Calculate the duration of the disturbance The formula for calculating the duration of the disturbance is: ； in, This is the current sampling point number. The sampling period is The time of disturbance occurrence is determined by frequency mutation detection; When the absolute value of the smoothed frequency deviation is greater than the first frequency threshold, the absolute value of the smoothed frequency change rate is greater than the first change rate threshold, and the duration of the disturbance is less than the preset time threshold, it is determined to be a load change disturbance. When the absolute value of the smoothed frequency deviation is within the range of the second frequency threshold, the absolute value of the smoothed frequency change rate is less than or equal to the second change rate threshold, and the duration of the disturbance is greater than or equal to the preset time threshold, it is determined to be a new energy fluctuation disturbance. Calculate the disturbance intensity coefficient The formula for calculating the disturbance intensity coefficient is as follows: ； in, This represents the peak frequency deviation up to the current moment. When the disturbance intensity coefficient is greater than the preset intensity threshold, it is determined to be a strong disturbance; otherwise, it is determined to be a weak disturbance.

8. The method for setting control parameters of an energy storage converter for frequency regulation according to claim 7, characterized in that, When the disturbance type is the aforementioned load mutation disturbance, the correction formula for the differentiated correction strategy is: ； ； ； in, , , These are the corrected virtual inertia coefficient, virtual damping coefficient, and droop coefficient, respectively. , , These are the initial virtual inertia coefficient, the initial virtual damping coefficient, and the initial droop coefficient, respectively. The smoothed rate of change of frequency. The smoothed frequency deviation, For real-time SOC adaptive coefficients.

9. The method for setting control parameters of an energy storage converter for frequency regulation according to claim 7, characterized in that, When the disturbance type is the aforementioned new energy fluctuation disturbance, the correction formula for the differentiated correction strategy is: ； ； ； in, This is the lower limit of the virtual inertia coefficient. This represents the lower limit of the virtual damping coefficient.

10. The method for setting control parameters of an energy storage converter for frequency regulation according to any one of claims 1 to 9, characterized in that, Also includes: A verification index system is established to perform closed-loop verification of the real-time control parameters. If all the verification indicators meet the preset threshold, the real-time control parameters are used as the final effective parameters. Otherwise, targeted feedback adjustments are made according to the type of indicators that do not meet the threshold until the preset threshold is met. The verification index system includes frequency deviation correction time. Power oscillation attenuation rate SOC change rate And parameter adjustment smoothness ; The formula for calculating the power oscillation attenuation rate is as follows: ； in, This represents the peak value of the first power oscillation after the disturbance. This is the peak value of the second power oscillation; The formula for calculating the rate of change of SOC is: ； in, For a complete frequency modulation cycle, The SOC value at the moment the disturbance occurs. The SOC value at the end of the frequency modulation cycle; The feedback adjustments include: When the frequency deviation correction time is greater than the preset correction time threshold, the droop coefficient is increased to a first adjustment factor multiple of the original value and the virtual inertia coefficient is increased to a second adjustment factor multiple of the original value. When the power oscillation attenuation rate is less than the preset attenuation threshold, the virtual damping coefficient is increased to the third adjustment factor multiple of the original value; When the absolute value of the SOC change rate is greater than the preset rate threshold, the virtual inertia coefficient and the droop coefficient are reduced to a fourth adjustment factor multiple of their original values. When the smoothness of the parameter adjustment is greater than the preset smoothing threshold, the parameter is smoothed by low-pass filtering.

Images