New energy power system minimum inertia demand rapid assessment method based on mechanism-data fusion

Abstract

The invention relates to the technical field of power system dynamic safety analysis, and discloses a mechanism-data fusion new energy power system minimum inertia demand rapid evaluation method, which comprises the following steps: in an offline stage, establishing an expansion system frequency response model containing new energy; generating a plurality of sample scenes in a given disturbance power range and a new energy permeability range; a model minimum inertia demand and a simulation minimum inertia demand are obtained through expansion of a system frequency response model and stepping type time domain simulation calculation, a polynomial regression method and a polynomial regression equation are utilized, and the coefficient and order of the polynomial regression equation are determined by adopting an optimization algorithm; during online evaluation, real-time disturbance power and new energy permeability of a power grid are obtained; and obtaining a model minimum inertia demand and a corresponding minimum inertia demand deviation compensation value, and outputting a final system minimum inertia demand evaluation result. According to the method, the final evaluation result has the calculation efficiency of the extended SFR model and the evaluation precision of stepping simulation.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of power system dynamic security analysis, in particular to a mechanism-data fusion new energy power system minimum inertia demand rapid evaluation method. BACKGROUND

[0002] In recent years, new energy has become an important trend of energy transformation by accessing power system in large scale through power electronic equipment. However, the power electronic power supply does not have the rotational inertia characteristics of the traditional synchronous generator, which leads to the decrease of the overall inertia level of the power system. Once the system is subjected to power disturbance, it may cause frequency collapse, and the risk of frequency stability continues to increase, which has become a core challenge for the safe operation of high proportion of new energy power system. Therefore, it is very important to quickly and accurately evaluate the minimum inertia demand of new energy power system for ensuring the safe and stable operation of the system.

[0003] The existing minimum inertia demand evaluation methods of power system mainly have the following kinds, one is to calculate the minimum inertia demand of the system under the maximum frequency deviation through step-by-step time domain simulation. The evaluation results of this kind of method are accurate and reliable, but the complexity of solving is significantly increased due to the fine modeling of system elements, which leads to huge time consumption in the evaluation process of minimum inertia demand, and it is difficult to meet the online demand; two is the minimum inertia demand evaluation method based on average system frequency (ASF) model. The core feature of the ASF model is that it completely retains the speed regulation link of all units in the system, and its calculation result is more accurate, but the evaluation speed is slower; three is the minimum inertia demand evaluation method of system based on system frequency response (SFR) model. This method is based on SFR model, considers the frequency regulation characteristics of power electronic power supply, and constructs a simplified equivalent frequency model through parameter aggregation. The evaluation speed of this method is fast, but the error of the minimum inertia demand obtained is large due to the high simplification of the model. SUMMARY

[0004] In order to overcome the shortcomings of the prior art, the present application provides a mechanism-data fusion new energy power system minimum inertia demand rapid evaluation method, which takes into account the calculation efficiency and evaluation accuracy of minimum inertia demand evaluation. First, in the mechanism part, the present application establishes an extended SFR model containing new energy, analyzes the conditional equation containing minimum inertia demand under the constraint of maximum frequency deviation, and solves the minimum inertia demand by Newton method; secondly, in the data part, for the given disturbance power and new energy penetration , the simulation minimum inertia demand corresponding to the detailed model is obtained in the offline stage, and the model minimum inertia demand under the simplified model is calculated according to the method of mechanism part, and the minimum inertia demand deviation , the above process is repeated to obtain a sample set. A polynomial regression is used to establish a nonlinear correlation between the disturbance power and the new energy penetration rate and the minimum inertia requirement deviation . Finally, in the online evaluation, the results of the mechanism part and the data part are combined to achieve a fast and accurate minimum inertia requirement evaluation result. An effective means is provided for frequency stability analysis of high-proportion new energy power systems.

