System safety assessment method and system for energy storage power station containing network-forming type energy storage converter

Abstract

The invention discloses a system safety assessment method and system for an energy storage power station containing a network-building type energy storage converter, and the method comprises the following steps: S1, collecting converter control parameters, power grid dynamic data and lithium battery mechanism data through multi-source heterogeneous data fusion, and carrying out the preprocessing through noise elimination and a normalization algorithm; s2, constructing a network construction characteristic quantitative evaluation system including indexes such as dynamic stability and voltage supporting capability; s3, establishing a three-layer evaluation system of an equipment layer, a control layer and a system layer, and calculating a total risk value of the system by using a weight distribution algorithm; s4, verifying system adaptability and optimizing parameters through time domain simulation in scenes such as a strong power grid and a weak power grid; and S5, generating an evaluation report containing parameter optimization suggestions, triggering graded early warning based on an index threshold value, and automatically generating a closed-loop control instruction. According to the invention, multi-dimensional safety assessment of the energy storage power station is realized, and systematic technical support is provided for safe operation of the network-forming type energy storage power station.

Description

TECHNICAL FIELD

[0001] The present application relates to energy storage power station safety evaluation technology, and in particular to a system safety evaluation method and system for an energy storage power station containing a grid-forming energy storage converter. BACKGROUND

[0002] With the continuous increase of new energy power generation proportion, the grid-forming energy storage converter as the core equipment to support the stability of the power grid, its safe operation directly affects the reliability of the power system. The existing energy storage power station safety evaluation technology has the following core defects: first, the dynamic characteristic evaluation is missing. The traditional method does not consider the influence of dynamic parameters such as moment of inertia and damping coefficient of the grid-forming converter on the stability of the system. CN202411714279 points out that a 200MW energy storage power station has a power oscillation peak of ±15% of the rated power under grid disturbance due to the non-optimization of the moment of inertia parameter, and the oscillation lasts for 1.2 seconds, resulting in a power grid frequency deviation exceeding 0.5Hz. Second, the LVRT scene evaluation is insufficient. The power angle stability of the flow energy storage grid-forming system in the low voltage ride through (LVRT) lacks a unified evaluation standard. CN202410987516 shows that if the current feedback mechanism is not introduced when the voltage drops to 0.5p.u., the power angle overshoot can reach 68°, and the existing technology does not include such dynamic response in the safety evaluation system. Third, the mechanism characteristics of lithium batteries are ignored. The traditional evaluation does not establish a correlation model between the aging characteristics of lithium batteries and system safety. CN202411010266 points out that the scheduling strategy without considering the mechanism characteristics such as lithium precipitation rate will cause the battery aging recession to be 38.8% higher than the optimized strategy. Fourth, the grid adaptability evaluation is weak. In a high proportion of new energy power grid, the improvement effect of the grid-forming energy storage on the short circuit ratio (SCR) lacks a quantitative method. CN202510220856 shows that the unoptimized energy storage configuration may result in an insufficient SCR improvement of less than 20%, which cannot effectively improve the stability of the weak power grid. Fifth, the multi-physical field coupling analysis is insufficient: the existing technology does not realize the coordinated evaluation of the converter control parameters, the dynamic characteristics of the power grid and the mechanism characteristics of the battery, and it is difficult to find the systemic risk caused by the coupling of multiple factors.

[0003] It should be noted that the information disclosed in the above background section is only for understanding the background of the present application, and therefore can include information that does not constitute the prior art known to those of ordinary skill in the art. SUMMARY

[0004] The main purpose of the present application is to overcome the defects in the background art, and to provide a system safety evaluation method and system for an energy storage power station containing a grid-forming energy storage converter.

[0005] To achieve the above-mentioned purpose, the present application adopts the following technical solutions:

[0006] In the first aspect of the application, a system safety evaluation method for a grid-forming energy storage converter energy storage power station comprises the following steps:

[0007] S1. Multi-source heterogeneous data fusion collection: real-time collection of control parameters, grid dynamic data and lithium battery mechanism data of the grid-forming energy storage converter, and preprocessing through noise elimination and normalization algorithm;

[0008] S2. Construction of grid characteristic quantitative evaluation index system: establishment of dynamic stability index, voltage support capability index, short circuit ratio improvement effect index and lithium battery aging safety index to quantify the multi-dimensional safety state of the system;

[0009] S3. Multi-physical field coupling risk evaluation model construction: establishment of a three-layer evaluation system of equipment layer, control layer and system layer, calculation of normalized risk values of each layer through weight distribution algorithm, and weighted summation to obtain the total risk value of the system;

[0010] S4. Scenario simulation verification and optimization: setting of strong grid, weak grid, high proportion of new energy and low voltage ride-through scenarios, verification of system adaptability through time domain simulation and optimization of control parameters;

[0011] S5. Safety evaluation report generation and intelligent early warning: generation of an evaluation report containing parameter optimization suggestions, triggering of graded early warning based on index threshold and automatic generation of closed-loop control instructions.

[0012] In the second aspect of the application, a system safety evaluation system for a grid-forming energy storage converter energy storage power station comprises:

[0013] Multi-source data acquisition and preprocessing module: it collects control parameters, grid dynamic data and lithium battery mechanism data of the grid-forming energy storage converter in real time, and preprocesses them through noise elimination and normalization algorithm;

[0014] Grid characteristic quantitative index calculation module: it establishes dynamic stability index, voltage support capability index, short circuit ratio improvement effect index and lithium battery aging safety index to quantify the multi-dimensional safety state of the system;

[0015] Multi-physical field coupling risk evaluation module: it establishes a three-layer evaluation system of equipment layer, control layer and system layer, calculates normalized risk values of each layer through weight distribution algorithm, and weighted summation to obtain the total risk value of the system;

[0016] Scenario simulation verification and optimization module: it sets strong grid, weak grid, high proportion of new energy and low voltage ride-through scenarios, verifies the system adaptability through time domain simulation and optimizes the control parameters;

[0017] Evaluation report and intelligent early warning module: it generates an evaluation report containing parameter optimization suggestions, triggers graded early warning based on index threshold and automatically generates closed-loop control instructions.

