VSC DC power supply system small signal stability analysis method and system

Abstract

The invention discloses a VSC DC power supply system small signal stability analysis method and system, and belongs to the field of DC power transmission. A power supply basic structure of a VSC direct current power supply system is determined; constructing a stability analysis module based on an impedance equivalent model by performing impedance modeling and simplified equivalence based on a power supply basic structure; performing power supply data sampling, judging a power supply scene, performing power supply stability evaluation in combination with the impedance equivalent model, and determining a system stability coefficient. According to the method, the technical problem that the stability analysis efficiency and accuracy are insufficient when an existing direct current power supply system faces complex loads and operation conditions is solved, and the technical effect of improving the stability evaluation precision and efficiency of the VSC direct current power supply system through impedance modeling and small signal analysis is achieved.

Description

Technical Field

[0001] The present invention relates to the technical field of DC power transmission, and particularly to a method and system for analyzing the small-signal stability of a DC power supply system of a VSC. Background Art

[0002] With the transformation of the global energy structure and the rapid development of renewable energy, the application of DC power transmission technology (HVDC) in power systems is becoming increasingly widespread. Especially in cross-regional power transmission and the integration of large-scale renewable energy sources such as wind power and photovoltaic power, the DC power transmission system (VSC-HVDC) has become one of the important power transmission solutions due to its good control performance and adaptability. However, in practical applications, when the VSC DC power supply system faces different loads and complex power environments, its stability problems gradually become prominent, especially the stability of the system under small-signal disturbances. Traditional small-signal stability analysis methods usually ignore the impedance characteristics of different devices in the DC power supply system and their mutual influence, resulting in inaccurate analysis results. Especially when the load changes greatly, it is difficult to find a balance between efficiency and accuracy. Summary of the Invention

[0003] The present invention provides a method and system for analyzing the small-signal stability of a DC power supply system of a voltage source converter (VSC) to solve the technical problem of insufficient efficiency and accuracy of stability analysis in existing DC power supply systems when facing complex loads and operating conditions.

[0004] The first aspect embodiment of the present invention provides a method for analyzing the small-signal stability of a DC power supply system of a VSC, including the following steps:

[0005] Determine the basic power supply structure of the VSC DC power supply system;

[0006] According to the basic power supply structure, model and simplify the impedance characteristics of each device in the VSC DC power supply system to construct an impedance equivalent model, use a multi-layer perceptron to construct a stability analysis module, and embed the impedance equivalent model into the stability analysis module;

[0007] Collect the real-time power supply data of the VSC DC power supply system, analyze the current power supply scenario, activate the impedance equivalent model according to the current power supply scenario, and use the stability analysis module to evaluate the power supply stability to determine the system stability coefficient of the VSC DC power supply system.

[0008] Optionally, in an embodiment of the present invention, the basic power supply structure includes a device connection structure and an impedance characteristic relationship. The device connection structure defines the positions of each device in the VSC DC power supply system and the connection manner therebetween. The impedance characteristic relationship is a linear relationship between the device control strategy and the impedance characteristics. According to the basic power supply structure, an impedance model is established for the impedance characteristics of each device in the VSC DC power supply system, and an impedance equivalent model is constructed by simplifying and equivalentizing the impedance model. The simplification and equivalentization process includes a primary simplification and equivalentization based on the stability influence degree and a secondary simplification and equivalentization under a constant power load. The specific simplification and equivalentization process includes:

[0009] Set the simplification and equivalentization conditions based on the stability influence degree, and perform a primary equivalent simplification on the impedance model according to the simplification and equivalentization conditions to obtain a first impedance equivalent model;

[0010] Taking the constant power load as a constraint scenario, perform a secondary equivalent simplification on the first impedance equivalent model according to the simplification and equivalentization conditions to obtain a second impedance equivalent model;

[0011] Integrate the first impedance equivalent model and the second impedance equivalent model to construct the impedance equivalent model.

[0012] Optionally, in an embodiment of the present invention, the simplification and equivalentization conditions include a structure condition and a parameter condition. Performing a primary equivalent simplification on the impedance model according to the simplification and equivalentization conditions to obtain a first impedance equivalent model includes:

[0013] Determine the device stability influence threshold according to the structure condition, and determine the parameter stability influence threshold and the preset frequency band threshold according to the parameter condition;

[0014] Perform a structural simplification and equivalentization on the impedance model according to the device stability influence threshold to obtain a one-step impedance equivalent model;

[0015] Perform a parameter simplification and equivalentization on the one-step impedance equivalent model according to the parameter stability influence threshold to obtain a two-step impedance equivalent model;

[0016] Perform a frequency band simplification and equivalentization on the two-step impedance equivalent model according to the preset frequency band threshold to obtain the first impedance equivalent model, where the frequency band simplification and equivalentization is to filter the impedance characteristics of high-frequency components with the preset frequency band threshold as a constraint.

[0017] Optionally, in an embodiment of the present invention, the structural simplification and equivalentization includes reactive device screening and simplification and series-parallel equivalentization, and the parameter simplification and equivalentization includes reactive parameter screening and simplification and parameter order reduction equivalentization. Among them, in the simplification and equivalentization process, reactive judgment constraints are performed based on the device stability influence threshold and the parameter stability influence threshold to determine the devices and parameters for simplification and equivalentization.

[0018] Optionally, in an embodiment of the present invention, a simulation is performed with a constant power load to obtain the device stability influence degree, parameter stability influence degree, and device frequency under the constraint scenario of the constant power load, and the second impedance equivalent model is obtained through secondary equivalent simplification with the same simplification equivalent process as the first impedance equivalent model.

[0019] Optionally, in an embodiment of the present invention, the impedance equivalent model is associated with a small-signal transfer block diagram, specifically including:

[0020] Determine the signal transfer nodes according to the impedance equivalent model, where the signal transfer nodes at least include interface nodes and device nodes, and the signal transfer nodes correspond to transfer linear relationships;

[0021] Construct a small-signal transfer block diagram for the signal transfer nodes, where the small-signal transfer block diagram characterizes the dynamic behavior under signal perturbation.

[0022] Optionally, in an embodiment of the present invention, the signal transfer nodes include linear nodes and non-linear nodes, including:

[0023] Traverse the signal transfer nodes and extract the non-linear nodes;

[0024] Perform approximate linearization of the discrete distribution for the non-linear nodes to determine the approximate linear relationship;

[0025] Establish a mapping between the approximate linear relationship and the non-linear nodes and add it to the transfer linear relationship.

