Parameter identification method and system of load model, computer device and storage medium

Abstract

The application discloses a parameter identification method and system of a load model, a computer device and a storage medium, and the method comprises the following steps: obtaining preprocessed power system disturbance data; converting the load model into a form of a state space equation and performing a simplification process; inputting the preprocessed power system disturbance data and initial values of each to-be-identified parameter into a nonlinear grey-box model to realize parameter identification of the load model. The application combines a physical model with data driving, can consider the physical mechanism of the load, can utilize a large amount of observed data, and improves the accuracy of parameter identification; meanwhile, compared with a traditional static load model and a WECC CLM model, the ZIP+IM model can not only better simulate the dynamic behavior of the load, but also has fewer parameters and lower identification difficulty. In addition, the grey-box model has clear physical significance, a short training period and fast solving speed, and has good application value.

Description

Technical Field

[0001] This invention relates to a parameter identification method, system, computer equipment, and storage medium for a load model, belonging to the field of power system technology. Background Technology

[0002] The power load model influences the stability calculation results of the power system, specifically affecting the analysis and calculation of transient stability, small disturbance stability, and voltage stability to varying degrees. Using an accurate load model in power system simulation calculations is beneficial for ensuring the safe and stable operation of the power grid and improving the reliability of power supply to users.

[0003] With the emergence of new smart grid technologies such as distributed generators, electric vehicles, and demand-side management, the load composition of power systems has become more complex, and the time-varying and uncertain nature of load fluctuations has continuously increased, posing new challenges to the accuracy of load modeling. The goal of load modeling is to develop simple mathematical models to approximate load behavior and represent the relationship between power and voltage in the load bus. Load modeling includes two main steps: 1) selecting the load model structure; 2) identifying the load model parameters. Load models are divided into static load, dynamic load, and composite load models. Composite load models consider both dynamic and static components of the load, providing a better simulation of load behavior. Physical models of composite loads mainly include the composite model consisting of static load and induction motors (ZIP+IM) and the Western Electricity Coordinating Council Load Model (WECC CLM). The WECC CLM model requires the identification of 131 parameters, making it complex and difficult to implement. In contrast, the ZIP+IM model has fewer parameters, not only simulating dynamic load behavior well but also significantly reducing the difficulty of model solving.

[0004] Load modeling and identification methods mainly fall into two categories: methods combining physical models and data-driven approaches, and purely data-driven methods such as machine learning. Methods combining physical and data-driven approaches consider the physical mechanisms of the load and are easy to simulate. However, purely data-driven methods such as machine learning lack physical meaning, data is difficult to obtain, and integration with simulation software is inconvenient. Currently, for the identification of ZIP+IM models, much research focuses on extracting typical parameters and solving them using genetic algorithms or reinforcement learning methods. These two methods suffer from drawbacks such as poor accuracy, long training cycles, and a lack of parameter universality. Summary of the Invention

[0005] In view of this, the present invention provides a method, system, computer device and storage medium for parameter identification of a load model. It identifies parameters by using a combination of physical model and data-driven approach, which has clear physical meaning, fast solution speed and good application value.

[0006] The first objective of this invention is to provide a method for parameter identification of a load model.

[0007] The second objective of this invention is to provide a parameter identification system for a load model.

[0008] A third objective of this invention is to provide a computer device.

[0009] A fourth objective of this invention is to provide a storage medium.

[0010] The first objective of this invention can be achieved by adopting the following technical solution:

[0011] A method for parameter identification of a load model, the method comprising:

[0012] Acquire preprocessed power system disturbance data;

[0013] The load model is transformed into a state-space equation and simplified to complete the construction of a nonlinear gray box model, which includes multiple parameters to be identified.

[0014] The preprocessed power system disturbance data and the initial values ​​of each parameter to be identified are input into the nonlinear gray box model to achieve parameter identification of the load model.

[0015] Furthermore, the load model is a ZIP+IM model;

[0016] The process of transforming the load model into a state-space equation form and simplifying it to complete the construction of the nonlinear gray box model specifically includes:

[0017] Based on the third-order induction motor equations of the ZIP+IM model, the active power and reactive power of the load nodes, the load model is transformed into the form of state-space equations, thus obtaining the state-space equations.

[0018] The state-space equations are simplified to obtain a nonlinear gray-box model.