[0005] The present application is realized by the following technical solutions:

[0006] A mechanism-data fusion new energy power system minimum inertia requirement fast evaluation method, comprising:

[0007] In the offline stage, an extended system frequency response model containing new energy is established;

[0008] Within a given disturbance power range and new energy penetration rate range, a plurality of sample scenarios are generated, each sample scenario being composed of disturbance power and new energy penetration rate ;

[0009] For each of the sample scenarios, the model minimum inertia requirement and the simulation minimum inertia requirement are calculated by the extended system frequency response model and the step-by-step time domain simulation, respectively, and the minimum inertia requirement deviation is calculated to constitute a training sample set;

[0010] Based on the training sample set, a polynomial regression equation is established by using a polynomial regression method, with disturbance power and new energy penetration rate as independent variables, and minimum inertia requirement deviation as dependent variable, and an optimization algorithm is used to determine the coefficients and order of the polynomial regression equation;

[0011] In the online evaluation, the real-time disturbance power and new energy penetration rate of the power grid are obtained; the real-time disturbance power and new energy penetration rate are input into the extended system frequency response model and the polynomial regression equation to obtain the model minimum inertia requirement and the corresponding minimum inertia requirement deviation compensation value , the model minimum inertia requirement and the corresponding minimum inertia requirement deviation are summed, and the final system minimum inertia requirement evaluation result is output.

[0012] As optimization, the structure of the extended system frequency response model comprises:

[0013] a system inertia module, a transfer function of the system inertia module being , an input of the system inertia module being a difference between a power disturbance and a frequency modulation power , an output of the system inertia module being a system frequency deviation ; ;

[0014] a frequency response module, a transfer function of the frequency response module being , an input of the frequency response module being the system frequency deviation , an output of the frequency response module being weighted by a feedforward coefficient to form the frequency modulation power , wherein, is a synchronous unit damping coefficient, is a composite damping coefficient; s is a Laplace operator, R is a regulation coefficient, K is a new energy frequency response coefficient, H is a system inertia, and T is a time constant.

[0015] As optimization, a specific process for calculating the model minimum inertia requirement by the extended system frequency response model comprises:

[0016] deriving a time-domain expression of the system frequency deviation according to the extended system frequency response model;

[0017] solving to obtain a time at which the system frequency reaches a lowest point according to a condition of a frequency change rate ;

[0018] constructing a conditional equation about the system inertia H based on the time with a constraint that a maximum frequency deviation does not exceed a set threshold;

[0019] solving the conditional equation by using a Newton method to obtain the model minimum inertia requirement .

[0020] As optimization, the polynomial regression evaluation process comprises:

[0021] inputting sample data, including a disturbance power and a new energy penetration rate ;

[0022] constructing a polynomial regression equation to establish the disturbance power and the new energy penetration rate to the minimum inertia requirement deviation a mapping relationship;

[0023] determining optimal parameters of the polynomial regression equation by using a genetic algorithm;

[0024] outputting the determined polynomial equation.

[0025] As optimization, the parameters of the polynomial regression equation determined by the genetic algorithm include the order and coefficients of the polynomial regression equation.

[0026] As optimization, the fitness function of the genetic algorithm is:

[0027] ;

[0028] wherein N represents the number of samples in the training sample set, represents the actual inertia deviation of the i th sample, represents the predicted inertia deviation of the i th sample, and MSE represents the mean square error.

[0029] As optimization, the simulation minimum inertia requirement is calculated by step-by-step time domain simulation.

[0030] A power system simulation model containing detailed models of synchronous generators, speed regulation systems and new energy power generation equipment is adopted, the system inertia H is gradually increased and time domain simulation is performed until the maximum frequency deviation of the system meets the safety constraint, at which time the corresponding system inertia is the simulation minimum inertia requirement .

[0031] As optimization, the specific process for obtaining the real-time disturbance power of the power grid is as follows: according to the real-time operation mode of the power grid and the set of expected faults, the most serious power disturbance value is determined as the disturbance power .

[0032] As optimization, the specific process for obtaining the real-time new energy penetration rate of the power grid is as follows: based on the real-time data of the energy management system, the ratio of the total new energy output to the total system load is calculated to obtain the new energy penetration rate .

