Full-control current source converter dc transmission line fault detection method and device

Abstract

The application provides a full-control current source converter DC transmission line fault detection method and device, relates to the high-voltage DC transmission technical field, and the method comprises the following steps: acquiring the actual first-end voltage, actual first-end current and actual last-end voltage of the full-control current source converter DC transmission line; taking the acquired actual first-end current and actual last-end voltage as inputs, outputting the predicted first-end voltage of the full-control current source converter DC transmission line through a preset digital twin reference model; determining whether the full-control current source converter DC transmission line is faulty based on correlation analysis on the actual first-end voltage and the predicted first-end voltage; wherein the digital twin reference model is used to represent the mapping relationship between the first-end current, last-end voltage and first-end voltage in the time domain. The application effectively improves the accuracy and efficiency of full-control current source converter DC transmission line fault identification.

Description

Technical Field

[0001] This application relates to the field of high voltage direct current transmission technology, specifically to a fault detection method for DC transmission lines using a fully controlled current source converter, a fault detection device for DC transmission lines using a fully controlled current source converter, a control method for DC transmission lines using a fully controlled current source converter, a control device for DC transmission lines using a fully controlled current source converter, a converter system, a machine-readable storage medium, and a terminal device. Background Technology

[0002] High-voltage direct current (HVDC) transmission technology plays a crucial role in renewable energy grid integration and inter-regional power grid interconnection due to its advantages in long-distance and large-capacity transmission. Full-controllable composite converters (F3C), as a novel DC transmission topology, possess stronger active power transmission capabilities and flexible control characteristics, but fault detection in their DC lines still faces challenges. Currently, fault detection methods for DC transmission lines mainly rely on traveling wave protection and differential undervoltage protection as primary protection, supplemented by current differential protection and low-voltage protection as backup protection. However, these methods have the following limitations: 1) Traveling wave protection detects faults based on the voltage or current traveling wave signal generated at the moment of the fault, judging the fault by measuring the time difference or amplitude change of the traveling wave arriving at both ends of the line. However, the traveling wave signal is easily affected by line distributed parameters, noise, and high-frequency attenuation, resulting in decreased sensitivity under high-resistance faults or complex network structures, requiring extremely high sampling frequencies. 2) Differential undervoltage protection judges faults by detecting the DC line voltage mutation rate and voltage amplitude drop. However, this method may fail to operate under high transition resistance faults; simultaneously, the voltage change rate is easily affected by line capacitive current, leading to malfunctions and making it difficult to meet high reliability requirements. 3) Current differential protection adopts the idea of ​​Kirchhoff's current law, judging the fault by comparing the current difference between the two ends of the line. Theoretically, it has absolute selectivity and is applicable to various fault types. However, due to the influence of line distributed capacitive current, its speed of operation needs to be improved. 4) Undervoltage protection is triggered by the voltage drop of the DC bus after a fault, which is simple to implement and does not require complex calculations. However, its sensitivity is affected by the fluctuation of the system operating voltage. Under high resistance faults at remote ends, the voltage drop is not obvious, which may lead to protection failure to operate.

[0003] In summary, existing protection methods all suffer from insufficient adaptability, and there is an urgent need for a new protection scheme that can integrate multiple information and adapt to dynamic operating conditions. Summary of the Invention

[0004] The purpose of this application is to provide a method and apparatus for fault detection of DC transmission lines using a fully controlled current source converter, in order to solve the problem that the existing technology is easily affected by system operation, resulting in insufficient adaptability.

[0005] To achieve the above objectives, the first aspect of this application provides a fault detection method for DC transmission lines using a fully controlled current source converter, comprising:

[0006] Obtain the actual starting voltage, actual starting current, and actual ending voltage of a DC transmission line with a fully controlled current source converter.

[0007] Using the actual head current and actual tail voltage as inputs, the predicted head voltage of the fully controlled current source converter DC transmission line is output through a preset digital twin reference model.

[0008] The correlation analysis between the actual and predicted head-end voltages is used to determine whether the DC transmission line of the fully controlled current source converter is faulty.