[0019] Furthermore, the active power and reactive power of the load node are as follows:

[0020]

[0021] Where P represents the active power of the load node, Q represents the reactive power of the load node, and V represents the voltage of the load node. ZIP Q represents the active power of the static load. ZIP P represents the reactive power of the static load. M Q represents the active power absorbed by the induction motor. MP represents the reactive power absorbed by the induction motor. Z P I P P Q represents the percentage of active power in the constant impedance, constant voltage, and constant power components of the static model, respectively. Z Q I Q P P represents the percentage of reactive power in the constant impedance, constant voltage, and constant power components of the static model, respectively. ZIP0 Q represents the active power of the static model in steady state. ZIP0 V0 represents the reactive power of the static model under steady state, and X represents the rated voltage of the load node. m ′ represents transient reactance, E m ' represents transient potential, δ m The power angle represents the transient potential.

[0022] Furthermore, the state-space equation is as follows:

[0023]

[0024]

[0025] Among them, X m T represents the magnetizing reactance. dm ' represents the transient time constant, ω m ω represents the mechanical angular velocity of the induction motor load. s T represents the angular frequency of the load node. m H represents the torque after load equivalent. m This represents the inertial time constant.

[0026] Furthermore, the nonlinear gray box model is as follows:

[0027]

[0028]

[0029] Among them, P i Let i represent the parameter to be identified, i = 1, 2, ..., 11.

[0030] Furthermore, the acquisition of preprocessed power system disturbance data specifically includes:

[0031] At the load node, a single-phase fault for a first preset time and a three-phase fault for a second preset time are set respectively. Multiple sets of voltage, frequency, active power and reactive power are collected during a third preset time, including before the fault, the fault segment and after the fault is cleared. The input data are voltage and frequency, and the output data are active power and reactive power.

[0032] High-order harmonics and noise are filtered out from each set of input and output data to obtain preprocessed power system disturbance data.

[0033] Furthermore, after inputting the preprocessed perturbation data and the initial value of each parameter to be identified into the nonlinear gray box model, the Levenberg-Marquardt iterative algorithm is used to update each parameter to be identified.

[0034] The second objective of this invention can be achieved by adopting the following technical solution:

[0035] A parameter identification system for a load model, the system comprising:

[0036] The acquisition unit is used to acquire preprocessed power system disturbance data;

[0037] The construction unit is used to transform the load model into the form of state-space equations and simplify it, thereby completing the construction of the nonlinear gray box model, which includes multiple parameters to be identified.

[0038] The identification unit is used to input the preprocessed power system disturbance data and the initial value of each parameter to be identified into the nonlinear gray box model to realize the parameter identification of the load model.

[0039] The third objective of this invention can be achieved by adopting the following technical solution:

[0040] A computer device includes a processor and a memory for storing a processor-executable program, wherein when the processor executes the program stored in the memory, it implements the parameter identification method described above.

[0041] The fourth objective of this invention can be achieved by adopting the following technical solution:

[0042] A storage medium storing a program that, when executed by a processor, implements the parameter identification method described above.

[0043] The present invention has the following advantages over the prior art:

[0044] 1. This invention combines a physical model with a data-driven approach, taking into account the physical mechanisms of the load and facilitating simulation. Currently, with the integration of new energy loads, the types of loads are becoming more diverse. Pure physical models involve precise mathematical modeling of various loads, and the integration methods are too complex. Meanwhile, pure data-driven methods such as machine learning lack physical meaning, data is difficult to obtain, and integration with simulation software is inconvenient. Combining a physical model with a data-driven approach allows for consideration of the physical mechanisms of the load while utilizing a large amount of observed data, thus improving the accuracy of the model.

[0045] 2. This invention features a short training cycle and fast solution speed. Currently, many studies on load model identification focus on extracting typical parameters and solving them using genetic algorithms or reinforcement learning methods. These two methods may result in poor accuracy, long training cycles, and parameters lacking universality. In contrast, the gray box model has a short training cycle, fast parameter solution speed, universality, and good application value.

[0046] 3. The load model in this embodiment of the invention is a physical model of a composite load, which simultaneously considers the dynamic and static components of the load, resulting in a better simulation effect of load behavior. Physical models of composite loads mainly include composite models consisting of static loads and induction motors (ZIP+IM) and the Western Electricity Coordinating Council Load Model (WECC CLM), etc. Among them, the WECC CLM model requires the identification of 131 parameters, which is quite complex and difficult to implement; while the ZIP+IM model has only 14 parameters, which not only simulates the dynamic behavior of the load better, but also greatly reduces the difficulty of model solving. Attached Figure Description

[0047] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on the structures shown in these drawings without creative effort.