[0026] Optionally, in an embodiment of the present invention, a stability analysis module is constructed using a multi-layer perceptron, including:

[0027] Collect the historical power supply data set and historical system stability coefficient set of the VSC DC power supply system, input the historical power supply data into the constructed impedance equivalent model for simplified response, identify the sample equivalent data set, use the historical system stability coefficient set to label the sample equivalent data set, and divide the labeling result to construct a sample training set and a sample validation set;

[0028] Use a multi-layer perceptron to construct the stability analysis module, including an input layer, a hidden layer, an output layer, etc. Input the sample training set into the initialized multi-layer perceptron model for forward propagation, perform layer-by-layer transfer through the input layer, hidden layer, and output layer, etc., calculate the predicted system stability coefficient of the VSC DC power supply system, calculate the loss value between the predicted system stability coefficient and the sample data using the mean square error loss function, and calculate the gradient of the loss with respect to the weights of each layer through backpropagation layer by layer;

[0029] The Adam optimizer is used to optimize the parameters of the multi-layer perceptron model, adjusting the weights to minimize the value of the loss function until the maximum number of iterations is reached or the loss value converges;

[0030] After the training is completed, the performance of the multi-layer perceptron model is tested using the sample validation set, and the accuracy of the multi-layer perceptron model in predicting the system stability coefficient is evaluated. When the accuracy meets the expected target, the current multi-layer perceptron model is output as the final stability analysis module.

[0031] Optionally, in an embodiment of the present invention, activating the impedance equivalent model according to the current power supply scenario includes:

[0032] Judging based on whether it is a constant power load scenario, and activating the first impedance equivalent model or the second impedance equivalent model in the stability analysis module according to the current power supply scenario; wherein, if the current power supply scenario is a constant power load scenario, the second impedance equivalent model in the impedance equivalent model is activated, and if the power supply scenario is not a constant power load scenario, the first impedance equivalent model in the impedance equivalent model is activated.

[0033] An embodiment of the second aspect of the present invention provides a small-signal stability analysis system for a VSC DC power supply system, including:

[0034] A determination module for determining the basic power supply structure of the VSC DC power supply system;

[0035] A construction module for modeling and simplifying the impedance characteristics of each device in the VSC DC power supply system according to the basic power supply structure to construct an impedance equivalent model, using a multi-layer perceptron to construct a stability analysis module, and embedding the impedance equivalent model into the stability analysis module;

[0036] An analysis module for collecting real-time power supply data of the VSC DC power supply system, analyzing the current power supply scenario, activating the impedance equivalent model according to the current power supply scenario, and using the stability analysis module to evaluate the power supply stability to determine the system stability coefficient of the VSC DC power supply system.

[0037] The small-signal stability analysis method and system for the DC power supply system of VSC in the embodiments of the present invention analyze by clarifying the basic power supply structure of the VSC DC power supply system, perform impedance modeling and simplified equivalence, and obtain a more refined secondary simplified equivalent model through a two-step simplified equivalence method. This step-by-step simplification process not only reduces the computational complexity but also ensures the accuracy and efficiency of the impedance equivalent model. By real-time sampling of power supply data and judgment of power supply scenarios, an appropriate impedance equivalent model can be selected for stability assessment according to different load conditions; finally, by combining the work of all these links, a stability coefficient can be effectively calculated, providing a scientific basis for the stability of the VSC DC power supply system. The application of this method not only improves the accuracy of system analysis but also enhances the efficiency of stability assessment of the power supply system, providing strong support for stability control in practical applications.

[0038] Additional aspects and advantages of the present invention will be given in part in the following description, become apparent in part from the following description, or be learned by practice of the present invention. Brief Description of the Drawings

[0039] The above and / or additional aspects and advantages of the present invention will become apparent and be readily understood from the following description of the embodiments in conjunction with the drawings, where:

[0040] Figure 1 is a flowchart of a method for analyzing the small-signal stability of a DC power supply system of VSC according to an embodiment of the present invention;

[0041] Figure 2 is a block diagram of a system for analyzing the small-signal stability of a DC power supply system of VSC according to an embodiment of the present invention. Detailed Embodiments

[0042] Embodiments of the present invention will be described in detail below. Examples of the embodiments are shown in the drawings, where the same or similar reference numerals denote the same or similar elements or elements having the same or similar functions throughout. The embodiments described below with reference to the drawings are exemplary and are intended to explain the present invention and should not be construed as limiting the present invention.

[0043] Figure 1 is a flowchart of a method for analyzing the small-signal stability of a DC power supply system of VSC according to an embodiment of the present invention.

[0044] As Figure 1 shown, the method for analyzing the small-signal stability of the DC power supply system of VSC includes the following steps:

[0045] In step S101, determine the basic power supply structure of the VSC DC power supply system.

[0046] Optionally, in an embodiment of the present invention, the basic power supply structure includes a device connection structure and an impedance characteristic relationship. The device connection structure defines the position of each device in the VSC DC power supply system and the connection manner between them, and the impedance characteristic relationship is a linear relationship between the device control strategy and the impedance characteristic.

[0047] Specifically, in the embodiment of the present invention, in the VSC DC power supply system, first, the basic power supply structure is determined. The basic power supply structure mainly includes two key parts, namely, the device connection structure and the impedance characteristic relationship. Among them, the device connection structure describes how each device in the VSC DC power supply system is connected and coordinated. The VSC DC power supply system usually consists of multiple devices, such as voltage source converters (VSCs), DC cables, inverters, transformers, loads, etc. These devices are electrically connected to form a complete power supply network. The device connection structure defines the position of each device in the VSC DC power supply system and the connection manner between them, which determines the transmission path of the power flow and the working mode of the system. The impedance characteristic relationship describes the key factors of the response of the devices in the VSC DC power supply system to current and voltage. The impedance characteristic of each device will affect the stability and performance of the entire VSC DC power supply system. In the VSC DC power supply system, the impedance characteristic usually shows how the device responds to the change of current or voltage. For example, the control strategy of the voltage source converter (VSC) will change its internal impedance, thereby affecting the overall impedance characteristic of the VSC DC power supply system. This impedance characteristic relationship is generated through the interaction between the control strategy of the device and its own electrical properties. In addition, the impedance characteristic relationship is a linear relationship, which means that the change between the device control strategy and its impedance characteristic is linearly correlated. This linear relationship simplifies the modeling and analysis of the VSC DC power supply system because the influence of the control strategy on the device impedance can be described by a simple mathematical formula, making the stability analysis of the VSC DC power supply system clearer and easier to calculate. In summary, determining the basic power supply structure is to clarify how the devices are connected to each other and how these devices affect their impedance characteristics through the control strategy, thereby affecting the dynamic performance of the entire VSC DC power supply system.

[0048] In step S102, according to the basic power supply structure, the impedance characteristic of each device in the VSC DC power supply system is modeled and a simplified equivalent impedance equivalent model is constructed. A stability analysis module is constructed by using a multi-layer perceptron, and the impedance equivalent model is built into the stability analysis module.