[0018] The present application has the following beneficial effects:

[0019] The present application provides a system safety evaluation method and system for a grid-forming energy storage converter energy storage power station, which can perform multi-dimensional safety evaluation on the dynamic characteristics of the grid-forming energy storage converter, the mechanism characteristics of lithium batteries, and the interaction with the power grid. Through multi-source data fusion acquisition, grid characteristic quantitative index construction, multi-physical field coupling risk evaluation model establishment, scenario simulation verification, and intelligent early warning mechanism, the present application can realize multi-dimensional and all-around safety evaluation of the energy storage power station. Experimental data show that the power oscillation peak value of the grid-forming converter based on neural network optimization can be reduced by 66.7%, and the oscillation duration is shortened by 65%; the scheduling strategy considering the mechanism of lithium batteries can reduce the aging degradation amount by 38.8%, and the system risk value is reduced by 38.1%. The method of the present application fills the gap in dynamic characteristic evaluation, lithium battery mechanism fusion, and power grid adaptability evaluation in the prior art, and provides systematic technical support for the safe operation of the grid-forming energy storage power station.

[0020] Other beneficial effects of the embodiments of the present application will be further described below. BRIEF DESCRIPTION OF DRAWINGS

[0021] Figure 1 The present application provides a system safety evaluation method for a grid-forming energy storage converter energy storage power station.

[0022] Figure 2 The present application provides a system safety evaluation method for a grid-forming energy storage converter energy storage power station. DETAILED DESCRIPTION

[0023] The embodiments of the present application will be described in detail below. It should be emphasized that the following description is only exemplary and is not intended to limit the scope of the present application and its applications.

[0024] The present application aims to overcome the deficiencies of the existing safety evaluation technology for energy storage power stations in dynamic characteristic evaluation, lithium battery mechanism fusion, and power grid adaptability evaluation, and provides a system safety evaluation method for a grid-forming energy storage converter energy storage power station. The method realizes multi-dimensional safety evaluation of the energy storage power station through the synergistic effect of multi-source heterogeneous data fusion acquisition, grid characteristic quantitative evaluation index system construction, multi-physical field coupling risk evaluation model construction, scenario simulation verification and optimization, and safety evaluation report generation and intelligent early warning, effectively reduces the power oscillation peak value and oscillation duration of the grid-forming converter, reduces the aging degradation amount of the lithium battery, and further reduces the system risk, thereby ensuring the safe and stable operation of the energy storage power station.

[0025] Reference Figure 1The embodiment of the present application provides a kind of system safety evaluation method of energy storage power station containing network type energy storage converter, comprising the following steps:

[0026] Step S1. Multi-source heterogeneous data fusion acquisition: real-time acquisition of control parameters, power grid dynamic data and lithium battery mechanism data of network type energy storage converter, and preprocessing by noise elimination and normalization algorithm.

[0027] In some embodiments, the preprocessing in step S1 includes: for high-frequency sampled converter control parameters, using state space modeling filtering algorithm to eliminate Gaussian noise through prediction-update iteration; for lithium battery mechanism data, using range normalization algorithm to map parameters to a unified interval.

[0028] Step S2. Network characteristic quantitative evaluation index system construction: dynamic stability index, voltage support capability index, short circuit ratio improvement effect index and lithium battery aging safety index are established to quantify the multi-dimensional safety state of the system.

[0029] In some embodiments, in step S2: the dynamic stability index is related to the moment of inertia, damping coefficient and grid frequency deviation of the converter; the voltage support capability index is related to the virtual impedance of the converter, the voltage recovery characteristics of the power grid and the power angle stability; the lithium battery aging safety index accumulatively calculates the aging recession amount caused by the growth of solid electrolyte interface film, lithium extraction and active material loss.

[0030] In some embodiments, the calculation of the lithium battery aging safety index is based on an electrochemical mechanism model, and the cumulative aging recession amount is dynamically quantified by real-time acquisition of state of charge, lithium extraction rate and active material loss rate.

[0031] Step S3. Multi-physical field coupling risk evaluation model construction: a three-layer evaluation system of device layer, control layer and system layer is established, and the normalized risk values of each layer are calculated by weight distribution algorithm, and the system total risk value is obtained by weighted summation.

[0032] In some embodiments, in step S3: the device layer evaluates the temperature of converter power device and the balance degree of lithium battery state of charge; the control layer evaluates the control loop gain margin and phase angle margin through small signal stability analysis; the system layer evaluates the power angle stability margin and frequency recovery time through fault scenario time domain simulation.

[0033] In some embodiments, the weight distribution algorithm uses the analytic hierarchy process: the judgment matrix of device layer, control layer and system layer is constructed, and the relative importance is determined by pairwise comparison; the initial weight vector is calculated by eigenvector method, and the final weight matrix is determined after consistency test.

[0034] Step S4. Scenario simulation verification and optimization: Set up strong grid, weak grid, high proportion of new energy and low voltage ride through scenarios, verify system adaptability through time domain simulation and optimize control parameters;

[0035] In some embodiments, in step S4: the strong grid scenario verifies the frequency deviation suppression capability; the weak grid scenario verifies the short circuit ratio improvement capability; the low voltage ride through scenario verifies the voltage recovery time and power angle overshoot suppression capability.

[0036] Step S5. Safety evaluation report generation and intelligent early warning: Generate an evaluation report containing parameter optimization suggestions, trigger a hierarchical early warning based on index thresholds and automatically generate closed-loop control instructions.

[0037] In some embodiments, the hierarchical early warning of step S5 includes: when the dynamic stability index or the voltage support capability index exceeds the threshold, triggering a secondary early warning and generating a control parameter adjustment instruction; when the lithium battery aging safety index exceeds the threshold, triggering a primary early warning and generating a dispatching strategy optimization instruction.

[0038] In some embodiments, the closed-loop control instruction execution process includes: identifying the core control parameters that need to be adjusted and determining the optimization range; generating parameter adjustment values based on optimization algorithms and issuing them in real time; re-collecting data to verify whether the indicators have returned to the safe range, and iterating optimization until the requirements are met when the indicators do not meet the requirements.

[0039] Referring to Figure 2 The embodiments of the present application also provide a system safety evaluation system for a grid-forming energy storage converter energy storage power station, comprising:

[0040] Multi-source data acquisition and preprocessing module: It acquires the control parameters of the grid-forming energy storage converter, the grid dynamic data and the lithium battery mechanism data in real time, and preprocesses them through noise elimination and normalization algorithm;

[0041] Grid characteristic quantitative index calculation module: It establishes dynamic stability index, voltage support capability index, short circuit ratio improvement effect index and lithium battery aging safety index to quantify the multi-dimensional safety state of the system;

[0042] Multi-physical field coupling risk assessment module: It establishes a three-layer evaluation system of device layer, control layer and system layer, calculates the normalized risk value of each layer through weight distribution algorithm, and obtains the total risk value of the system by weighted summation;

[0043] Scenario simulation verification and optimization module: It sets up strong grid, weak grid, high proportion of new energy and low voltage ride through scenarios, verifies the system adaptability through time domain simulation and optimizes the control parameters;

[0044] The evaluation report and intelligent early warning module generate an evaluation report containing parameter optimization suggestions, trigger a hierarchical early warning based on index threshold values, and automatically generate closed-loop control instructions.