[0048] Figure 1 This is a flowchart of the parameter identification method for the load model in Embodiment 1 of the present invention.

[0049] Figure 2 This is an equivalent structural diagram of the ZIP+IM model in Embodiment 1 of the present invention.

[0050] Figure 3 This is the equivalent circuit diagram of the ZIP+IM model in Embodiment 1 of the present invention.

[0051] Figure 4 This is a flowchart illustrating the identification process of the nonlinear gray box model in Embodiment 1 of the present invention.

[0052] Figure 5(a) is a comparison of active and reactive power of the single-phase fault results in Embodiment 1 of the present invention.

[0053] Figure 5(b) is a comparison of active and reactive power in the three-phase fault results of Embodiment 1 of the present invention.

[0054] Figure 6 This is a structural block diagram of the parameter identification system for the load model in Embodiment 2 of the present invention.

[0055] Figure 7 This is a structural block diagram of the computer device according to Embodiment 3 of the present invention. Detailed Implementation

[0056] To make the objectives, technical solutions, and advantages of the embodiments of the present invention clearer, the technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, not all embodiments. All other embodiments obtained by those skilled in the art based on the embodiments of the present invention without creative effort are within the scope of protection of the present invention.

[0057] Example 1:

[0058] like Figure 1 As shown in the figure, this embodiment provides a parameter identification method for a load model, which includes the following steps:

[0059] S101. Obtain preprocessed power system disturbance data.

[0060] S1011. Select the load model.

[0061] In terms of load model structure, it is mainly divided into static load model, dynamic load model, and composite load model including static and dynamic load components. Among them, the composite load model considers both dynamic and static components of the load and has better application effect. Therefore, the ZIP+IM model is selected as the load model in this embodiment. The equivalent structure of the ZIP+IM model is as follows: Figure 2 As shown, the equivalent circuit of the ZIP+IM model is as follows: Figure 3 As shown.

[0062] S1012. Collect disturbance data generated when a short circuit or other large disturbance occurs in the power system, i.e., power system disturbance data. Specifically, based on the IEEE-39 node standard example, select node 9 as the load node, and use a ZIP+IM model connected in series at node 9 as the simulation model for simulation. Set up a single-phase fault for a first preset time and a three-phase fault for a second preset time at the adjacent node 8 of the load, and collect 6250 sets of voltage V, frequency f, active power P and reactive power Q of the load node within a third preset time including before the fault, the fault segment and after the fault is cleared. Among them, voltage V and frequency f are used as input data, and active power P and reactive power Q are used as output data. V is also called the voltage of the load node, and f, P and Q are similarly represented.

[0063] In this embodiment, the first preset time and the second preset time are 0.5s and 0.2s, respectively, and the third preset time is 5s; the power system disturbance data is generated by the software RSCAD driving RTDS.

[0064] S1013. Filter out the high-order harmonics and noise of each set of input and output data to obtain the preprocessed power system disturbance data.

[0065] In this embodiment, a Butterworth low-pass filter is used to filter out the high-order harmonics and noise of each set of input and output data in step S1011, thereby obtaining the preprocessed power system disturbance data.

[0066] S102. The load model is transformed into a state-space equation and simplified to complete the construction of the nonlinear gray box model.

[0067] Specifically, the expression for the static load component (static model) of the ZIP+IM model is shown in equation (5.1):

[0068]

[0069] Among them, P ZIP Q represents the active power of the static load. ZIP P represents the reactive power of the static load. ZIP0 Q represents the active power of the static model in steady state. ZIP0 V represents the reactive power in the static model under steady state, and P represents the voltage at the load node. Z P I P P Q represents the percentage of active power in the constant impedance, constant voltage, and constant power components of the static model, respectively. Z Q I Q P These represent the percentage of reactive power in the constant impedance, constant voltage, and constant power components of the static model, respectively, while V0 represents the rated voltage of the load node.

[0070] Specifically, the dynamic load part (dynamic model) of the ZIP+IM model is represented by the third-order induction motor equations. In the third-order induction motor equations, the electromagnetic transient process of the stator winding is ignored. The changes of the voltage, amplitude, and angular velocity of the induction motor with time are shown in equation (5.2):

[0071]

[0072] Among them, E m ' represents transient potential, δ m The power angle X represents the transient potential. m X represents the magnetizing reactance. m ′ represents transient reactance, T dm ' represents the transient time constant, ω m T represents the mechanical angular velocity of the induction motor load. mH represents the torque after load equivalent. m This represents the inertial time constant.