[0049] Specifically, in the embodiments of the present invention, after determining the basic power supply structure of the VSC DC power supply system, a stability analysis module based on an impedance equivalent model is constructed through impedance modeling and simplified equivalence. The core of this process is to convert the complex impedance characteristics in the VSC DC power supply system into a more easily analyzable model through simplification, thereby assisting in stability assessment.

[0050] First, model the impedance characteristics of each device (such as voltage source converters, cables, loads, etc.) in the VSC DC power supply system. The impedance model of each device reflects its response behavior to current and voltage changes. These impedance models connect the dynamic behaviors of the entire VSC DC power supply system and provide a basis for subsequent stability analysis. Since the actual VSC DC power supply system contains a large number of complex components, directly analyzing the impedance of all components in detail is often computationally intensive and complex. Therefore, through the simplified equivalence method, the complex impedance relationship can be converted into a relatively simple equivalent model, which is convenient for stability analysis.

[0051] Optionally, in an embodiment of the present invention, according to the basic power supply structure, model the impedance characteristics of each device in the VSC DC power supply system to obtain an impedance model, and perform simplified equivalence on the impedance model to construct an impedance equivalent model. Among them, the simplified equivalence process includes a primary simplified equivalence based on the stability influence degree and a secondary simplified equivalence under constant power loads. The specific simplified equivalence process includes:

[0052] Set the simplified equivalence conditions based on the stability influence degree, and perform a primary equivalent simplification on the impedance model according to the simplified equivalence conditions to obtain a first impedance equivalent model;

[0053] Taking the constant power load as a constraint scenario, perform a secondary equivalent simplification on the first impedance equivalent model according to the simplified equivalence conditions to obtain a second impedance equivalent model;

[0054] Integrate the first impedance equivalent model and the second impedance equivalent model to construct an impedance equivalent model.

[0055] Specifically, in an embodiment of the present invention, when simplifying and equilibrating, it is first necessary to construct an impedance model of the VSC DC power supply system, that is, to mathematically model the impedance characteristics such as resistance and reactance of each device (such as voltage source converter, cable, load, etc.) in the VSC DC power supply system. The process of constructing the impedance model is as follows: First, for the voltage source converter, it is necessary to obtain the influence of its internal control strategy and switching frequency on the impedance characteristics. The input impedance of the voltage source converter usually depends on its regulation mode, such as current mode or voltage mode control, and the change of the external load. Through circuit analysis and control strategy modeling, the equivalent impedance of the VSC can be derived. This process usually requires the use of frequency domain analysis and small signal models to solve its response to voltage and current changes. Secondly, for the cable part, it is necessary to obtain the transmission characteristics of the cable, including resistance, capacitance and inductance. The impedance characteristics of the cable are mainly determined by its length, structure and frequency characteristics of the current. At low frequencies, the impedance of the cable is pure resistance, while at high frequencies, due to the influence of inductance and capacitance, the impedance of the cable will change. In the model, these characteristics are usually described by transmission line theory, taking into account the distributed parameters and transmission losses of the cable, so as to derive its comprehensive impedance. For loads, especially constant power loads, the impedance characteristics of the load are closely related to the working state of the load. The characteristic of constant power loads is that their power demand does not change with voltage changes, so the relationship between their voltage and current is usually nonlinear. When constructing the model, the impedance characteristics of the load can be processed by linearization or approximation methods to facilitate subsequent stability analysis. In the entire impedance modeling process, the coupling effect between various devices needs to be considered. The impedance characteristics of different devices affect each other. For example, the connection between VSC and cable will interact with each other, and its impedance characteristics cannot be simply considered independently. It is necessary to determine the impedance parameters of each device through the combination of circuit model and control strategy, and merge them into an overall impedance model to provide a basis for subsequent stability analysis. Through the above steps, the impedance model finally constructed can fully describe the response of each device in the VSC DC power supply system to current and voltage changes, and provide a reliable theoretical basis for subsequent stability evaluation and control strategy optimization.

[0056] After constructing the impedance model, simplified equivalent conditions based on the stability influence degree are set. This stability influence degree refers to the degree of influence of each device on the stability of the VSC DC power supply system in the VSC DC power supply system, and is obtained by simulating the VSC DC power supply system through simulation and experimental data in combination with the actual working conditions. This simplified equivalent condition is used to simplify the complexity of the model while maintaining the key characteristics of the VSC DC power supply system. The simplified equivalent condition includes a structural condition and a parameter condition. Among them, the structural condition refers to which devices in the VSC DC power supply system play a greater role in influencing stability and which can be ignored. These conditions help to determine which devices need to be modeled in detail and which can be simplified. The parameter condition is a simplified condition determined according to the parameters and working conditions of the device. For example, some parameters may have a relatively small influence on the stability of the VSC DC power supply system (such as the control delay of the converter, the parameters of redundant circuit elements, etc.), so they can be ignored or combined.

[0057] Based on the above simplified equivalent conditions, the impedance model is equivalently simplified once. In this step, devices with relatively small influence on the stability of the VSC DC power supply system are screened out according to the stability influence degree and frequency band of the devices, and these devices are equivalently simplified to generate a first impedance equivalent model. This model is relatively simplified, but still can reflect the influence of key devices and components on the stability in the VSC DC power supply system. Subsequently, under the constraint condition of a constant power load, the first impedance equivalent model is equivalently simplified a second time. In the constant power load scenario, the load demand is fixed, and this condition can help to further simplify the model. In this process, the equivalent simplification method is the same as the first equivalent simplification, and is also carried out through the stability influence degree and frequency band. The second simplification mainly removes the possible redundant parts in the first equivalent model, making the model more concise, while ensuring that the stability characteristics under the constant power load are retained. Through this process, a second impedance equivalent model is obtained. Finally, the first impedance equivalent model and the second impedance equivalent model are integrated to form a final impedance equivalent model. This model combines the key information extracted from the first and second simplifications and can accurately reflect the stability characteristics of the VSC DC power supply system under different load conditions, providing a basis for subsequent stability analysis.

[0058] Optionally, in an embodiment of the present invention, the simplified equivalent condition includes a structural condition and a parameter condition. Equivalently simplifying the impedance model once according to the simplified equivalent condition to obtain a first impedance equivalent model includes:

[0059] Determining the device stability influence threshold according to the structural condition, and determining the parameter stability influence threshold and the preset frequency band threshold according to the parameter condition;

[0060] The impedance model is structurally simplified and equivalent according to the device stability influence threshold to obtain a one-step impedance equivalent model;

[0061] The one-step impedance equivalent model is parameter-simplified and equivalent according to the parameter stability influence threshold to obtain a two-step impedance equivalent model;

[0062] The two-step impedance equivalent model is frequency-band simplified and equivalent according to the preset frequency-band threshold to obtain a first impedance equivalent model, where the frequency-band simplification and equivalence is to filter the impedance characteristics of high-frequency components with the preset frequency-band threshold as a constraint.