[0033] The application further discloses a mechanism-data fusion new energy power system minimum inertia requirement rapid evaluation system, which is used for executing the mechanism-data fusion new energy power system minimum inertia requirement rapid evaluation method.

[0034] The offline modeling unit is used to build an extended system frequency response model including new energy sources in the offline phase; within a given range of disturbance power and new energy penetration rate, it generates multiple sample scenarios, each sample scenario consisting of disturbance power... and the penetration rate of new energy Composition; For each of the sample scenarios, the minimum inertia requirement of the model is calculated using the extended system frequency response model and step-time domain simulation, respectively. and simulation minimum inertia requirements And calculate its minimum inertia requirement deviation. To form a training sample set; based on the training sample set, a multinomial regression method is used to establish a perturbation power... and the penetration rate of new energy As the independent variable, with minimum inertia demand deviation The polynomial regression equation for the dependent variable is given, and the coefficients and order of the polynomial regression equation are determined using an optimization algorithm.

[0035] The online assessment unit is used to acquire the real-time disturbance power of the power grid. and the penetration rate of new energy Real-time disturbance power and the penetration rate of new energy The minimum inertia requirement of the model is obtained by inputting the data into the extended system frequency response model and the polynomial regression equation. and the corresponding minimum inertia requirement deviation compensation value Minimum inertia requirement for the model and the corresponding minimum inertia requirement deviation Summation is performed to output the final system minimum inertia requirement assessment result. .

[0036] Compared with the prior art, the present invention has the following advantages and beneficial effects:

[0037] This invention utilizes the computational speed of the SFR model (mechanistic model) to achieve minimum inertia requirements. Millisecond-level rapid preliminary assessment, through offline learning of the required deviation between the detailed model and the mechanistic model. The preliminary assessment results are corrected online with high precision, resulting in the final assessment results. It combines the computational efficiency of extended SFR models with the evaluation accuracy of step-by-step simulations.

[0038] When applied online, this invention requires only one analytical calculation of the extended SFR model and one calculation of the polynomial function, resulting in extremely low computational load. It completely avoids the time-consuming processes of complex iterations and differential equation solving in step-by-step simulation, making the evaluation speed several orders of magnitude faster than traditional step-by-step time-domain simulation methods. This meets the requirements of online safety analysis and real-time decision-making for power grids, allowing the dispatch center to refresh the minimum inertia requirements of the entire network at the minute or even second level, and dynamically track changes in system operation.

[0039] This invention is achieved through The definition directly quantifies the error in the extended SFR model caused by simplifications such as aggregation and linearization. Utilizing the powerful nonlinear fitting capability of multinomial regression, it accurately learns the relationship between this error and operating conditions. and The complex mapping relationship of ) clearly acknowledges and precisely quantifies the inherent errors in the simplified mechanism model, and corrects them through data-driven methods, making the results closer to the actual system characteristics. Thus, the evaluation results are no longer idealized theoretical values, but rather more credible and reliable lower bounds of inertia that have been corrected for actual system characteristics, providing a solid basis for scheduling decisions.

[0040] At the mechanistic level, this invention will Embedded as a core parameter in the extended SFR model; at the data level, and Interactive items ( ) and higher-order terms ( , By introducing regression equations, this method captures complex correlations that are difficult to describe using traditional methods. Even in power systems with high or even very high proportions of renewable energy integration, it maintains excellent evaluation performance, demonstrating strong technological foresight and lifecycle, thus accurately reflecting the renewable energy penetration rate. The complex nonlinear effects of changes on the system's inertia requirements.

[0041] Although the sample generation and model training processes in this invention are complex, they are both one-time or periodic background tasks that do not affect online speed. In online applications, the process is clear, the calculations are simple, and it can be directly integrated into existing energy management systems (EMS) or decision support systems. Input parameters ( and All of these can be easily obtained from real-time power grid data. This method is not just a theoretical algorithm, but also a practical tool that can provide real-time, key safety indicators for the daily dispatch and operation of the power grid, thus forming a complete and implementable technical solution from offline training to online application.