[0009] The digital twin reference model is used to represent the time-domain mapping relationship between the start-up current, the end-up voltage, and the start-up voltage.

[0010] A second aspect of this application provides a control method for a fully controlled current source converter DC transmission line, comprising:

[0011] If a fault is determined in the DC transmission line of the fully controlled current source converter using the above-mentioned fault detection method, the fully controlled current source converter is locked, and / or the DC transmission line of the fully controlled current source converter is disconnected.

[0012] A third aspect of this application provides a fault detection device for a fully controlled current source converter DC transmission line, employing the fault detection method for a fully controlled current source converter DC transmission line as described above. The device comprises:

[0013] The data acquisition module is configured to acquire the actual starting voltage, actual starting current, and actual ending voltage of the DC transmission line of the fully controlled current source converter.

[0014] The first-end voltage prediction module is configured to take the actual first-end current and actual end voltage as inputs, and output the predicted first-end voltage of the fully controlled current source converter DC transmission line through a preset digital twin reference model.

[0015] The digital twin reference model is used to represent the time-domain mapping relationship between the start-up current, the end-up voltage, and the start-up voltage.

[0016] The fault detection module is configured to determine whether the DC transmission line of the fully controlled current source converter is faulty based on the correlation analysis between the actual head-end voltage and the predicted head-end voltage.

[0017] A fourth aspect of this application provides a control device for a fully controlled current source converter DC transmission line, comprising:

[0018] The control module is configured to, if a fault is determined in the DC transmission line of the fully controlled current source converter using the aforementioned fault detection method, control the fully controlled current source converter to lock out, and / or control the DC transmission line of the fully controlled current source converter to disconnect.

[0019] A fifth aspect of this application provides a converter system, comprising:

[0020] A fully controlled current source converter; and a controller communicatively connected to the fully controlled current source converter;

[0021] The controller is configured to perform either the fault detection method for a fully controlled current source converter DC transmission line as described above, or the control method for a fully controlled current source converter DC transmission line as described above.

[0022] In a sixth aspect, this application provides a machine-readable storage medium storing instructions that, when executed by a processor, configure the processor to perform either the fully controlled current source converter DC transmission line fault detection method described above, or the fully controlled current source converter DC transmission line control method described above.

[0023] In a seventh aspect, this application provides a terminal device including a memory, a processor, and a computer program stored in the memory and executable on the processor. When the processor executes the computer program, it implements the steps of the fault detection method for a fully controlled current source converter DC transmission line as described above, or executes the control method for a fully controlled current source converter DC transmission line as described above.

[0024] This application establishes a two-terminal frequency domain relationship based on a frequency-dependent model, representing the linear mapping relationship between the starting-end voltage and current and the ending voltage and current of a fully controlled current source converter DC transmission line. This relationship is then converted into a single-input-output relationship with the starting-end current and ending voltage as inputs and the starting-end voltage as the output. Based on this single-input-output relationship, a digital twin reference model of the fully controlled current source converter DC transmission line is constructed. This allows for the extraction of electrical information from both ends of the transmission line, and the predicted starting-end voltage can be obtained through the digital twin reference model. By performing correlation analysis between the predicted starting-end voltage and the actual starting-end voltage, the presence of a fault in the fully controlled current source converter DC transmission line can be determined, effectively improving the accuracy and efficiency of fault identification in fully controlled current source converter DC transmission lines.

[0025] Other features and advantages of the embodiments of this application will be described in detail in the following detailed description section. Attached Figure Description

[0026] The accompanying drawings are provided to further illustrate the embodiments of this application and form part of the specification. They are used together with the following detailed description to explain the embodiments of this application, but do not constitute a limitation on the embodiments of this application. In the drawings:

[0027] Figure 1 A flowchart illustrating the method for fault detection of DC transmission lines using a fully controlled current source converter, as provided in a preferred embodiment of this application.

[0028] Figure 2 A schematic diagram of a digital twin reference model provided for a preferred embodiment of this application;

[0029] Figure 3 A schematic diagram of a digital twin actual model provided for a preferred embodiment of this application;

[0030] Figure 4 A schematic diagram of the fault detection device for a fully controlled current source converter DC transmission line provided in a preferred embodiment of this application;

[0031] Figure 5 A schematic diagram of a terminal device provided for a preferred embodiment of this application.