[0063] Specifically, in the embodiments of the present invention, when constructing the impedance equivalent model of the VSC DC power supply system, through a series of simplification steps, combined with the equipment stability influence threshold, the parameter stability influence threshold, and the preset frequency band threshold, the characteristics that have a greater impact on the stability of the VSC DC power supply system are accurately extracted, thereby improving the efficiency and calculation speed of subsequent modules. Specifically, first, according to the structural conditions, the influence degree of each device on the system stability is evaluated, and the equipment stability influence threshold is set in combination with historical experience and expert suggestions. Similarly, by analyzing the operating parameters of each device (such as resistance, inductance, capacitance, frequency, etc.), the parameter stability influence threshold and the preset frequency band threshold are set in combination with historical experience and expert suggestions. Subsequently, based on the set equipment stability influence threshold, the impedance model is simplified and equivalent in structure. In this process, the equipment stability influence degree of each device is extracted from the simulated stability influence degree, and the equipment stability influence degree is compared with the equipment stability influence threshold. Those devices with an influence degree less than the equipment stability influence threshold are screened out, and relevant parameters are removed or series-parallel equivalent from the impedance model according to the screening results, obtaining a one-step impedance equivalent model. This one-step impedance equivalent model can reduce the complexity of the VSC DC power supply system without significantly losing the stability characteristics of the VSC DC power supply system. Then, based on the parameter stability influence threshold, the one-step impedance equivalent model is further simplified. In this process, the parameter stability influence degree of each device is extracted from the simulated stability influence degree, and the parameter stability influence degree is compared with the parameter stability influence threshold. Those devices with an influence degree less than the parameter stability influence threshold are screened out, and relevant parameters are removed or parameter order reduction equivalent from the one-step impedance equivalent model according to the screening results, obtaining a two-step impedance equivalent model. This two-step impedance equivalent model further simplifies the complexity of the VSC DC power supply system compared with the one-step impedance equivalent model, removes those less important parameters, and retains the core characteristics required for the stability analysis of the VSC DC power supply system. Then, based on the preset frequency band threshold, the two-step impedance equivalent model is simplified in frequency band. The purpose of the frequency band simplification is to remove those high-frequency components that are unimportant in terms of frequency and retain the key characteristics in the low-frequency and medium-frequency ranges. In this process, the device frequency obtained by simulation for each device is compared with the preset frequency band threshold, those devices with a frequency greater than the preset frequency band threshold are screened out, and relevant parameters are removed from the two-step impedance equivalent model according to the screening results, obtaining a first impedance equivalent model. This first impedance equivalent model can effectively remove the irrelevant high-frequency part, obtain a more concise model, and at the same time ensure the retention of the stability characteristics of the key frequency band, providing support for subsequent stability analysis.

[0064] Optionally, in an embodiment of the present invention, simulation is performed with a constant power load to obtain the equipment stability influence degree, parameter stability influence degree, and equipment frequency under the constraint scenario of the constant power load, and a second impedance equivalent model is obtained through secondary equivalent simplification with the same simplified equivalent process as the first impedance equivalent model.

[0065] Specifically, in the embodiment of the present invention, for the secondary equivalent simplification, the specific steps are the same as those of the above-mentioned primary equivalent simplification, except that the equipment stability influence degree, parameter stability influence degree, and equipment frequency used are obtained based on the simulation under the constant power load.

[0066] Optionally, in an embodiment of the present invention, the structural simplified equivalent includes the reactive power equipment screening simplification and the series-parallel equivalent, and the parameter simplified equivalent includes the reactive power parameter screening simplification and the parameter order reduction equivalent. Among them, in the process of simplified equivalent, reactive power determination constraints are performed based on the equipment stability influence threshold and the parameter stability influence threshold to determine the equipment and parameters for simplified equivalent.

[0067] Specifically, in the embodiments of the present invention, when constructing and simplifying the impedance model of the VSC DC power supply system, in order to improve the calculation efficiency and retain the key stability characteristics, the strategies of structural simplification equivalence and parameter simplification equivalence are adopted. Specifically, the simplification process can be achieved through methods such as reactive device screening and simplification, series-parallel equivalence, reactive parameter screening and simplification, and parameter order reduction equivalence. The selection of these simplification methods is based on the device stability influence threshold and the parameter stability influence threshold to ensure that the information crucial to the stability of the VSC DC power supply system is not lost during simplification. For the reactive device screening and simplification, it is to remove the relevant parameters of the devices whose device stability influence degree is less than the second device stability influence threshold in the device stability influence threshold, reducing unnecessary calculations. For the series-parallel equivalence, it is to perform impedance equivalence for the devices whose device stability influence degree is less than the first device stability influence threshold but greater than or equal to the second device stability influence threshold in the device stability influence threshold according to their series or parallel connection, and replace the relevant parameters in the model with the equivalent impedance parameters, thereby reducing the complexity of the model. For the reactive parameter screening and simplification, it is to remove the relevant parameters of the devices whose parameter stability influence degree is less than the second parameter stability influence threshold in the parameter stability influence threshold, reducing unnecessary calculations. For the parameter order reduction equivalence, it is to reduce the dimension or merge the devices whose parameter stability influence degree is less than the first parameter stability influence threshold but greater than or equal to the second parameter stability influence threshold through order reduction. For example, the control strategy of the voltage source converter may involve multiple parameters, but only a few of them have a significant impact on the dynamic stability of the VSC DC power supply system. Through parameter order reduction, those parameters with less influence can be removed, thereby simplifying the model and improving the calculation efficiency. Based on the device stability influence threshold and the parameter stability influence threshold, the purpose of the reactive determination constraint is to ensure that only the devices and parameters that play an important role in the stability analysis are modeled, while the devices and parameters with less influence are simplified, ensuring that while minimizing unnecessary parameters, the characteristics crucial to the stability analysis of the VSC DC power supply system are still retained.

[0068] Optionally, in an embodiment of the present invention, the impedance equivalent model is associated with a small-signal transfer block diagram, specifically including:

[0069] Determine the signal transfer nodes according to the impedance equivalent model, where the signal transfer nodes at least include interface nodes and device nodes, and the signal transfer nodes correspond to transfer linear relationships;

[0070] Construct a small-signal transfer block diagram for the signal transfer nodes, where the small-signal transfer block diagram characterizes the dynamic behavior under signal perturbation.