[0045] The core defects of the existing energy storage power station safety evaluation technology are addressed by the innovative multi-dimensional collaborative evaluation system. By integrating network-type converter dynamic parameters (moment of inertia, damping coefficient), lithium battery aging mechanism characteristics (lithium extraction rate, active material loss), and grid interaction characteristics (short circuit ratio, power angle stability), the system systematically addresses issues such as the lack of dynamic characteristic evaluation in traditional methods, the neglect of LVRT scene response, the weak correlation of battery aging, and the insufficient quantification of grid adaptability. An innovative multi-dimensional quantitative evaluation system is constructed, including dynamic stability indicators and voltage support capacity indicators. Combined with a three-layer coupled risk evaluation model at the device layer, control layer, and system layer, the collaborative analysis of converter control parameters, grid dynamic response, and battery aging state is achieved. Based on time-domain simulation verification under strong / weak grid, high proportion of new energy, and low voltage ride-through scenarios, the system stability is significantly improved (power oscillation peak reduced by 66.7%, oscillation duration shortened by 65%) by real-time calibration of parameters and generation of closed-loop control instructions using optimization algorithms. By dynamically correlating the aging degradation amount and scheduling strategy through a lithium battery mechanism model, battery aging is effectively suppressed (degradation amount reduced by 38.8%) and system risk is reduced (decrease of 38.1%), filling the gap in multi-physical field coupling evaluation and providing systematic technical support for the safe operation of network-type energy storage power stations.

[0046] The specific embodiments of the present application, algorithm examples, and experimental verification are further described below.

[0047] A system safety evaluation method for a network-type energy storage converter energy storage power station, comprising the following steps:

[0048] S1. Multi-source heterogeneous data fusion acquisition

[0049] Converter control parameters: real-time acquisition of moment of inertia J, damping coefficient D, virtual impedance Z, etc. control parameters, real-time calibration of parameter drift using optimization algorithms, and sampling frequency not less than 10 kHz. The optimization algorithm includes but is not limited to: gradient descent and its variants, momentum-enhanced algorithm, adaptive learning rate algorithm, bionic and swarm intelligence algorithm, neural network structure optimization algorithm, and any one of the leading optimization strategies.

[0050] Grid dynamic data: monitoring of parameters such as short circuit ratio SCR, voltage drop depth, and frequency deviation at the access point, real-time updating of grid strength indicators based on the short circuit ratio calculation model, and calculation period of 100 ms.

[0051] Battery mechanism data: acquisition of lithium battery state of charge (SOC), lithium extraction rate j LP, active material loss rate, etc. parameters, establish an electrochemical model and a battery aging characteristic library, and the sampling interval is 5 minutes.

[0052] Data preprocessing: Kalman filtering algorithm is used for state space modeling of high-frequency sampling data, and Gaussian noise is eliminated through the prediction-update two-step; Min-Max normalization (range normalization) is used for battery mechanism data (5-minute interval), and SOC, lithium extraction rate and other parameters are mapped to the [0, 1] interval.

[0053] The specific implementation process of the Kalman filtering algorithm is as follows.

[0054] The state equation of Kalman filtering is x(k) = Ax(k-1) + Bu(k-1) + w(k-1), and the measurement equation is z(k) = Hx(k) + v(k). Wherein, x(k) represents the system state at time k, in this method, it specifically refers to the state of the high-frequency sampling network-forming converter control parameters; A represents the state transition matrix; B represents the control input matrix; x(k-1) represents the system state at time k-1, that is, the converter control parameter state at the previous moment, which is the basis for calculating the current state x(k); u(k-1) represents the control input at time k-1, which refers to the external control signal that affects the converter parameters; w(k-1) represents the process noise at time k-1, which is the random error introduced by system modeling or external disturbance, and obeys the Gaussian distribution with mean value 0 and covariance Q, used to represent the uncertainty in the state transition process; z(k) represents the observation value at time k, specifically refers to the high-frequency sampling data obtained in real time through sensors or data acquisition devices. H represents the measurement matrix, which is used to establish the mapping relationship between the system state and the observation value, and its dimension is determined by the number of system state variables and the number of observation values. v(k) is the measurement noise obeying Gaussian distribution with mean value 0 and covariance R.

[0055] Kalman filtering eliminates noise through "prediction-update" iteration, and the steps are as follows:

[0056] Prediction step: based on the state estimation at the last moment, the current state and error covariance are predicted.

[0057] State prediction:

[0058] The state prediction value at time k is based on the information at time k-1.

[0059] Error covariance prediction: P(k|k-1) = AP(k-1|k-1)A T +Q

[0060] P(k|k-1) is the error covariance matrix of the predicted state.

[0061] Update step: correct the prediction result with current observation value to get the optimal estimation.

[0062] Calculate Kalman gain: K(k) = P(k|k-1)HT[HP(k|k-1)H T +R] -1

[0063] K(k) is the weight coefficient of balancing the reliability of prediction and observation.

[0064] State update:

[0065] is the optimal state estimation after fusing the observation value at time k.

[0066] Error covariance update: P(k|k) = [I-K(k)H]P(k|k-1)

[0067] I is the identity matrix, and the updated covariance reflects the uncertainty of the optimal estimation.

[0068] The specific implementation of Min-Max normalization, its calculation formula is:

[0069] Where x is the original parameter value, x min and x max are the minimum and maximum values of the parameter in the sample set, and x' is the normalized value.

[0070] Application scenario: for example, when processing the state of charge (SOC) of lithium battery, if the original range of SOC is 20% ~ 80% (i.e. x min = 20%, x max = 80%), then when x = 50%, the normalized x' = (80-20) / (50-20) = 0.5.

[0071] S2. Construction of network characteristic quantitative evaluation index system

[0072] Dynamic stability index:

[0073] Where Δω max is the maximum frequency deviation (Hz), f n is the rated frequency (50 Hz), T J = J / P n is the moment of inertia time constant (s), Pn represents the rated power, T D = D / P n is the damping time constant (s), k z is the virtual impedance influence coefficient (take 0.2), and |Z| is the virtual impedance modulus (Ω).

[0074] Engineering application example: when the grid frequency deviation Δω max = 0.3 Hz, J = 500 kg·m 2 , D = 2000, virtual impedance Z = 0.2 Ω, it is calculated that

[0075] δ stab = 0.3 / 50×(500 / 100) / (2000 / 100)×(1+0.2×0.2) = 0.0012

[0076] Voltage support capability index:

[0077] Where, U rec is the voltage recovery value after fault (p.u.), U fault is the voltage drop value at fault (p.u.), δ max is the maximum swing of power angle (°). The support capability of grid-forming converter under voltage drop is quantified, and when V support < 0.9, the voltage support risk warning is triggered.