[0032] Explanation of reference numerals in the attached figures

[0033] 10 - Terminal device, 100 - Processor, 101 - Memory, 102 - Computer program. Detailed Implementation

[0034] To make the objectives, technical solutions, and advantages of the embodiments of this application clearer, the technical solutions of the embodiments of this application will be clearly and completely described below with reference to the accompanying drawings. It should be understood that the specific embodiments described herein are only for illustration and explanation of the embodiments of this application and are not intended to limit the embodiments of this application. All other embodiments obtained by those skilled in the art based on the embodiments of this application without creative effort are within the scope of protection of this application.

[0035] It should be noted that the technical solutions of the various embodiments of this application can be combined with each other, but only if they are based on the ability of those skilled in the art to implement them. When the combination of technical solutions is contradictory or cannot be implemented, it should be considered that such combination of technical solutions does not exist and is not within the scope of protection claimed by this application.

[0036] Understandably, the F3C-HVDC DC transmission system was proposed in the context of high-proportion renewable energy access and is suitable for renewable energy transmission scenarios. However, after renewable energy access, the fault characteristics of DC lines change, and the original protection schemes cannot be directly applied. Moreover, the original protection schemes themselves have inherent defects. Therefore, it is necessary to fully combine the renewable energy access scenario with the new characteristics of F3C converters to detect new DC line faults in order to reliably realize the protection of the F3C-HVDC DC transmission system.

[0037] To solve the above problems, such as Figure 1 As shown, the first aspect of this application provides a fault detection method for DC transmission lines using a fully controlled current source converter, comprising:

[0038] S100: Obtain the actual starting voltage, actual starting current, and actual ending voltage of the DC transmission line of the fully controlled current source converter.

[0039] S200, taking the actual head current and actual tail voltage as input, outputs the predicted head voltage of the fully controlled current source converter DC transmission line through the preset digital twin reference model;

[0040] Among them, the digital twin reference model is used to represent the time-domain mapping relationship between the start-up current, the end-up voltage, and the start-up voltage;

[0041] S300. Based on the correlation analysis between the actual and predicted head-end voltages, determine whether the DC transmission line of the fully controlled current source converter is faulty.

[0042] Thus, this application only needs to extract the electrical information at both ends of the transmission line, and the predicted head-end voltage of the transmission line can be obtained through the digital twin reference model. By performing correlation analysis between the predicted head-end voltage and the actual head-end voltage collected in real time, it is possible to determine whether there is a fault in the DC transmission line of the fully controlled current source converter, which effectively improves the accuracy and efficiency of fault identification of the DC transmission line of the fully controlled current source converter.

[0043] The digital twin reference model is constructed through the following steps: Based on the frequency-dependent model, a two-terminal frequency domain relationship is established to represent the linear mapping relationship between the head-end voltage, head-end current and the terminal voltage, terminal current of the DC transmission line of the fully controlled current source converter. The two-terminal frequency domain relationship is converted into a single-input output relationship with the head-end current and terminal voltage as inputs and the head-end voltage as output. Based on the head-end current, terminal voltage and head-end voltage under fault-free operation, the single-input output relationship is fitted into a digital twin reference model in the time domain.