[0071] Specifically, in the embodiments of the present invention, in order to make the analysis more comprehensive, the simplified impedance equivalent model is also combined with a small-signal transfer block diagram, which can accurately characterize the dynamic response behavior under small-signal disturbances and provide more accurate stability analysis results. This small-signal transfer block diagram is a graphical tool for describing how the VSC DC power supply system responds to small disturbances. Through this transfer block diagram, it is possible to clearly show how small-signal disturbances propagate through various devices and nodes in the VSC DC power supply system and affect the stability of the entire VSC DC power supply system.

[0072] In the stability analysis of the VSC DC power supply system, constructing signal transfer nodes and a small-signal transfer block diagram is to accurately characterize the dynamic response behavior of the VSC DC power supply system under signal disturbances. Specifically, by identifying the impedance equivalent model, the signal transfer nodes can be determined. Signal transfer nodes refer to the key points where electrical signals (such as voltage, current, etc.) are transferred in the VSC DC power supply system, including but not limited to interface nodes and device nodes. These nodes are the interfaces of various devices and components in the VSC DC power supply system, and through these nodes, signals can flow or be transferred in the system. Among them, interface nodes refer to the nodes in the VSC DC power supply system that connect to external signal sources or loads, such as input interface nodes, output interface nodes, etc. Device nodes refer to the connection points of different devices or components in the VSC DC power supply system. Each device (such as a voltage source converter, filter, cable, load, etc.) has its corresponding device node, and these nodes determine how the electrical characteristics of the device transfer signals. Each signal transfer node has its corresponding linear relationship, which means that the signal changes passing through these nodes are linear and can be represented by a transfer function, usually in the following form: Among them, H(s) is the transfer function, which represents the relationship between the input signal and the output signal. Vin(s) and Vout(s) are the Laplace transforms of the input and output signals respectively. Subsequently, according to the settings of the signal transfer nodes, the relationships between these nodes are connected through the transfer function to form a small-signal transfer block diagram. This small-signal transfer block diagram is a tool used to describe the dynamic behavior of the VSC DC power supply system under small disturbances or small-signal conditions, showing the dynamic relationship from the input signal (such as voltage, current) to the output of the VSC DC power supply system, usually including the interaction between multiple device nodes and signal gain or attenuation. In summary, through the construction of signal transfer nodes and the small-signal transfer block diagram, the dynamic behavior of the VSC DC power supply system under small disturbances can be accurately described. The signal transfer nodes help define the signal transfer path and gain relationship in the VSC DC power supply system, while the small-signal transfer block diagram is used to specifically characterize the dynamic response of the VSC DC power supply system. In this way, it is possible to deeply understand how the VSC DC power supply system responds to external disturbances and provide an important basis for stability analysis.

[0073] Optionally, in an embodiment of the present invention, the signal transfer nodes include linear nodes and non-linear nodes, including:

[0074] Traverse the signal transfer nodes and extract the non-linear nodes;

[0075] For the non-linear nodes, perform approximate linearization of the discrete distribution to determine the approximate linear relationship;

[0076] Establish the mapping between the approximate linear relationship and the non-linear nodes and add it to the transfer linear relationship.

[0077] Specifically, in the embodiments of the present invention, in the stability analysis of the VSC DC power supply system, the signal transfer nodes include not only linear nodes but also nonlinear nodes. To accurately describe the dynamic behavior of the VSC DC power supply system under small-signal disturbances, the nonlinear nodes must be processed so that they can be effectively combined and analyzed with the linear nodes in the transfer block diagram. First, all signal transfer nodes are traversed, and through simulation, it can be determined which device characteristics are nonlinear. Once the nonlinear nodes are identified, these nodes are approximately linearized. Since the actual nonlinear relationship is relatively complex, multiple possible operating points (representing different operating states or input signal conditions) are statistically analyzed through a discrete distribution method, and the behavior of the nonlinear nodes near each operating point is approximated by Taylor series expansion to obtain multiple local linearization models. Subsequently, these local linearization models are combined through interpolation methods or weighted averaging to determine the approximate linear relationship of the approximate nonlinear node. After completing the approximate linearization, a mapping relationship is established for each nonlinear node and the approximate linear relationship to convert its nonlinear behavior into a linear relationship, and this mapping relationship clearly describes the linear approximation method of the nonlinear node under different operating points. Finally, all the obtained mapping relationships are added to the transfer linear relationship to form a complete small-signal transfer block diagram together with the original linear nodes, providing a basis for subsequent stability analysis.

[0078] Optionally, in an embodiment of the present invention, a stability analysis module is constructed using a multi-layer perceptron, including:

[0079] Collect the historical power supply data set and the historical system stability coefficient set of the VSC DC power supply system, input the historical power supply data into the constructed impedance equivalent model for simplified response, identify the sample equivalent data set, label the sample equivalent data set using the historical system stability coefficient set, and partition the labeling results to construct a sample training set and a sample validation set;

[0080] Use a multi-layer perceptron to construct a stability analysis module, including an input layer, a hidden layer, an output layer, etc. Input the sample training set into the initialized multi-layer perceptron model for forward propagation, and calculate the predicted system stability coefficient of the VSC DC power supply system by passing through the input layer, hidden layer, and output layer layer by layer. Calculate the loss value between the predicted system stability coefficient and the sample data using the mean square error loss function, and calculate the gradient of the loss with respect to the weights of each layer layer by layer through backpropagation;

[0081] Use the Adam optimizer to optimize the parameters of the multi-layer perceptron model, adjust the weights to minimize the value of the loss function until the maximum number of iterations is reached or the loss value converges;

[0082] After the training is completed, the performance of the multi-layer perceptron model is tested using the sample validation set, and the accuracy of the multi-layer perceptron model in predicting the system stability coefficient is evaluated. When the accuracy meets the expected goal, the current multi-layer perceptron model is output as the final stability analysis module.