[0042] The present application provides a mechanism-data fusion new energy power system minimum inertia demand rapid evaluation method, compared with the prior art, the method provided by the present application balances the calculation efficiency and the evaluation accuracy through the comparative analysis of the extended SFR system model containing new energy and the detailed full-time domain simulation model, which realizes the rapid preliminary evaluation of the minimum inertia demand by using the extended SFR model, and guarantees the evaluation accuracy by means of the detailed full-time domain simulation model; the method provides more comprehensive and accurate theoretical support for the minimum inertia demand evaluation and frequency stability characteristic analysis of the high-proportion new energy power system, and has the efficiency and accuracy of the minimum inertia demand evaluation. BRIEF DESCRIPTION OF DRAWINGS

[0043] The drawings described herein are used to provide further understanding of the embodiments of the present application, constitute a part of the present application, and do not constitute a limitation on the embodiments of the present application. In the drawings:

[0044] Figure 1 Schematic diagram of the extended SFR model containing new energy;

[0045] Figure 2 Flowchart of the sample set generation;

[0046] Figure 3 Flowchart of the minimum inertia demand polynomial regression evaluation method. DETAILED DESCRIPTION

[0047] In order to make the purpose, technical scheme and advantages of the present application more clear, the present application is further described in detail below in combination with the embodiments and drawings, the illustrative embodiments of the present application and the description thereof are used to explain the present application, and do not constitute a limitation on the present application.

[0048] The present embodiment 1 provides a mechanism-data fusion new energy power system minimum inertia demand rapid evaluation method, which comprises:

[0049] In the offline stage, the calculation is performed in the mechanism part and the data part.

[0050] S1, in the given disturbance power range and new energy penetration rate range, a plurality of sample scenarios are generated, each sample scenario is composed of disturbance power and new energy penetration rate ;

[0051] S2, for each of the sample scenarios, the model minimum inertia demand and the simulation minimum inertia demand are calculated by the extended system frequency response model and the stepwise time domain simulation respectively, and the minimum inertia demand deviation is calculated to constitute a training sample set.

[0052] In some embodiments, the model minimum inertia requirement is calculated by the extended system frequency response model The specific process includes:

[0053] S2.1, deriving a time-domain expression of system frequency deviation according to the extended system frequency response model ;

[0054] S2.2, solving the time when the system frequency reaches the lowest point according to the condition of frequency change rate ; ;

[0055] S2.3, constructing a conditional equation about system inertia H based on the time with the constraint that the maximum frequency deviation does not exceed the set threshold;

[0056] S2.4, solving the conditional equation by using Newton method to obtain the model minimum inertia requirement .

[0057] The more specific process is as follows:

[0058] I. Mechanism part:

[0059] 1. Establishing an extended system frequency response model SFR containing new energy, obtaining a minimum inertia requirement conditional equation under the constraint of maximum frequency deviation, and solving by using Newton method to obtain .

[0060] In some embodiments, the structure of the extended system frequency response model includes:

[0061] A system inertia module, the transfer function of the system inertia module is , the input of the system inertia module is the difference between power disturbance and frequency modulation power , and the output of the system inertia module is system frequency deviation ; ;

[0062] A frequency response module, the transfer function of the frequency response module is , the input of the frequency response module is system frequency deviation , and the output of the frequency response module is weighted by feedforward coefficient to form the frequency modulation power , wherein, is the synchronous unit damping coefficient, is the composite damping coefficient; s is the Laplace operator, R is the regulation coefficient, K is the new energy frequency response coefficient, H is the system inertia, and T is the time constant.

[0063] As Figure 1 shown, and by Figure 1 The transfer function of which can be obtained as:

[0064] ; (1)

[0065] Where R is the adjustment coefficient, T is the time constant, is the synchronous unit damping coefficient, is the composite damping coefficient, is the system natural angular frequency, is the system damping ratio.