[0044] Before step S100, such as Figure 2As shown, this application pre-constructs a digital twin reference model of a fully controlled current source converter DC transmission line, namely the F3C-HVDC DC transmission system. Specifically, the digital twin reference model is constructed through the following steps: First, a digital twin model of the F3C-HVDC DC transmission system is established, and the parameter changes of the DC transmission line are simulated using a frequency-dependent model. Specifically, the frequency-dependent model can be calculated and solved using the Marti model. It is understood that the calculation process of the frequency-dependent model is existing technology, and it is not limited here. This application uses the frequency-dependent model to solve for the time-domain analytical expression of the line-head voltage of the F3C-HVDC DC transmission system under fault-free conditions, and uses this expression as the digital twin reference model. Specifically, frequency-dependent models are used to simulate parameter variations in DC transmission lines, and a two-terminal frequency domain relationship is established to represent the linear mapping relationship between the head-end voltage and current and the terminal voltage and current of a fully controlled current source converter DC transmission line. This two-terminal frequency domain relationship is then converted into a single-input-output relationship with the head-end current and terminal voltage as inputs and the head-end voltage as the output. Based on the head-end current, terminal voltage, and head-end voltage under fault-free operation, the single-input-output relationship is fitted into a digital twin reference model in the time domain. Simultaneously, such as Figure 3 As shown, the real-time measured head-end voltage of the F3C-HVDC DC transmission system is used as the actual model, and fault detection is performed by comparing it with a reference model. This is understandable, as... Figure 2 and Figure 3 As shown, the first end refers to the side closer to the rectifier station, that is, the connection point between the DC output terminal of the converter and the first end of the line, and the last end refers to the side closer to the inverter station, that is, the connection point between the end of the line and the DC input terminal of the inverter.

[0045] In this application, a two-terminal frequency domain relationship is established based on a frequency-dependent model to represent the linear mapping relationship between the starting-end voltage and starting-end current and the ending-end voltage and ending-end current of a fully controlled current source converter DC transmission line, including:

[0046] Based on the frequency-dependent model, the frequency-dependent characteristics of resistance, inductance, conductance, and capacitance per unit length in a fully controlled current source converter DC transmission line are determined. Based on these frequency-dependent characteristics, the relationships between the traveling wave, reverse traveling wave, terminal voltage, and terminal current at both ends of the fully controlled current source converter DC transmission line in the frequency domain are determined. Specifically, the Marti model is used to accurately simulate the frequency-dependent characteristics of resistance R(ω), inductance L(ω), conductance G(ω), and capacitance C(ω) per unit length, and based on this, a precise distributed parameter model describing the voltage and current relationship between the beginning (k-end) and end (m-end) of the line is established in the frequency domain.

[0047] The construction of the distributed parameter model includes: establishing a first mapping relationship between the traveling wave, reverse traveling wave, terminal voltage, and terminal current at both ends of a fully controlled current source converter DC transmission line based on the frequency domain; establishing a second mapping relationship between the first-end voltage and the terminal voltage and current of the fully controlled current source converter DC transmission line; and establishing a two-terminal frequency domain relationship based on the first and second mapping relationships to represent the linear mapping relationship between the first-end voltage and the terminal voltage and current of the fully controlled current source converter DC transmission line.

[0048] Specifically, the general solution of the distributed parameter model in the frequency domain can be expressed as a combination of a traveling wave and an anti-traveling wave, that is, the two-terminal frequency domain relationship can be expressed as:

[0049]

[0050] in, Let be the propagation constant. R(ω) is the characteristic impedance, in Ω; L(ω) is the resistance per unit length, in rad / m; G(ω) is the inductance per unit length; C(ω) is the conductance per unit length; and l is the length of the transmission line, in meters. This is the voltage at the beginning of the transmission line. This refers to the voltage at the end of the transmission line. This refers to the current at the beginning of the transmission line. Let ω represent the terminal current of the transmission line, j be an imaginary number, ω represent the angular frequency, cosh represent the hyperbolic cosine function, and sinh represent the hyperbolic sine function. These functions are used to describe the voltage-current relationship of distributed parameter lines.

[0051] Understandably, the method of this application can be applied to protection devices, for example, deployed in the controller of a protection device. To facilitate real-time calculations in the protection device, this application transforms the above-mentioned two-terminal frequency domain relationship model into one based on the first-terminal current I. k (ω) and terminal voltage V m (ω) represents the input, and the starting voltage V is... kThe single-input single-output model is defined by (ω). The transformation of the two-terminal frequency domain relationship into a single-input output relationship with the first-terminal current and the last-terminal voltage as inputs and the first-terminal voltage as the output includes: based on the second mapping relationship, using the last-terminal current of the fully controlled current-source converter DC transmission line as an intermediate quantity, the first mapping relationship is transformed into a third mapping relationship between the first-terminal voltage and the last-terminal voltage and the first-terminal current of the fully controlled current-source converter DC transmission line; based on the third mapping relationship, the first-terminal current and the last-terminal voltage of the fully controlled current-source converter DC transmission line are mapped to the current frequency domain response and voltage frequency domain response of the first-terminal voltage, respectively, resulting in a single-input output relationship with the first-terminal voltage as the output.