[0083] Specifically, in the embodiment of the present invention, a historical power supply data set and a historical system stability coefficient set are collected from the historical database. The historical power supply data is input into the constructed impedance equivalent model for simplified response to identify a sample equivalent data set, including the variation relationship between voltage and current, load change rate, equivalent power, equivalent impedance, etc. Then, the historical system stability coefficient set is used to label the sample equivalent data set, and the labeled results are divided to construct a sample training set and a sample validation set. Subsequently, a stability analysis module is constructed using a multi-layer perceptron (MLP), including an input layer, a hidden layer, an output layer, etc. The sample training set is input into the initialized MLP model for forward propagation, and is passed layer by layer through the input layer, hidden layer, and output layer, etc., to calculate the predicted system stability coefficient of the VSC DC power supply system. Then, the mean square error (MSE) loss function is used to calculate the loss value between the predicted system stability coefficient and the sample data, and the gradient of the loss with respect to the weights of each layer is calculated layer by layer through backpropagation. After that, the Adam optimizer is used to optimize the model parameters, adjust the weights to minimize the value of the loss function, and repeat the above process until the maximum number of iterations is reached or the loss value converges. After the training is completed, the performance of the model is tested using the sample validation set, and the accuracy of the model in predicting the system stability coefficient is evaluated. If the accuracy meets the expected goal, the current MLP model is output as the final stability analysis module. Otherwise, hyperparameters such as the learning rate and the number of neurons in the hidden layer are adjusted to further improve the stability prediction effect. Finally, the impedance equivalent model is built into the stability analysis module to jointly perform a more accurate dynamic behavior analysis of the VSC DC power supply system under small signal disturbances, providing data support for subsequent stability evaluation.

[0084] In step S103, the real-time power supply data of the VSC DC power supply system is collected, the current power supply scenario is analyzed, the impedance equivalent model is activated according to the current power supply scenario, and the power supply stability is evaluated using the stability analysis module to determine the system stability coefficient of the VSC DC power supply system.

[0085] Specifically, in an embodiment of the present invention, when evaluating the power supply stability, first, real-time operating state data of the VSC DC power supply system is collected through power supply data sampling to obtain the power supply data of the VSC DC power supply system, including voltage, current, load status, control signals, etc. Subsequently, based on the sampled power supply data, the current power supply scenario is analyzed. If the power supply scenario is a constant power load scenario, the second impedance equivalent model in the impedance equivalent model is activated to simplify the response of the power supply data and identify the equivalent data, including the variation relationship between voltage and current, the load change rate, the equivalent power, and the equivalent impedance. If the power supply scenario is not a constant power load scenario, the first impedance equivalent model is activated to perform the same operation. After that, the equivalent data is transmitted to the stability analysis module for power supply stability evaluation, and the stability coefficient of the VSC DC power supply system is calculated. A higher stability coefficient indicates that the VSC DC power supply system can quickly stabilize and effectively cope with external disturbances, while a lower stability coefficient indicates that the VSC DC power supply system is vulnerable to the influence of disturbances and may have the risk of instability.

[0086] Optionally, in an embodiment of the present invention, activating the impedance equivalent model according to the current power supply scenario includes:

[0087] Judging based on whether it is a constant power load scenario, and activating the first impedance equivalent model or the second impedance equivalent model in the stability analysis module according to the current power supply scenario; wherein, if the current power supply scenario is a constant power load scenario, the second impedance equivalent model in the impedance equivalent model is activated, and if the power supply scenario is not a constant power load scenario, the first impedance equivalent model in the impedance equivalent model is activated.

[0088] Specifically, in the embodiments of the present invention, when evaluating the power supply stability, the power supply data is first analyzed to determine the current power supply scenario. The determination of the power supply scenario is mainly based on the load type. For example, if the current load state of the VSC DC power supply system shows the characteristics of a constant power load (i.e., the load power is constant and does not change with the voltage), the current scenario can be determined as a constant power load scenario. The constant power load has certain response characteristics to the changes in voltage and current, and a specific model needs to be used for analysis. Subsequently, according to the determined power supply scenario, a suitable target impedance equivalent model is selected for further analysis. In the stability evaluation of the VSC DC power supply system, the impedance equivalent model is used to simplify the electrical characteristics of the VSC DC power supply system and help analyze the dynamic behavior of the VSC DC power supply system under specific working conditions. When the power supply scenario is a constant power load scenario, the second impedance equivalent model in the stability analysis module is activated as the target impedance equivalent model. Otherwise, the first impedance equivalent model is activated as the target impedance equivalent model. Then, based on this target impedance equivalent model and combined with the stability analysis module, the power supply stability evaluation is carried out through the same steps as described above to calculate the system stability coefficient of the VSC DC power supply system, ensuring that the VSC DC power supply system can operate stably under different loads and working conditions.

[0089] Next, a small-signal stability analysis system for the DC power supply system of the VSC according to an embodiment of the present invention will be described with reference to the accompanying drawings.

[0090] Figure 2 It is a block diagram of the small-signal stability analysis system for the DC power supply system of the VSC according to the embodiment of the present invention.

[0091] As Figure 2 shown, the small-signal stability analysis system 10 for the DC power supply system of the VSC includes: a determination module 100, a construction module 200, and an analysis module 300.

[0092] Among them, the determination module 100 is used to determine the basic power supply structure of the VSC DC power supply system. The construction module 200 is used to model and simplify the impedance characteristics of each device in the VSC DC power supply system according to the basic power supply structure to construct an impedance equivalent model, use a multi-layer perceptron to construct a stability analysis module, and embed the impedance equivalent model into the stability analysis module. The analysis module 300 is used to collect the real-time power supply data of the VSC DC power supply system, analyze the current power supply scenario, activate the impedance equivalent model according to the current power supply scenario, and use the stability analysis module to evaluate the power supply stability to determine the system stability coefficient of the VSC DC power supply system.

[0093] Optionally, in an embodiment of the present invention, the power supply basic structure includes a device connection structure and an impedance characteristic relationship. The device connection structure defines the positions of each device in the VSC DC power supply system and the connection manner between them. The impedance characteristic relationship is a linear relationship between the device control strategy and the impedance characteristics. According to the power supply basic structure, the impedance characteristics of each device in the VSC DC power supply system are modeled to obtain an impedance model, and the impedance model is simplified and equivalent to construct an impedance equivalent model. Among them, the simplification and equivalent process includes a primary simplification and equivalent based on the stability influence degree and a secondary simplification and equivalent under a constant power load. The specific simplification and equivalent process includes:

[0094] Set the simplification and equivalent conditions based on the stability influence degree, and perform a primary equivalent simplification on the impedance model according to the simplification and equivalent conditions to obtain a first impedance equivalent model;

[0095] Taking the constant power load as a constraint scenario, perform a secondary equivalent simplification on the first impedance equivalent model according to the simplification and equivalent conditions to obtain a second impedance equivalent model;

[0096] Integrate the first impedance equivalent model and the second impedance equivalent model to construct an impedance equivalent model.