[0078] Short-circuit ratio improvement effect index:

[0079] Where, SCR before , SCR after are the short-circuit ratios before and after the energy storage is connected, respectively, and when SCR improve < 15%, it is evaluated as insufficient weak grid adaptability.

[0080] Lithium battery aging safety index:

[0081] Where, are the aging degradation amounts caused by solid electrolyte interface film growth, lithium precipitation, and active material loss at time t, respectively. Based on the lithium battery mechanism model, when τ L > 0.5% / month, it is determined as an aging acceleration risk. The specific measurement method of the three parameters is mainly based on electrochemical model simulation, combined with actual monitoring of lithium battery mechanism data (such as SOC, lithium precipitation rate) and aging characteristic library calibration, and finally the aging degradation amount is quantified.

[0082] S3. Multi-physical field coupling risk assessment model construction

[0083] Establish a three-layer evaluation system:

[0084] Device layer: evaluate the hardware status such as converter power device temperature (threshold ≤ 85℃), lithium battery SOC uniformity (deviation ≤ 5%).

[0085] Control layer: Evaluate the virtual synchronous control loop gain margin (require ≥ 6dB), phase margin (require ≥ 45°) by small signal stability analysis.

[0086] System layer: Evaluate the system power angle stability margin (require ≥ 15°), frequency recovery time (require ≤ 0.5s) by time domain simulation of three-phase short circuit, LVRT and other fault scenarios.

[0087] Risk quantification algorithm:

[0088] Where w i is the index weight (0.3 for equipment layer, 0.4 for control layer, 0.3 for system layer), r i is the normalized risk value (0-1), and the weight matrix is determined by the analytic hierarchy process (AHP).

[0089] 1. Establish a hierarchical model

[0090] Clearly define the evaluation objectives and hierarchical relationships:

[0091] Target layer: Determine the weight matrix of the equipment layer, control layer, and system layer, which is used to calculate the total system risk: R = ∑w i ·r i

[0092] Criteria layer: The three evaluation layers to be assigned weights, namely the equipment layer (evaluating hardware status such as converter temperature and battery SOC balance), the control layer (evaluating virtual synchronous control loop stability), and the system layer (evaluating power angle stability margin under fault scenarios).

[0093] 2. Construct the judgment matrix

[0094] Compare each element (equipment layer, control layer, system layer) of the criteria layer with each other through expert scoring, use the 1-9 scale method (1 represents equal importance, 9 represents extreme importance) to quantify the relative importance, and form the judgment matrix A, as shown in Table 1.

[0095] Table 1

[0096] Evaluation layer Device layer Control layer System layer Device layer 1 a 12 ]]> a 13 ]]> Control layer a 21 = 1 / a 12 ]] 1 a 23 ]]> System layer a 31 = 1 / a 13 ]] a 32 = 1 / a 23 ]] 1

[0097] The maximum weight is "equipment layer 0.3, control layer 0.4, system layer 0.3", and the judgment matrix is as follows:

[0098] The control layer is slightly more important than the equipment layer (e.g., a 21 = 1.33 means that the importance of the control layer is 1.33 times that of the equipment layer); the equipment layer and the system layer are equally important (a 13 = 1); and the control layer is slightly more important than the system layer (a 23 = 1.33).

[0099] 3. Calculate weight vector (eigenvector method)

[0100] 1) Normalization: sum each column element of the judgment matrix A, and divide each element by the sum of the column to obtain the normalized matrix

[0101] 2) Calculate weight: average each row element of to obtain the weight vector w=(w1, w2, w3) T (w1 is the device layer weight, w2 is the control layer weight, and w3 is the system layer weight).

[0102] 3) Consistency check:

[0103] Calculate the maximum eigenvalue λ max of the judgment matrix;

[0104] Calculate the consistency index

[0105] Check the table to get the average random consistency index RI=0.58 (when n=3);

[0106] Calculate the consistency ratio CR=RICI. When CR<0.1, the judgment matrix meets the consistency requirement, and the weight is valid.

[0107] The weight "0.3, 0.4, 0.3" meets the requirement of CR<0.1 after inspection, so it is used as the final weight matrix.

[0108] 4. Determine the final weight matrix

[0109] The device layer weight w1=0.3, the control layer weight w2=0.4, and the system layer weight w3=0.3 are used in the system risk quantification formula R=0.3r1+0.4r2+0.3r3 (r1, r2, r3 are the normalized risk values of the device layer, the control layer, and the system layer, respectively).

[0110] S4. Scenario simulation verification and optimization

[0111] Typical scenario settings:

[0112] Strong grid scenario (SCR=3.5): evaluate the inertia support effect of the converter, and require the maximum frequency deviation to be 0.2 Hz.

[0113] Weak grid scenario (SCR=1.8): verify the short circuit ratio improvement capability, and require SCR improve ≥25%.

[0114] High proportion of new energy scene (new energy proportion ≥ 70%): evaluate the frequency modulation effect of photovoltaic / wind power cooperation, require the frequency deviation recovery time ≤0.8s after joint frequency modulation, verify the inertia cooperation characteristics of grid type energy storage and new energy station through time domain simulation, and quantify the maximum frequency deviation and recovery characteristics.

[0115] LVRT scene (voltage drop to 0.5p.u.): verify the low voltage ride through capability of the converter, the LVRT control strategy of the flow energy storage, and require the voltage recovery time ≤1.0s and the power angle overshoot ≤30°, and inhibit the fault current through virtual impedance compensation and current feedback mechanism.

[0116] S5. Generation of safety evaluation report and intelligent early warning

[0117] Evaluation report content:

[0118] Converter parameter optimization suggestion: according to dynamic stability index δ stab Adjust the moment of inertia J, recommended range J ∈ [0.6Pn, 1.2Pn], where Pn is the rated power. Weak link identification of power grid: based on short circuit ratio improvement effect index, calculate the transmission capacity improvement value ΔP trans = 0.6·SCR improve ·P rated (rated power).

[0119] Intelligent early warning mechanism:

[0120] 1. When the dynamic stability index δ stab >0.003 or the voltage support capability index V support <0.9, trigger secondary warning, and automatically generate control parameter adjustment instructions.