[0066] ; (2)

[0067] According to the inverse Laplace transform, the time domain expression of formula (1) is:

[0068] ; (3)

[0069] Where, ; (4)

[0070] From formula (3), we can know that:

[0071] ; (5)

[0072] When the power system frequency reaches the maximum deviation, we can get:

[0073] ; (6)

[0074] Substituting formula (6) into formula (3), we can get the maximum frequency deviation of the system:

[0075] ; (7)

[0076] Taking the maximum frequency deviation of no more than 1 Hz as an example, the minimum inertia demand condition equation under the maximum frequency deviation constraint of the power system is:

[0077] ; (8)

[0078] Solve the equation by Newton method, and get the minimum inertia demand of the extended SFR model .

[0079] II. Data part

[0080] In some embodiments, the specific process of calculating the simulation minimum inertia demand by step-by-step time domain simulation is as follows:

[0081] The power system simulation model comprises a synchronous generator, a speed regulation system and a new energy power generation device detailed model, the system inertia H is gradually increased and time domain simulation is performed until the maximum frequency deviation of the system meets the safety constraint, and the system inertia corresponding to the simulation minimum inertia requirement is the simulation minimum inertia requirement .

[0082] For a randomly obtained sample scene, the system minimum inertia requirement condition equation of the extended SFR model is analyzed and calculated, and the system minimum inertia requirement under the corresponding maximum frequency deviation is calculated through step-by-step time domain simulation and . The minimum inertia requirement deviation formula is obtained as follows:

[0083] ; (9)

[0084] It is judged whether the obtained sample quantity meets the preset requirement, and if not, the above sample generation process is repeated until the sample quantity meets the preset requirement. The characteristics of each sample are and , and the label is . The sample set generation flow chart is shown in Figure 2 .

[0085] S3, based on the training sample set, a polynomial regression method is used to establish a polynomial regression equation with disturbance power and new energy penetration rate as independent variables, and minimum inertia requirement deviation as dependent variable, and an optimization algorithm is used to determine the coefficients and order of the polynomial regression equation.

[0086] In some embodiments, the polynomial regression evaluation process comprises:

[0087] S3.1, input sample data, including disturbance power and new energy penetration rate ;

[0088] S3.2, construct a polynomial regression equation to establish the mapping relationship of the disturbance power and new energy penetration rate to the minimum inertia requirement deviation ;

[0089] S3.3, determine the optimal parameters of the polynomial regression equation by using a genetic algorithm;

[0090] S3.4, output the determined polynomial equation.

[0091] The relationship of about disturbance power and new energy penetration rate a polynomial regression equation, whose equation is:

[0092] ; (10)

[0093] It should be noted that the subscript of the coefficient vector , the left number corresponds to the disturbance order of the disturbance, and the right number corresponds to the order of the penetration rate. Both orders are mutually arranged and combined from 0 to n.

[0094] A plurality of data sets are trained to construct a polynomial equation, and a GA algorithm is used to determine the order and coefficients of the polynomial regression equation. The flowchart of the minimum inertia requirement polynomial regression evaluation method is shown in Figure 3 .

[0095] In some embodiments, the parameters of the polynomial regression equation determined by the genetic algorithm include the order and coefficients of the polynomial regression equation.

[0096] The fitness function of the genetic algorithm is:

[0097] ;

[0098] Where N represents the number of samples in the training sample set, represents the actual inertia deviation of the i-th sample, represents the predicted inertia deviation of the i-th sample, and MSE represents the mean square error.

[0099] The order n of the polynomial and the coefficient vector are collectively encoded as a chromosome individual of the genetic algorithm.

[0100] The fitness function value is used to evaluate the pros and cons of each chromosome individual. The method defines the fitness function as the inverse of the mean square error (MSE) between the predicted value and the true value. The optimization goal is to maximize the fitness function, which is equivalent to minimizing the mean square error MSE.

[0101] Through selection, crossover, mutation and other operations, the population is iteratively evolved, and finally converges to the individual with the highest fitness. The order and the coefficient vector decoded by the individual are the optimal solution, thereby determining the final polynomial regression evaluation equation.