[0097] Optionally, in an embodiment of the present invention, the simplification and equivalent conditions include a structure condition and a parameter condition. Performing a primary equivalent simplification on the impedance model according to the simplification and equivalent conditions to obtain a first impedance equivalent model includes:

[0098] Determine the device stability influence threshold according to the structure condition, and determine the parameter stability influence threshold and the preset frequency band threshold according to the parameter condition;

[0099] Perform a structural simplification and equivalent on the impedance model according to the device stability influence threshold to obtain a one-step impedance equivalent model;

[0100] Perform a parameter simplification and equivalent on the one-step impedance equivalent model according to the parameter stability influence threshold to obtain a two-step impedance equivalent model;

[0101] Perform a frequency band simplification and equivalent on the two-step impedance equivalent model according to the preset frequency band threshold to obtain a first impedance equivalent model, where the frequency band simplification and equivalent is to filter the impedance characteristics of high-frequency components with the preset frequency band threshold as a constraint.

[0102] Optionally, in an embodiment of the present invention, the structural simplification and equivalent includes reactive device screening and simplification and series-parallel equivalence, and the parameter simplification and equivalent includes reactive parameter screening and simplification and parameter order reduction equivalence. Among them, in the simplification and equivalent process, reactive determination constraints are performed based on the device stability influence threshold and the parameter stability influence threshold to determine the devices and parameters for simplification and equivalence.

[0103] Optionally, in an embodiment of the present invention, simulation is performed with a constant power load to obtain the equipment stability influence degree, parameter stability influence degree, and equipment frequency under the constraint scenario of the constant power load, and a second impedance equivalent model is obtained through secondary equivalent simplification with the same simplification equivalent process as the first impedance equivalent model.

[0104] Optionally, in an embodiment of the present invention, the impedance equivalent model is associated with a small-signal transfer block diagram, specifically including:

[0105] Determine signal transfer nodes according to the impedance equivalent model, where the signal transfer nodes at least include interface nodes and equipment nodes, and the signal transfer nodes correspond to transfer linear relationships;

[0106] Construct a small-signal transfer block diagram for the signal transfer nodes, where the small-signal transfer block diagram characterizes the dynamic behavior under signal perturbation.

[0107] Optionally, in an embodiment of the present invention, the signal transfer nodes include linear nodes and non-linear nodes, including:

[0108] Traverse the signal transfer nodes and extract non-linear nodes;

[0109] For the non-linear nodes, perform approximate linearization of the discrete distribution to determine the approximate linear relationship;

[0110] Establish a mapping between the approximate linear relationship and the non-linear nodes and add it to the transfer linear relationship.

[0111] Optionally, in an embodiment of the present invention, a stability analysis module is constructed using a multi-layer perceptron, including:

[0112] Collect the historical power supply data set and historical system stability coefficient set of the VSC DC power supply system, input the historical power supply data into the constructed impedance equivalent model for simplified response, identify the sample equivalent data set, label the sample equivalent data set using the historical system stability coefficient set, and divide the labeled results to construct a sample training set and a sample validation set;

[0113] Use a multi-layer perceptron to construct a stability analysis module, including an input layer, a hidden layer, an output layer, etc. Input the sample training set into the initialized multi-layer perceptron model for forward propagation, perform layer-by-layer transfer through the input layer, hidden layer, and output layer, etc., calculate the predicted system stability coefficient of the VSC DC power supply system, calculate the loss value between the predicted system stability coefficient and the sample data using the mean square error loss function, and calculate the gradient of the loss with respect to the weights of each layer layer-by-layer through backpropagation;

[0114] Use the Adam optimizer to optimize the parameters of the multi-layer perceptron model, adjust the weights to minimize the value of the loss function until the maximum number of iterations is reached or the loss value converges;

[0115] After the training is completed, use the sample validation set to test the performance of the multi-layer perceptron model, and evaluate the accuracy of the multi-layer perceptron model in predicting the system stability coefficient. When the accuracy meets the expected target, the current multi-layer perceptron model is used as the final stability analysis module for output.

[0116] Optionally, in an embodiment of the present invention, activating the impedance equivalent model according to the current power supply scenario includes:

[0117] Determine whether it is a constant power load scenario. According to the current power supply scenario, activate the first impedance equivalent model or the second impedance equivalent model in the stability analysis module. Among them, if the current power supply scenario is a constant power load scenario, activate the second impedance equivalent model in the impedance equivalent model. If the power supply scenario is not a constant power load scenario, activate the first impedance equivalent model in the impedance equivalent model.

[0118] It should be noted that the foregoing explanation of the embodiments of the small-signal stability analysis method for the VSC DC power supply system also applies to the small-signal stability analysis system of the VSC DC power supply system in this embodiment, and will not be repeated here.

[0119] According to the small-signal stability analysis method and system of the VSC DC power supply system proposed in the embodiments of the present invention, determine the basic power supply structure, construct a stability analysis module based on the impedance equivalent model, and perform stability evaluation by analyzing the power supply scenario through power supply data sampling. In the process of simplifying the equivalent, through setting the simplification conditions of the stability influence degree, perform the primary and secondary simplifications of the impedance model, and then form the first impedance equivalent model and the second impedance equivalent model. Simplify the structure and parameters of the impedance model, and perform reactive power determination constraints based on the equipment stability influence threshold and the parameter stability influence threshold. By associating the small-signal transfer block diagram, construct the signal transfer node and analyze the small-signal transfer block diagram to further improve the stability analysis accuracy. At the same time, combined with the power supply scenario determination, activate the target impedance equivalent model, perform stability evaluation on the VSC DC power supply system, and determine the system stability coefficient. Thus, the technical problem of the insufficient efficiency and accuracy of the stability analysis of the existing DC power supply system in the face of complex loads and operating conditions is solved, and the technical effect of improving the accuracy and efficiency of the stability evaluation of the VSC DC power supply system through impedance modeling and small-signal analysis is achieved.

[0120] In the description of this specification, the descriptions referring to terms such as "one embodiment", "some embodiments", "example", "specific example", or "some examples", etc. mean that the specific features, structures, materials, or characteristics described in connection with the embodiment or example are included in at least one embodiment or example of the present invention. In this specification, the schematic representations of the above terms do not necessarily refer to the same embodiment or example. Moreover, the specific features, structures, materials, or characteristics described may be combined in any one or N embodiments or examples in a suitable manner. In addition, without contradiction, those skilled in the art can combine and combine the different embodiments or examples described in this specification and the features of different embodiments or examples.

[0121] In addition, the terms "first" and "second" are used for descriptive purposes only and should not be construed as indicating or implying relative importance or implicitly specifying the quantity of the indicated technical features. Thus, the features defined with "first" and "second" may explicitly or implicitly include at least one such feature. In the description of the present invention, the meaning of "N" is at least two, such as two, three, etc., unless otherwise specifically defined.

[0122] Any process or method description shown in the flowchart or described in other ways herein may be understood to represent a module, segment, or portion of code including one or N executable instructions for implementing a customized logical function or process, and the scope of the preferred embodiments of the present invention includes additional implementations, where the functions may be performed in a substantially simultaneous manner or in a reverse order according to the functions involved, rather than in the order shown or discussed, which should be understood by those skilled in the art to which the embodiments of the present invention pertain.