[0121] Process of automatically generating control parameter adjustment instructions:

[0122] 1.1 Real-time index monitoring and threshold judgment

[0123] The system continuously collects and calculates the dynamic stability index δ stab and the voltage support capability index V support :

[0124] Dynamic stability index δ stab : based on the real-time collected moment of inertia J, damping coefficient D, virtual impedance Z and maximum frequency deviation Δω max of the converter, calculated according to the formula in real time (where fn is the rated frequency, T J =J / Pn, T D =D / Pn are time constants, and kz is the virtual impedance influence coefficient).

[0125] voltage support capability indicator V support : based on voltage drop value U fault , post-fault voltage recovery value U rec and power angle maximum swing δ max , according to the formula real-time calculation.

[0126] When δ stab > 0.003 or V support < 0.9 is monitored, a secondary warning is triggered, and an automatic control parameter adjustment process is started.

[0127] 1.2 Adjustment parameter identification and optimization range determination

[0128] According to the type of the indicator triggered by the warning, the core control parameters that need to be adjusted are locked:

[0129] If δ stab exceeds the standard (reflecting insufficient frequency dynamic stability): focus on adjusting the converter moment of inertia J, damping coefficient D and virtual impedance Z (all three are core influencing factors of δ stab ). The recommended optimization range of the moment of inertia J is J ∈ [0.6Pn, 1.2Pn], and the parameter drift is calibrated in real time through a neural network optimization algorithm (sampling frequency ≥ 10 kHz, ensuring timeliness of adjustment).

[0130] If V support exceeds the standard (reflecting insufficient voltage support capability): focus on adjusting the virtual impedance Z (affecting voltage recovery characteristics) and current feedback mechanism parameters (suppressing power angle overshoot during faults), and refer to the strategy of "virtual impedance compensation and current feedback to suppress fault current" in the LVRT scenario.

[0131] 1.3 Instruction generation and execution

[0132] Based on the parameter optimization range and the optimal value output by the neural network algorithm, the system automatically generates specific adjustment instructions:

[0133] The instruction content specifies the parameter adjustment value (such as "adjust the moment of inertia J from the current value to X kg·m 2 ", "correct the virtual impedance Z to Y Ω"), and attaches adjustment basis (such as "because δ stab = 0.0035 > 0.003, J needs to be improved to reduce frequency deviation").

[0134] The instruction is issued in real time through the converter control system, with an execution period matching the data sampling frequency (≤ 100 ms), ensuring rapid response to grid dynamic changes.

[0135] 1.4 Closed-loop verification and iterative optimization

[0136] After parameter adjustment, the system immediately reacquires data and calculates δ stab and V support , verify if the index returns to the safe range (δ stab ≤ 0.003 and V support ≥ 0.9):

[0137] If the index meets the standard, terminate the adjustment process;

[0138] If it does not meet the standard, based on the newly acquired parameter deviation data, repeat steps 2-3, further optimize the parameter value (such as fine-tune the damping coefficient D to enhance the damping effect) through the neural network algorithm, until the index meets the safety requirements, forming a closed-loop control.

[0139] 2. When the lithium battery aging safety index τ L > 0.8% per month, trigger a first-level early warning, adjust the scheduling strategy based on the mechanism model to suppress aging acceleration. The process is as follows:

[0140] 2.1 Mechanism model construction and parameter input

[0141] Core model: Establish a lithium battery electrochemical mechanism model, integrate the quantitative relationship of three major aging paths of solid electrolyte interface film (SEI) growth, lithium precipitation (LP), and active material (AM) loss, and clearly define the correlation formula between each factor (such as lithium precipitation rate j LP , state of charge SOC, and charge / discharge rate) and aging degradation.

[0142] Data input: Real-time acquisition of lithium battery mechanism data (sampling interval 5 minutes), including state of charge SOC, lithium precipitation rate j LP (unit A / cm 2 ), and active material loss rate, as dynamic input parameters for the model.

[0143] 2.2 Aging risk assessment and early warning triggering

[0144] Aging index calculation: Real-time calculation of lithium battery aging safety index τ L (cumulative aging degradation) based on the mechanism model, with the formula

[0145] wherein, are the aging degradation amounts caused by SEI film growth, lithium precipitation, and active material loss at time t, respectively.

[0146] Early warning criterion: When τ L > 0.8% per month (first-level early warning threshold), it is determined that there is a risk of accelerated aging, and the scheduling strategy adjustment process is started.

[0147] 2.2.1 Core measures of scheduling strategy adjustment

[0148] Identify the key aging inducers (such as excessive lithium precipitation rate) based on mechanism model recognition, and optimize the scheduling strategy accordingly:

[0149] (1) Limit the lithium precipitation rate: Through the mechanism model, the safe threshold of lithium precipitation rate j LP (such as j LP ≤0.012A / cm 2 ) is determined, and the charging and discharging strategy is adjusted (such as reducing the fast charging rate, optimizing the SOC working interval) to avoid excessive growth of lithium plating layer. For example, in Example 2, the mechanism optimization strategy controls j LP from 0.018A / cm 2 (conventional strategy) to 0.012A / cm 2 or less, directly reducing the aging caused by lithium precipitation.

[0150] (2) Dynamically adjust the inverter parameters: Combine the real-time correction of the battery aging state to modify the moment of inertia J of the grid-connected inverter (recommended J∈[0.6Pn, 1.2Pn]), balance the battery aging protection and system dynamic stability. For example, in Example 2, according to the aging state, J is dynamically adjusted from 350kg·m 2 to 420kg·m 2 , making the dynamic stability index δ stab improve by 40%.

[0151] (3) Optimize the charging and discharging timing: Based on the mechanism model, predict the aging rate under different working conditions (such as reducing the charging and discharging depth during high temperature period), develop peak shifting scheduling strategy, and reduce the active material loss rate.

[0152] 2.2.2 Effect verification and closed-loop iteration

[0153] (1) Index verification: After adjustment, continuously monitor τ L , j LP and system stability index (such as δ stab ), verify whether the aging degradation is reduced (such as in Example 2, the 30-day aging degradation is reduced from 0.85% to 0.52%, a decrease of 38.8%), and at the same time ensure that the system risk value R (weighted sum of device layer, control layer and system layer risk) is also reduced (such as from 0.21 to 0.13, a decrease of 38.1%).

[0154] (2) Model calibration: feedback the actual operation data to the mechanism model, update the aging characteristic library (such as the corresponding relationship between different j LP and ), continuously optimize the strategy parameters, and form a closed loop of "monitoring-evaluation-adjustment-verification".

[0155] A system safety evaluation system for a grid-connected energy storage inverter energy storage power station, which specifically comprises the following.

[0156] Multi-source data acquisition and preprocessing module

[0157] Responsible for real-time acquisition of three types of key data: network-forming energy storage converter control parameters (moment of inertia J, damping coefficient D, etc.), power grid dynamic data (short circuit ratio SCR, voltage drop depth, etc.), and lithium battery mechanism data (SOC, lithium extraction rate, etc.).