[0073] Specifically, according to V and E m ′ and δ m The expressions for the active and reactive power absorbed by the induction motor are derived, as shown in equation (5.3):

[0074]

[0075] Furthermore, the active power and reactive power of the load node are the sum of the static load portion and the dynamic load portion, as shown in equation (5.4):

[0076]

[0077] Furthermore, based on equations (5.2) and (5.4), the ZIP+IM model is expressed in the form of a state-space equation, which is the nonlinear third-order state equation of the ZIP+IM model, as shown in equation (5.5); in equation (5.5), the state vector x = [E m ′ δ m ω m ] T The inputs are the voltage and angular frequency u = [V ω] of the load node. s ] T The output is the active and reactive power of the load node, y = [PQ]. T .

[0078]

[0079] Furthermore, to reduce computational complexity, a constant P is used. i (i = 1, 2, ..., 11) represents the parameters to be identified in the coefficient matrix of the state-space equation, i.e., the state-space equation is simplified, and the simplified state-space equation is obtained, as shown in (5.6):

[0080]

[0081] In this embodiment, the simplified state-space equations are used as a nonlinear gray-box model.

[0082] S103. Input the preprocessed power system disturbance data and the initial value of each parameter to be identified into the nonlinear gray box model to realize the parameter identification of the load model.

[0083] like Figure 4As shown, after inputting the nonlinear gray box model, the nonlinear gray box model provides the output response (fitted response) of active power and reactive power, which is compared with the actual output response of active power and reactive power. When the error is large, each corresponding parameter to be identified is updated until the error of active power and reactive power meets the predetermined threshold (the average absolute value of the vertical distance between the active power and reactive power and the fitted curve of 6250 sets of original data points is less than 1%), and then each identified parameter is output; otherwise, the comparison and update process is repeated. In the comparison and update process, i.e., the iteration process, the Levenberg-Marquardt iterative algorithm is used. This algorithm combines the advantages of the gradient method and Newton's method, has a fast optimization speed and is not easy to get trapped in local optima. Therefore, this algorithm is used to minimize the average value of the vertical distance between the data points and the fitted curve.

[0084] Specifically, the preprocessed power system disturbance data and the initial values ​​of each parameter to be identified are imported into the nonlinear gray box model. Specifically, the input and output data of 6250 groups of single-phase faults and the corresponding initial values ​​of each parameter to be identified are imported into the nonlinear gray box model to obtain each identified parameter of the single-phase fault. At the same time, based on each identified parameter of the single-phase fault and the input data of 6250 groups of single-phase faults, the corresponding results are fitted and calculated, as shown in Figure 5(a). The input and output data of 6250 groups of three-phase faults and the corresponding initial values ​​of each parameter to be identified are imported into the nonlinear gray box model to obtain each identified parameter of the three-phase fault. At the same time, based on each identified parameter of the three-phase fault and the input data of 6250 groups of three-phase faults, the corresponding results are fitted and calculated, as shown in Figure 5(b).

[0085] As shown in Figures 5(a) and 5(b), the two blue curves represent the actual active power (reference value) and reactive power (reference value) obtained from the simulation model, respectively; the two yellow curves represent the fitted active power and reactive power values ​​obtained from the parameters identified by the nonlinear gray box model. It can be clearly seen that the fitted power curve has a small error compared with the actual curve (original curve or reference curve), indicating that the parameters identified by the nonlinear gray box model have high accuracy.

[0086] It is worth noting that: the initial values ​​of the identification parameters for single-phase faults are the same as those for three-phase faults; the initial values ​​are related to the voltage level and are typical data of the load at that voltage level; the iterative process of gray box solution requires an initial value, and using typical data as the initial value can ensure that the initial value is not too far from the true value of the parameter to be identified, thereby improving the iteration speed; MATLAB software is used for programming to identify each parameter to be identified.

[0087] Specifically, typical data can be found in [1] Zali M,S, Milanovic, et al. Generic Model of Active Distribution Network for Large Power System Stability Studies[J].IEEE Transactions on Power Systems, 2013, 28(3):3126-3133.

[0088] It is worth noting that the nonlinear gray box model, after more than 6,000 sets of data iterations, is sufficient to identify most power system disturbance data. Therefore, the nonlinear gray box model constructed in this embodiment has a certain degree of universality. Only when the load of the load node changes significantly does it need to identify the parameters again.

[0089] Those skilled in the art will understand that all or part of the steps in the methods of the above embodiments can be implemented by a program instructing related hardware, and the corresponding program can be stored in a computer-readable storage medium.