[0102] Of course, the optimization algorithm can be other algorithms, such as simulated annealing algorithm, ant colony algorithm, dung beetle algorithm, black-winged kite algorithm, etc. Here, it will not be repeated.

[0103] S4, in online evaluation, the real-time disturbance power and the new energy penetration rate of the power grid are obtained; the real-time disturbance power and new energy penetration rate input into the extended system frequency response model and the polynomial regression equation to obtain the model minimum inertia requirement and the corresponding minimum inertia requirement deviation compensation value , the model minimum inertia requirement and the corresponding minimum inertia requirement deviation summing up and outputting the final system minimum inertia requirement evaluation result .

[0104] In some embodiments, the disturbance power of the power grid in real time is obtained The specific process is: according to the real-time operation mode of the power grid and the set of expected faults, the most serious power disturbance value is determined as the disturbance power .

[0105] In some embodiments, the new energy penetration rate of the power grid in real time is obtained The specific process is: based on the real-time data of the energy management system, the ratio of the total new energy output to the total system load is calculated to obtain the new energy penetration rate .

[0106] When the power grid is actually online, the real-time disturbance power ΔP L and the new energy penetration rate λ are obtained, which are input into the extended SFR system model and the polynomial regression equation to quickly obtain the minimum inertia requirement H1 and the minimum inertia requirement deviation ΔH at this time, and the two are summed to obtain the mechanism-data fusion new energy power system minimum inertia requirement.

[0107] In summary, the general process of the method of the present application is:

[0108] 1) mechanism part

[0109] 1-1) establish an extended SFR model containing new energy, obtain the minimum inertia requirement condition equation under the maximum frequency deviation constraint, and solve it by using Newton method to obtain ;

[0110] 2) data part

[0111] 2-1) a set of disturbance powers and new energy penetration rates are randomly generated in the given disturbance power range and new energy penetration rate range, the system minimum inertia requirement and under the corresponding maximum frequency deviation are calculated by solving the system minimum inertia requirement condition equation of the extended SFR model and by step-by-step time domain simulation

[0112] 2-3) Calculate the minimum inertia requirement deviation of the time-domain simulation model and the extended SFR model under each set of sample scenarios ;

[0113] 2-3) Calculate the minimum inertia requirement deviation sample set of a predetermined sample number obtained by repeating 2-1 to 2-2, wherein each sample is characterized by and , and the label is ;

[0114] 2-4) Take the disturbance power and the new energy penetration rate as independent variables, construct a polynomial feature set containing high-order terms and interaction terms of the independent variables, and expand the independent variable dimension to reflect the nonlinear relationship;

[0115] 2-5) Obtain the regression equation of the extended polynomial feature by using the polynomial regression method, and use genetic algorithm (GA) on the sample set to optimize and determine the order and coefficients of the polynomial regression equation;

[0116] 3) Actual online use part

[0117] 3-1) In actual use, the disturbance power and the new energy penetration rate of the power grid are obtained by using online data, and the minimum inertia requirement of the current system is quickly calculated by using the minimum inertia requirement analytical expression obtained by the mechanism part;

[0118] 3-2) Input the same disturbance power and new energy penetration rate to the polynomial regression equation about obtained by the data part to obtain the minimum inertia requirement deviation at this time;

[0119] 3-3) Sum the minimum inertia requirement and the minimum inertia requirement deviation to obtain the minimum inertia requirement of the new energy power system at this time .

[0120] In summary, the present application proposes a mechanism-data fusion new energy power system minimum inertia requirement rapid evaluation method. Compared with the prior art, the method proposed in the present application not only improves the calculation efficiency of the minimum inertia requirement, but also improves the minimum inertia requirement evaluation accuracy, and provides an effective means for frequency stability analysis of high-proportion new energy power systems.