[0158] Calibrate converter parameter drift through neural network optimization algorithm, eliminate high-frequency data noise through Kalman filter, and process battery data through Min-Max normalization to provide high-quality data input for subsequent evaluation.

[0159] Network characteristic quantitative index calculation module

[0160] Based on the preprocessed data, calculate four core evaluation indicators: dynamic stability index (δ stab ), voltage support capability index (V support ), short circuit ratio improvement effect index (SCR improve ), and lithium battery aging safety index (τ L ) to quantify the safety status of each dimension of the system.

[0161] Multi-physical field coupling risk assessment module

[0162] Build a three-layer evaluation system of "device layer-control layer-system layer" to evaluate hardware status (such as converter temperature, battery SOC balance), control loop stability (such as gain margin, phase angle margin), and system-level fault response (such as power angle stability margin, frequency recovery time), and calculate the total risk of the system through risk quantification algorithm (R = Σ weight × normalized risk value).

[0163] Scenario simulation verification and optimization module

[0164] For typical scenarios such as strong grid, weak grid, high proportion of new energy, LVRT, etc., verify the adaptability of the system under different working conditions (such as inertia support effect, short circuit ratio improvement ability, low voltage ride-through performance) through time domain simulation, and optimize converter parameters (such as moment of inertia J) and scheduling strategies (such as lithium extraction rate control of lithium battery) based on simulation results.

[0165] Evaluation report and intelligent early warning module

[0166] Generate an evaluation report containing converter parameter optimization suggestions and identification of weak links in the power grid; start hierarchical early warning based on preset thresholds (such as δ stab > 0.003 triggers secondary early warning, τ L > 0.8% per month triggers primary early warning), automatically generate control parameter adjustment instructions or scheduling strategy optimization schemes, and achieve proactive prevention and control of safety risks.

[0167] The application establishes a multi-dimensional coupling evaluation system, which first couples the control parameters such as the rotational inertia and damping coefficient of the network-type converter with the system short-circuit ratio, power angle stability, and lithium battery aging index, and breaks through the traditional single device evaluation mode.

[0168] 1. Synergistic collection and fusion of multi-dimensional parameters:

[0169] The system first synchronously collects key parameters in three core dimensions through multi-source heterogeneous data fusion technology:

[0170] Network-type converter dimension: real-time collection of control parameters such as rotational inertia (J), damping coefficient (D), and virtual impedance (Z) (sampling frequency 10 kHz);

[0171] Power grid system dimension: monitoring of system characteristic parameters such as short-circuit ratio (SCR), power angle maximum swing (δ max ), voltage drop depth, and frequency deviation (100 ms update grid strength index);

[0172] Lithium battery dimension: collection of mechanism parameters such as lithium precipitation rate (j LP ), active material loss rate, and aging recession amount (τ L ) (5-minute sampling interval).

[0173] Through preprocessing methods such as Kalman filtering and Min-Max normalization, noise is eliminated and data scale is unified, laying a data foundation for multi-dimensional coupling.

[0174] 2. Cross-dimensional association design of quantitative indicators:

[0175] By constructing core evaluation indicators, parameters in different dimensions are directly associated to realize the mapping of "parameter-indicator-risk":

[0176] Dynamic stability index (δ stab ): coupling of converter rotational inertia (J), damping coefficient (D), and grid frequency deviation (Δω max ), in the formula, the ratio of rotational inertia time constant (T J =J / Pn), damping time constant (T D =D / Pn), and frequency deviation quantifies the influence of converter control parameters on grid frequency stability;

[0177] Voltage support capability index (V support ): association of converter virtual impedance (Z), grid voltage recovery characteristics (U rec / U fault ), and power angle stability (δ max), which reflects the comprehensive effect of the converter on the grid voltage support by coupling the power angle swing with the voltage recovery capability through an exponential function;

[0178] Short circuit ratio improvement effect index (SCR improve ): Coupling the converter configuration with the grid strength, quantifying the improvement of the converter on the weak grid stability through the difference of the SCR before and after the energy storage access;

[0179] Lithium battery aging safety index (τ L ): Coupling the battery lithium precipitation Active material loss mechanism parameters with the converter control strategy (such as dynamic adjustment of the moment of inertia), evaluating the mutual influence of battery aging and system stability (such as in Example 2, dynamic adjustment of J can reduce τ L while improving δ stab ).

[0180] 3. Synergistic calculation of the three-layer coupling evaluation model:

[0181] Through the three-layer evaluation system of "device layer-control layer-system layer", the deep coupling of multi-dimensional parameters is realized:

[0182] Device layer: Evaluate the hardware status of the converter power device temperature, lithium battery SOC balance, etc., and correlate the battery aging and the hardware load of the converter;

[0183] Control layer: Through small signal stability analysis, evaluate the gain margin and phase angle margin of the virtual synchronous control loop, and correlate the converter control parameters (J, D) and control stability;

[0184] System layer: Simulate three-phase short circuit, LVRT and other fault scenarios, evaluate the system power angle stability margin and frequency recovery time, and correlate the synergistic response of the grid short circuit ratio (SCR), converter parameters and battery aging state.

[0185] Finally, through the risk quantification algorithm (R = Σ weight × normalized risk value), the risk contributions of each dimension are coupled and calculated to obtain the total risk value of the system, realizing the cross-dimensional evaluation from a single parameter to the system level risk.

[0186] 4. Coupling verification of scenario simulation:

[0187] In typical scenarios such as strong / weak grid, high proportion of new energy, LVRT, etc., the coupling effect of multi-dimensional parameters is verified: for example, in the weak grid scenario (SCR = 1.8), the adjustment of the converter J, D parameters is evaluated simultaneously on the improvement of the SCR (grid dimension), frequency stability (δ stab , converter dimension) and battery aging rate (τ L )., the comprehensive effect of battery dimension) to ensure that the optimization direction of each dimension parameter is consistent (such as improving SCR without accelerating battery aging).

[0188] Experimental Example 1: Network-type converter power oscillation suppression evaluation based on neural network optimization

[0189] Test conditions: energy storage power station parameters: 100 MW / 200 MWh, converter initial parameters J = 400 kg·m 2 , D = 1800, Z = 0.1 + j0.3Ω, connected to 220kV power grid (initial SCR = 2.8, new energy ratio 60%).