Claims

1. A small-signal stability analysis method for a DC power supply system of a VSC, characterized in that It includes the following steps: Determine the basic power supply structure of the VSC DC power supply system; According to the basic power supply structure, model the impedance characteristics of each device in the VSC DC power supply system, simplify and equivalently construct an impedance equivalent model, use a multi-layer perceptron to construct a stability analysis module, and embed the impedance equivalent model into the stability analysis module; Collect the real-time power supply data of the VSC DC power supply system, analyze the current power supply scenario, activate the impedance equivalent model according to the current power supply scenario, and use the stability analysis module to evaluate the power supply stability to determine the system stability coefficient of the VSC DC power supply system.

2. The method according to claim 1, wherein The basic power supply structure includes a device connection structure and an impedance characteristic relationship. The device connection structure defines the position of each device in the VSC DC power supply system and the connection method between them. The impedance characteristic relationship is the linear relationship between the device control strategy and the impedance characteristics. According to the basic power supply structure, model the impedance characteristics of each device in the VSC DC power supply system to obtain an impedance model, and simplify and equivalently construct an impedance equivalent model. Among them, the simplification and equivalent process includes a primary simplification and equivalent based on the stability influence degree and a secondary simplification and equivalent under a constant power load. The specific simplification and equivalent process includes: Set the simplification and equivalent conditions based on the stability influence degree, and perform a primary equivalent simplification on the impedance model according to the simplification and equivalent conditions to obtain a first impedance equivalent model; Taking the constant power load as a constraint scenario, perform a secondary equivalent simplification on the first impedance equivalent model according to the simplification and equivalent conditions to obtain a second impedance equivalent model; Integrate the first impedance equivalent model and the second impedance equivalent model to construct the impedance equivalent model.

3. The method according to claim 2, wherein The simplification and equivalent conditions include a structure condition and a parameter condition. Performing a primary equivalent simplification on the impedance model according to the simplification and equivalent conditions includes: Determine the device stability influence threshold according to the structure condition, and determine the parameter stability influence threshold and the preset frequency band threshold according to the parameter condition; Perform a structural simplification and equivalent on the impedance model according to the device stability influence threshold to obtain a one-step impedance equivalent model; Perform a parameter simplification and equivalent on the one-step impedance equivalent model according to the parameter stability influence threshold to obtain a two-step impedance equivalent model; Perform a frequency band simplification and equivalent on the two-step impedance equivalent model according to the preset frequency band threshold to obtain the first impedance equivalent model, where the frequency band simplification and equivalent is to filter the impedance characteristics of high-frequency components with the preset frequency band threshold as a constraint.

4. The method according to claim 3, wherein The structural simplification and equivalent includes reactive device screening and simplification and series-parallel equivalent, and the parameter simplification and equivalent includes reactive parameter screening and simplification and parameter order reduction equivalent. Among them, in the simplification and equivalent process, reactive determination constraints are performed based on the device stability influence threshold and the parameter stability influence threshold to determine the devices and parameters for simplification and equivalent.

5. The method according to claim 3, wherein Perform simulations with a constant - power load to obtain the equipment stability influence degree, parameter stability influence degree, and equipment frequency under the constraint scenario of the constant - power load. Perform secondary equivalent simplification on the second impedance equivalent model through the same simplified equivalent process as the first impedance equivalent model.

6. The method according to claim 1, wherein The impedance equivalent model is associated with a small - signal transfer block diagram, specifically including: Determine signal transfer nodes according to the impedance equivalent model, where the signal transfer nodes at least include interface nodes and equipment nodes, and the signal transfer nodes correspond to transfer linear relationships; Construct a small - signal transfer block diagram for the signal transfer nodes, where the small - signal transfer block diagram characterizes the dynamic behavior under signal perturbation.

7. The method according to claim 6, characterized in that, The signal transfer nodes include linear nodes and non - linear nodes, including: Traverse the signal transfer nodes and extract non - linear nodes; Perform approximate linearization of the discrete distribution for the non - linear nodes to determine approximate linear relationships; Establish the mapping between the approximate linear relationships and the non - linear nodes and add them to the transfer linear relationships.

8. The method according to claim 1, wherein Use a multi - layer perceptron to construct a stability analysis module, including: Collect the historical power supply data set and historical system stability coefficient set of the VSC DC power supply system. Input the historical power supply data into the constructed impedance equivalent model for simplified response, identify the sample equivalent data set, label the sample equivalent data set using the historical system stability coefficient set, and divide the labeled results to construct a sample training set and a sample validation set; Use a multi - layer perceptron to construct the stability analysis module, including an input layer, a hidden layer, an output layer, etc. Input the sample training set into the initialized multi - layer perceptron model for forward propagation, perform layer - by - layer transfer through the input layer, hidden layer, and output layer, etc., calculate the predicted system stability coefficient of the VSC DC power supply system, calculate the loss value between the predicted system stability coefficient and the sample data using the mean - square error loss function, and calculate the gradient of the loss with respect to the weights of each layer through backpropagation layer by layer; Use the Adam optimizer to optimize the parameters of the multi - layer perceptron model, adjust the weights to minimize the value of the loss function until the maximum number of iterations is reached or the loss value converges; After training, use the sample validation set to test the performance of the multi - layer perceptron model, evaluate the accuracy of the multi - layer perceptron model in predicting the system stability coefficient. When the accuracy meets the expected target, output the current multi - layer perceptron model as the final stability analysis module.

9. The method according to claim 2, wherein Activate the impedance equivalent model according to the current power supply scenario, including: Determine whether it is a constant - power load scenario. According to the current power supply scenario, activate the first impedance equivalent model or the second impedance equivalent model in the stability analysis module; where if the current power supply scenario is a constant - power load scenario, activate the second impedance equivalent model in the impedance equivalent model, and if the power supply scenario is not a constant - power load scenario, activate the first impedance equivalent model in the impedance equivalent model.

10. A small-signal stability analysis system for a DC power supply system of a VSC, characterized in that, Include: A determination module for determining the basic power supply structure of the VSC DC power supply system; A building block for modeling and simplifying the equivalent impedance characteristics of each device in a VSC DC power supply system according to the basic power supply structure, constructing a stability analysis module using a multi-layer perceptron, and embedding the impedance equivalent model into the stability analysis module; An analysis module for collecting real-time power supply data of a VSC DC power supply system, analyzing the current power supply scenario, activating the impedance equivalent model according to the current power supply scenario, and using the stability analysis module to evaluate the power supply stability to determine the system stability coefficient of the VSC DC power supply system.

Images