[0121] Embodiment 2 discloses a mechanism-data fusion new energy power system minimum inertia demand rapid evaluation system for performing the mechanism-data fusion new energy power system minimum inertia demand rapid evaluation method described in Embodiment 1, comprising:

[0122] an offline modeling unit for establishing an extended system frequency response model containing new energy in an offline stage; generating a plurality of sample scenarios in a given disturbance power range and new energy penetration rate range, each sample scenario being composed of disturbance power and new energy penetration rate ; for each of the sample scenarios, respectively calculating model minimum inertia demand and simulation minimum inertia demand by the extended system frequency response model and step-by-step time domain simulation, and calculating minimum inertia demand deviation to constitute a training sample set; based on the training sample set, using a polynomial regression method to establish a polynomial regression equation with disturbance power and new energy penetration rate as independent variables and minimum inertia demand deviation as dependent variable, and using an optimization algorithm to determine the coefficients and order of the polynomial regression equation;

[0123] an online evaluation unit for obtaining real-time disturbance power and new energy penetration rate of the power grid; inputting the real-time disturbance power and new energy penetration rate into the extended system frequency response model and the polynomial regression equation to obtain model minimum inertia demand and corresponding minimum inertia demand deviation compensation value , summing up the model minimum inertia demand and corresponding minimum inertia demand deviation , and outputting the final system minimum inertia demand evaluation result .

[0124] The specific embodiments described above further detail the purposes, technical solutions and beneficial effects of the present application. It should be understood that the above description is only a specific embodiment of the present application and is not intended to limit the protection scope of the present application. Any modification, equivalent replacement, improvement, etc. made within the spirit and principles of the present application should be included in the protection scope of the present application.

Claims

1. A mechanism-data fusion-based rapid assessment method for minimum inertia requirements of new energy power systems, characterized in that, include: During the offline phase, an extended system frequency response model incorporating new energy sources is established. Within a given range of disturbance power and renewable energy penetration rate, multiple sample scenarios are generated. Each sample scenario is determined by the disturbance power. and the penetration rate of new energy constitute; For each of the aforementioned sample scenarios, the minimum inertia requirement of the model is calculated using the extended system frequency response model and stepped time-domain simulation, respectively. and simulation minimum inertia requirements And calculate its minimum inertia requirement deviation. , to form a training sample set; Based on the training sample set, a multinomial regression method is used to establish a perturbation power... and the penetration rate of new energy As the independent variable, with minimum inertia demand deviation The polynomial regression equation for the dependent variable is given, and the coefficients and order of the polynomial regression equation are determined using an optimization algorithm. During online assessment, the real-time disturbance power of the power grid is obtained. and the penetration rate of new energy Real-time disturbance power and the penetration rate of new energy The minimum inertia requirement of the model is obtained by inputting the data into the extended system frequency response model and the polynomial regression equation. and the corresponding minimum inertia requirement deviation compensation value Minimum inertia requirement for the model and the corresponding minimum inertia requirement deviation Summation is performed to output the final system minimum inertia requirement assessment result. .

2. The method for rapid assessment of minimum inertia requirements of a new energy power system based on mechanism-data fusion as described in claim 1, characterized in that, The structure of the extended system frequency response model includes: The system inertia module, the transfer function of which is: The input to the system inertia module is a power disturbance. With frequency modulation power difference The output of the system inertia module is the system frequency deviation. ; The frequency response module, wherein the transfer function of the frequency response module is: The input to the frequency response module is the system frequency deviation. The output of the frequency response module is fed forward via a feedforward coefficient. The frequency modulation power is formed after weighting. ,in, This refers to the damping coefficient of the synchronous generator unit. is the composite damping coefficient; s is the Laplace operator; R is the droop coefficient; K is the new energy frequency response coefficient; H is the system inertia; and T is the time constant.

3. The method for rapid assessment of minimum inertia requirements of a new energy power system based on mechanism-data fusion as described in claim 1, characterized in that, The minimum inertia requirement of the model is calculated using the extended system frequency response model. The specific process includes: Based on the extended system frequency response model, the time-domain expression for the system frequency deviation is derived. ; According to the rate of change of frequency Given the given conditions, the time it takes for the system frequency to reach its lowest point can be obtained by solving the problem. ; With maximum frequency deviation The constraint is that it does not exceed a set threshold, based on the time. Construct conditional equations for the system inertia H; The condition equations are solved using Newton's method to obtain the minimum inertia requirement of the model. .