[0190] Evaluation steps:

[0191] Dynamic characteristic test: simulate a 15% power grid voltage drop, the power oscillation peak reaches ±12% rated power without optimization, lasting 1.0s; after neural network optimization, J = 550 kg·m 2 , D = 2200, the optimized virtual impedance is adjusted to Z = 0.15 + j0.4Ω, the power oscillation peak is reduced to ±4%, the duration is shortened to 0.35s, and the oscillation decay rate is improved by about 3 times.

[0192] Stability index calculation:

[0193] Before optimization:

[0194] After optimization:

[0195] The index drop reaches 76.3%, verifying the effect of coordinated optimization of moment of inertia and virtual impedance.

[0196] Short circuit ratio improvement effect:

[0197] After connecting the optimized network-type energy storage, the SCR is improved from 2.8 to 3.2, with an improvement of 14.3%, meeting the basic requirement of weak grid SCR improvement ≥10%.

[0198] The test results are shown in Table 2:

[0199] Table 2

[0200]

[0201]

[0202] Experimental Example 2: Network-type energy storage scheduling safety evaluation based on lithium battery mechanism

[0203] 1. Preliminary preparation and parameter setting

[0204] (1) Hardware environment: 100MWh ternary lithium battery pack (single capacity 280Ah, 1000 series, 36 parallel), BMS (battery management system) supports SOC, lithium analysis rate monitoring (sampling interval 5 minutes).

[0205] (2) Initial parameter setting: SOC working interval 20% ~ 80%, conventional strategy lithium analysis rate j LP = 0.018A / cm 2 , optimization strategy limit j LP ≤ 0.012A / cm 2 ; initial moment of inertia of converter 350kg·m 2 .

[0206] 2. Lithium battery mechanism data collection and preprocessing

[0207] (1) Collect parameters: record SOC (such as 30%, 50%, etc.), lithium analysis rate j LP , active material loss rate (initial 0.5% / month) every 5 minutes;

[0208] (2) Min-Max normalization processing: take SOC as an example, x min = 20%, x max = 80%, at a certain time SOC = 50%, after normalization:

[0209] 3. Aging recession amount τ L Calculation

[0210] Conventional strategy calculation (30 days): recession amount caused by lithium analysis every day (empirical coefficient, based on battery characteristics library), j LP = 0.018A / cm 2 , then every day 30 days cumulative SEI film growth (measured value), active material loss (measured value); then the total decay amount τ L = 0.54% + 0.21% + 0.10% = 0.85%.

[0211] Mechanism optimization strategy calculation (30 days): j LP = 0.012A / cm 2 ; every day 30 days cumulative SEI film growth active material loss then the total decay amount τ L = 0.36% + 0.10% + 0.06% = 0.52%.

[0212] (4) Dynamic stability index:

[0213] Conventional strategy J = 350 kg·m 2 The dynamic stability index δ is calculated as stab = 0.0035.

[0214] Optimized strategy J = 420 kg·m 2 The dynamic stability index δ is calculated as stab = 0.0021.

[0215] (5) System risk calculation:

[0216] Conventional strategy: equipment layer risk 0.25, control layer risk 0.22, system layer risk 0.18, total risk value R = 0.21 (weight ratio 0.3:0.4:0.3).

[0217] Optimized strategy: equipment layer risk 0.15, control layer risk 0.13, system layer risk 0.11, total risk value R = 0.13, reduced by 38.1% compared with the conventional strategy, in line with the risk assessment model calculation logic.

[0218] 4. Dispatching strategy optimization verification

[0219] Adjustment measures: when j LP > 0.012 A / cm 2 , the BMS sends an instruction to reduce the charging current (from 1C to 0.8C);

[0220] Effect verification: continuous operation for 30 days, daily comparison of τ L cumulative value, confirming that the optimized strategy reduces by 38.8% compared with the conventional strategy.

[0221] The test results are shown in Table 3:

[0222] Table 3

[0223] Evaluation index Conventional strategy Mechanism optimization strategy Promotion amplitude 30-day decay amount τ L ]] 0.85％ 0.52％ -38.8％ Lithium plating rate j LP (A / cm 2 )]]> 0.018 (uncontrolled) 0.012 (precise control) -33.3％ moment of inertia J (kg-m 2 )]]> 350 (fixed value) 420 (dynamic adjustment) +20％ Dynamic stability indicator δ stab ]]> 0.0035 0.0021 -40％ System risk value R 0.21 0.13 -38.1％

[0224] The experimental data show that the power oscillation peak value of the grid-connected type converter based on neural network optimization can be reduced by 66.7%, and the oscillation duration is shortened by 65%; the dispatching strategy considering the lithium battery mechanism can reduce the aging recession amount by 38.8%, and the system risk value is reduced by 38.1%. In summary, the method of the present application fills the gap in dynamic characteristic evaluation, lithium battery mechanism integration and grid adaptability evaluation in the prior art, and provides systematic technical support for safe operation of grid-connected type energy storage power stations.

[0225] The embodiment of the present application also provides a storage medium for storing a computer program, which is executed to perform at least the method described above.

[0226] The embodiment of the present application also provides a control device, comprising a processor and a storage medium for storing a computer program; wherein the processor is used to execute the computer program and perform at least the method described above.

[0227] The embodiment of the present application also provides a processor, which executes a computer program and performs at least the method described above.

[0228] The storage medium can be implemented by any type of nonvolatile storage device, or a combination thereof. The nonvolatile storage device can be a read-only memory (ROM), a programmable read-only memory (PROM), an erasable programmable read-only memory (EPROM), an electrically erasable programmable read-only memory (EEPROM), a ferromagnetic random access memory (FRAM), a flash memory, a magnetic surface memory, an optical disc or a compact disc read-only memory (CD-ROM). The magnetic surface memory can be a disk memory or a tape memory. The storage medium described in the embodiment of the present application is intended to include but is not limited to these and any other suitable types of memory.

[0229] In several embodiments provided by the present application, it should be understood that the disclosed system and method can be implemented in other manners. The described device embodiments are merely schematic, for example, the division of the units is only a logical function division, and there can be another division manner in actual implementation, for example, a plurality of units or components can be combined or integrated into another system, or some features can be ignored or not executed. In addition, the displayed or discussed coupling between the components can be indirect coupling or direct coupling through some interface, device or unit, and can be electrical, mechanical or other forms.

[0230] The units described as separated components can or can not be physically separated, and the components displayed as units can or can not be physical units, that is, they can be located in one place, or can be distributed on a plurality of network units; some or all of the units can be selected according to actual needs to achieve the purposes of the embodiments.

[0231] In addition, each functional unit in each embodiment of the present application can be integrated in one processing unit, or each unit can be separately as a unit, or two or more units can be integrated in one unit; the integrated unit can be realized in the form of hardware or in the form of hardware plus software functional unit.