[0090] It should be noted that although the method operations of the above embodiments are described in a specific order in the accompanying drawings, this does not require or imply that these operations must be performed in that specific order, or that all the operations shown must be performed to achieve the desired result. On the contrary, the order of execution of the described steps may be changed. Additionally or alternatively, certain steps may be omitted, multiple steps may be combined into one step, and / or one step may be broken down into multiple steps.

[0091] Example 2:

[0092] like Figure 6 As shown, this embodiment provides a parameter identification system for a load model. The system includes an acquisition unit 601, a construction unit 602, and an identification unit 603. The specific functions of each unit are as follows:

[0093] Acquisition unit 601 is used to acquire preprocessed power system disturbance data;

[0094] The construction unit 602 is used to transform the load model into the form of state-space equations and simplify it, thereby completing the construction of the nonlinear gray box model, which includes multiple parameters to be identified.

[0095] The identification unit 603 is used to input the preprocessed power system disturbance data and the initial value of each parameter to be identified into the nonlinear gray box model to realize the parameter identification of the load model.

[0096] Example 3:

[0097] like Figure 7 As shown, this embodiment provides a computer device, which includes a processor 702, a memory, an input device 703, a display device 704, and a network interface 705 connected via a system bus 701. The processor 702 provides computing and control capabilities. The memory includes a non-volatile storage medium 706 and internal memory 707. The non-volatile storage medium 706 stores an operating system, computer programs, and a database. The internal memory 707 provides an environment for the operation of the operating system and computer programs in the non-volatile storage medium 706. When the computer program is executed by the processor 702, it implements the parameter identification method of Embodiment 1 described above, as follows:

[0098] Acquire preprocessed power system disturbance data;

[0099] The load model is transformed into a state-space equation and simplified to complete the construction of a nonlinear gray box model, which includes multiple parameters to be identified.

[0100] The preprocessed power system disturbance data and the initial values ​​of each parameter to be identified are input into the nonlinear gray box model to achieve parameter identification of the load model.

[0101] Example 4:

[0102] This embodiment provides a storage medium, which is a computer-readable storage medium, storing a computer program. When the computer program is executed by a processor, it implements the parameter identification method of Embodiment 1 above, as follows:

[0103] Acquire preprocessed power system disturbance data;

[0104] The load model is transformed into a state-space equation and simplified to complete the construction of a nonlinear gray box model, which includes multiple parameters to be identified.

[0105] The preprocessed power system disturbance data and the initial values ​​of each parameter to be identified are input into the nonlinear gray box model to achieve parameter identification of the load model.

[0106] It should be noted that the computer-readable storage medium in this embodiment can be a computer-readable signal medium or a computer-readable storage medium, or any combination thereof. The computer-readable storage medium can be, for example, but not limited to, an electrical, magnetic, optical, electromagnetic, infrared, or semiconductor system, apparatus, or device, or any combination thereof. More specific examples of a computer-readable storage medium may include, but are not limited to: an electrical connection having one or more wires, a portable computer disk, a hard disk, random access memory (RAM), read-only memory (ROM), erasable programmable read-only memory (EPROM or flash memory), optical fiber, portable compact disk read-only memory (CD-ROM), optical storage device, magnetic storage device, or any suitable combination thereof.

[0107] In this embodiment, the computer-readable storage medium can be any tangible medium containing or storing a program that can be used by or in connection with an instruction execution system, apparatus, or device. In this embodiment, the computer-readable signal medium can include a data signal propagated in baseband or as part of a carrier wave, carrying a computer-readable program. Such propagated data signals can take various forms, including but not limited to electromagnetic signals, optical signals, or any suitable combination thereof. The computer-readable signal medium can also be any computer-readable storage medium other than a computer-readable storage medium, which can send, propagate, or transmit a program for use by or in connection with an instruction execution system, apparatus, or device. The computer program contained on the computer-readable storage medium can be transmitted using any suitable medium, including but not limited to: wires, optical cables, RF (radio frequency), etc., or any suitable combination thereof.

[0108] The computer-readable storage medium described above can be used to write computer programs for executing this embodiment in one or more programming languages ​​or combinations thereof. These programming languages ​​include object-oriented programming languages—such as Java, Python, and C++—and conventional procedural programming languages—such as C or similar programming languages. The program can be executed entirely on the user's computer, partially on the user's computer, as a standalone software package, partially on the user's computer and partially on a remote computer, or entirely on a remote computer or server. In cases involving remote computers, the remote computer can be connected to the user's computer via any type of network, including a local area network (LAN) or a wide area network (WAN), or it can be connected to an external computer (e.g., via the Internet using an Internet service provider).