4. The method for rapid assessment of minimum inertia requirements of a new energy power system based on mechanism-data fusion as described in claim 1, characterized in that, The polynomial regression evaluation process includes: Input sample data, including disturbance power and the penetration rate of new energy ; Construct a polynomial regression equation to establish the disturbance power. and the penetration rate of new energy Minimum inertia requirement deviation The mapping relationship; The optimal parameters of the polynomial regression equation are determined using a genetic algorithm. Output the defined polynomial equation.

5. The method for rapid assessment of minimum inertia requirements of a new energy power system based on mechanism-data fusion as described in claim 4, characterized in that, The parameters of the polynomial regression equation that need to be determined by the genetic algorithm include the order and coefficients of the polynomial regression equation.

6. The method for rapid assessment of minimum inertia requirements of a new energy power system based on mechanism-data fusion as described in claim 4, characterized in that, The fitness function of the genetic algorithm is: ； Where N represents the number of samples in the training sample set. This represents the actual inertia deviation of the i-th sample. denoted as the prediction inertia bias of the i-th sample, and MSE as the mean square error.

7. The method for rapid assessment of minimum inertia requirements of a new energy power system based on mechanism-data fusion as described in claim 1, characterized in that, The minimum inertia requirement for simulation was obtained through step-time domain simulation calculation. The specific process is as follows: A power system simulation model incorporating detailed models of synchronous generators, speed control systems, and new energy power generation equipment is used. The system inertia H is gradually increased while time-domain simulation is performed until the maximum frequency deviation of the system meets the safety constraints. The system inertia corresponding to this point is the minimum inertia requirement for the simulation. .

8. The method for rapid assessment of minimum inertia requirements of a new energy power system based on mechanism-data fusion as described in claim 1, characterized in that, Obtain real-time disturbance power of the power grid The specific process is as follows: Based on the real-time operation mode of the power grid and the set of anticipated faults, the most severe power disturbance value is determined as the disturbance power. .

9. The method for rapid assessment of minimum inertia requirements of a new energy power system based on mechanism-data fusion as described in claim 1, characterized in that, Obtain the real-time penetration rate of new energy sources in the power grid. The specific process is as follows: Based on real-time data from the energy management system, the ratio of total new energy output to total system load is calculated to obtain the new energy penetration rate. .

10. A mechanism-data fusion-based rapid assessment system for minimum inertia requirements of a new energy power system, used to execute the mechanism-data fusion-based rapid assessment method for minimum inertia requirements of a new energy power system as described in any one of claims 1-9, characterized in that, include: The offline modeling unit is used to build an extended system frequency response model including new energy sources in the offline stage; Within a given range of disturbance power and renewable energy penetration rate, multiple sample scenarios are generated. Each sample scenario is determined by the disturbance power. and the penetration rate of new energy Composition; For each of the sample scenarios, the minimum inertia requirement of the model is calculated using the extended system frequency response model and step-time domain simulation, respectively. and simulation minimum inertia requirements And calculate its minimum inertia requirement deviation. , to form a training sample set; Based on the training sample set, a multinomial regression method is used to establish a perturbation power... and the penetration rate of new energy As the independent variable, with minimum inertia demand deviation The polynomial regression equation for the dependent variable is given, and the coefficients and order of the polynomial regression equation are determined using an optimization algorithm. The online assessment unit is used to acquire the real-time disturbance power of the power grid. and the penetration rate of new energy Real-time disturbance power and the penetration rate of new energy The minimum inertia requirement of the model is obtained by inputting the data into the extended system frequency response model and the polynomial regression equation. and the corresponding minimum inertia requirement deviation compensation value Minimum inertia requirement for the model and the corresponding minimum inertia requirement deviation Summation is performed to output the final system minimum inertia requirement assessment result. .

Images