[0232] Those skilled in the art can understand that all or part of the steps of the above-mentioned method embodiments can be completed by program instruction related hardware, and the foregoing program can be stored in a computer readable storage medium, and the program executes the steps of the above-mentioned method embodiments when executed; and the foregoing storage medium includes mobile storage device, read-only memory (ROM, Read-Only Memory), random access memory (RAM, Random Access Memory), magnetic disk or optical disk and various storage program codes.

[0233] Alternatively, the integrated unit of the present application can be stored in a computer readable storage medium if it is realized in the form of software functional module and sold or used as an independent product. Based on such understanding, the technical solutions of the embodiments of the present application can be embodied in the form of software product, which is stored in a storage medium and includes a plurality of instructions for making a computer device (which can be a personal computer, a server, or a network device, etc.) execute all or part of the methods described in the embodiments of the present application. The foregoing storage medium includes mobile storage device, ROM, RAM, magnetic disk or optical disk and various storage program codes.

[0234] The methods disclosed in the several method embodiments of the present application can be combined arbitrarily without conflict to obtain new method embodiments.

[0235] The features disclosed in the several product embodiments of the present application can be combined arbitrarily without conflict to obtain new product embodiments.

[0236] The features disclosed in the several method or device embodiments of the present application can be combined arbitrarily without conflict to obtain new method embodiments or device embodiments.

[0237] The above is a further detailed description of the present application in combination with specific preferred embodiments, and the specific implementation of the present application cannot be limited to these descriptions. For those skilled in the art of the present application, without departing from the concept of the present application, a number of equivalent substitutions or obvious modifications can be made, and the same performance or use should be regarded as belonging to the protection scope of the present application.

Claims

1. A system safety assessment method for a grid-type energy storage converter power station, characterized in that, Includes the following steps: S1. Multi-source heterogeneous data fusion acquisition: Real-time acquisition of control parameters of grid-type energy storage converter, grid dynamic data and lithium battery mechanism data, and preprocessing through noise cancellation and normalization algorithms; S2. Construction of a quantitative evaluation index system for network characteristics: Establish dynamic stability index, voltage support capability index, short-circuit ratio improvement effect index and lithium battery aging safety index to quantify the multi-dimensional safety status of the system; S3. Construction of Multiphysics Coupled Risk Assessment Model: Establish a three-layer assessment system of equipment layer, control layer and system layer, calculate the normalized risk value of each layer through weight allocation algorithm, and obtain the total risk value of the system by weighted summation; S4. Scenario-based simulation verification and optimization: Set up scenarios of strong power grid, weak power grid, high proportion of new energy and low voltage ride-through, and verify the system adaptability and optimize control parameters through time domain simulation; S5. Safety Assessment Report Generation and Intelligent Early Warning: Generates an assessment report containing parameter optimization suggestions, triggers tiered early warnings based on indicator thresholds, and automatically generates closed-loop control commands.

2. The method as described in claim 1, characterized in that, The preprocessing in step S1 includes: For the converter control parameters sampled at high frequencies, a state-space modeling filtering algorithm is used to eliminate Gaussian noise through prediction-update iteration; For lithium battery mechanism data, a range normalization algorithm is used to map the parameters to a unified range.

3. The method as described in claim 1 or 2, characterized in that, In step S2: The dynamic stability index is related to the converter's moment of inertia, damping coefficient, and grid frequency deviation. The voltage support capability index is related to the converter's virtual impedance, grid voltage recovery characteristics, and power angle stability. The lithium battery aging safety index is calculated cumulatively based on the aging degradation caused by the growth of the solid electrolyte interface film, lithium plating, and loss of active materials.

4. The method according to any one of claims 1 to 3, characterized in that, In step S3: The equipment layer evaluates the temperature of the power devices in the converter and the state-of-charge balance of the lithium battery. The control layer evaluates the gain margin and phase margin of the control loop through small-signal stability analysis; The system layer evaluates the power angle stability margin and frequency recovery time through time-domain simulation of fault scenarios.

5. The method as described in claim 4, characterized in that, The weight allocation algorithm uses the analytic hierarchy process (AHP). Construct judgment matrices for the device layer, control layer, and system layer, and determine their relative importance through pairwise comparisons; The initial weight vector is calculated using the eigenvector method, and the final weight matrix is ​​determined after a consistency check.

6. The method according to any one of claims 1 to 5, characterized in that, In step S4: Verification of frequency deviation suppression capability in high-voltage power grid scenarios; Verification of short-circuit ratio improvement capability in weak power grid scenarios; The voltage recovery time and power angle overshoot suppression capability were verified in a low-voltage ride-through scenario.

7. The method according to any one of claims 1 to 6, characterized in that, The graded early warning in step S5 includes: When the dynamic stability index or voltage support capability index exceeds the threshold, a level two early warning is triggered and a control parameter adjustment command is generated. When the aging safety index of the lithium battery exceeds the threshold, a first-level warning is triggered and a scheduling strategy optimization instruction is generated.

8. The method as described in claim 7, characterized in that, The closed-loop control command execution process includes: Identify the core control parameters that need adjustment and determine the optimization range; Parameter adjustment values ​​are generated based on optimization algorithms and distributed in real time. Re-collect data to verify whether the indicators have returned to a safe range. If they do not meet the standards, iterate and optimize until the requirements are met.

9. The method according to any one of claims 1 to 8, characterized in that, The calculation of the lithium battery aging safety index is based on an electrochemical mechanism model. The cumulative aging degradation is dynamically quantified by real-time collection of state of charge, lithium plating rate, and active material loss rate.

10. A system safety assessment system for a grid-type energy storage converter power station, characterized in that, include: Multi-source data acquisition and preprocessing module: It acquires control parameters of grid-type energy storage converter, grid dynamic data and lithium battery mechanism data in real time, and performs preprocessing through noise cancellation and normalization algorithms; The network characteristic quantification index calculation module establishes dynamic stability index, voltage support capability index, short-circuit ratio improvement effect index and lithium battery aging safety index, quantifying the multi-dimensional safety status of the system. Multiphysics Coupled Risk Assessment Module: It establishes a three-layer assessment system of equipment layer, control layer and system layer, calculates the normalized risk value of each layer through weight allocation algorithm, and obtains the total risk value of the system by weighted summation; Scenario-based simulation verification and optimization module: It sets up strong power grid, weak power grid, high proportion of new energy and low voltage ride-through scenarios, and verifies the system adaptability and optimizes control parameters through time domain simulation; Assessment Report and Intelligent Early Warning Module: It generates assessment reports containing parameter optimization suggestions, triggers graded early warnings based on indicator thresholds, and automatically generates closed-loop control commands.

Images