[0109] In summary, this invention combines a physical model with data-driven approaches, considering both the physical mechanisms of the load and utilizing a large amount of observed data, thus improving the accuracy of parameter identification. Furthermore, compared to traditional static load models and WECC CLM models, the ZIP+IM model not only simulates dynamic load behavior better but also has fewer parameters, making identification easier. In addition, the gray-box model has clear physical meaning, a short training cycle, and fast solution speed, making it highly valuable for practical applications.

[0110] The above description is merely a preferred embodiment of the present invention, but the scope of protection of the present invention is not limited thereto. Any equivalent substitutions or modifications made by those skilled in the art within the scope disclosed in the present invention, based on the technical solution and inventive concept of the present invention, shall fall within the scope of protection of the present invention.

Claims

1. A method for parameter identification of a load model, characterized in that, The method includes: Acquire preprocessed power system disturbance data; The load model is transformed into a state-space equation and simplified to complete the construction of a nonlinear gray box model, which includes multiple parameters to be identified. The load model is a ZIP+IM model; The process of transforming the load model into a state-space equation form and simplifying it to complete the construction of the nonlinear gray box model specifically includes: Based on the third-order induction motor equations of the ZIP+IM model, the active power and reactive power of the load nodes, the load model is transformed into the form of state-space equations, thus obtaining the state-space equations. The active power and reactive power of the load node are given by the following formula: Where P represents the active power of the load node, Q represents the reactive power of the load node, and V represents the voltage of the load node. ZIP Q represents the active power of the static load. ZIP P represents the reactive power of the static load. M Q represents the active power absorbed by the induction motor. M P represents the reactive power absorbed by the induction motor. Z P I P P Q represents the percentage of active power in the constant impedance, constant voltage, and constant power components of the static model, respectively. Z Q I Q P P represents the percentage of reactive power in the constant impedance, constant voltage, and constant power components of the static model, respectively. ZIP0 Q represents the active power of the static model in steady state. ZIP0 V0 represents the reactive power of the static model under steady state, and X represents the rated voltage of the load node. m ′ represents transient reactance, E m ' represents transient potential, δ m The power angle representing the transient potential; The state-space equation is as follows: Among them, X m T represents the magnetizing reactance. dm ' represents the transient time constant, ω m ω represents the mechanical angular velocity of the induction motor load. s T represents the angular frequency of the load node. m H represents the torque after load equivalent. m Represents the inertial time constant; The state-space equations are simplified to obtain a nonlinear gray box model; The preprocessed power system disturbance data and the initial values ​​of each parameter to be identified are input into the nonlinear gray box model to achieve parameter identification of the load model.

2. The parameter identification method according to claim 1, characterized in that, The nonlinear gray box model is as follows: Among them, P i Let i represent the parameter to be identified, i = 1, 2, ..., 11.

3. The parameter identification method according to claim 1, characterized in that, The acquisition of preprocessed power system disturbance data specifically includes: At the load node, a single-phase fault for a first preset time and a three-phase fault for a second preset time are set respectively. Multiple sets of voltage, frequency, active power and reactive power are collected during a third preset time, including before the fault, the fault segment and after the fault is cleared. The input data are voltage and frequency, and the output data are active power and reactive power. High-order harmonics and noise are filtered out from each set of input and output data to obtain preprocessed power system disturbance data.

4. The parameter identification method according to claim 1, characterized in that, After inputting the preprocessed perturbation data and the initial value of each parameter to be identified into the nonlinear gray box model, the Levenberg-Marquardt iterative algorithm is used to update each parameter to be identified.

5. A parameter identification system for a load model, characterized in that, The system for implementing the parameter identification method according to any one of claims 1-4, the system comprising: The acquisition unit is used to acquire preprocessed power system disturbance data; The construction unit is used to transform the load model into the form of state-space equations and simplify it, thereby completing the construction of the nonlinear gray box model, which includes multiple parameters to be identified. The identification unit is used to input the preprocessed power system disturbance data and the initial value of each parameter to be identified into the nonlinear gray box model to realize the parameter identification of the load model.

6. A computer device comprising a processor and a memory for storing a processor-executable program, characterized in that, When the processor executes the program stored in the memory, it implements the parameter identification method according to any one of claims 1-4.

7. A storage medium storing a program, characterized in that, When the program is executed by the processor, it implements the parameter identification method according to any one of claims 1-4.

Images