[0052] Specifically, by combining the expressions for the two-terminal frequency domain relationships and eliminating I... m (ω) yields the frequency domain transfer function in the following form, where the conversion process is not limited here. That is, the single-input-output relationship model of this application can be expressed as:

[0053]

[0054] in, , , where tanh represents the ratio of the hyperbolic sine function to the hyperbolic cosine function, and sech represents the reciprocal of the hyperbolic cosine function.

[0055] If the time-domain impulse response is obtained by directly performing inverse Fourier transform on H1(ω) and H2(ω) and then convolving it, the computational load is enormous and unsuitable for real-time protection. Therefore, this application fits the frequency domain response of the entire system through a system identification method. In this application, based on the first-end current, the last-end voltage, and the first-end voltage under fault-free operation, the single-input-output relationship is fitted into a digital twin reference model in the time domain, including:

[0056] S210. Continuously acquire the starting-end voltage, starting-end current, and ending-end voltage of the fully controlled current source converter DC transmission line under fault-free operation. Under the fault-free normal operation of the F3C-HVDC system, continuously acquire a large amount of {I k (ω),V m (ω),V k (ω)} data pairs, where a sliding overlap sampling strategy can be used during data acquisition to enhance data continuity and avoid transient loss.

[0057] S220. Based on the acquired data pairs, the current frequency domain response and voltage frequency domain response of the first-terminal voltage are fitted using a parameter identification algorithm, such as the Prony algorithm, to express the current frequency domain response and voltage frequency domain response of the first-terminal voltage as a sum of complex exponential terms, thus obtaining the fitting result. This application introduces an adaptive order selection mechanism on the basis of the traditional Prony algorithm. Specifically, during fitting, the initial order M0 is set to 10, and gradually increased to M... max =30, and the AIC criterion is used to select the optimal order to avoid overfitting. During fitting, a regularization term is introduced to improve fitting stability. After introducing the regularization term, the frequency domain response fitting objective can be expressed as:

[0058]

[0059] in, This is the regularization strength parameter.

[0060] S230. The fitted results are subjected to operations such as inverse Discrete Fourier transform to obtain the current-time domain impulse response and voltage-time domain impulse response of the first-terminal voltage. This yields a digital twin reference model of the single-input-output relationship in the time domain, while naturally suppressing errors caused by high-frequency noise through the fitting process. Furthermore, after fitting, cross-validation is used to divide the data into training and test sets to ensure the model's generalization ability. The model is considered effective when the fitting error is <2%.

[0061] Specifically, the digital twin reference model can be represented as:

[0062]

[0063] in, This represents the predicted head-end voltage output by the digital twin reference model. This represents the actual current at the beginning of the circuit. Let represent the actual terminal voltage, t represent time t, and e be the natural constant. , , , , , , , and All are model coefficients, among which, , Let represent the amplitude coefficients of the m / n-th oscillation component, respectively. , Let represent the attenuation coefficients of the m / n-th oscillation component, respectively. , These represent the angular frequencies of the m / n-th oscillation component, respectively. , Let M and N represent the initial phases of the m / n-th oscillation component, respectively, and M and N represent the fitting order, the specific values ​​of which can be determined by the model accuracy requirements. , , , , , , , and The coefficients can be predetermined. For example, they can be calculated offline in advance through the Prony identification process described above when the system is fault-free. They collectively represent the comprehensive dynamic characteristics of the specific line under the healthy operating environment of a specific F3C system.

[0064] This application uses the Marti model to simulate the real-world changes in line parameters with frequency (R(ω), L(ω), G(ω), C(ω)), establishing precise frequency domain relationships and ensuring high accuracy from the outset. However, directly using frequency-dependent models for time-domain simulation suffers from slow computation speed, failing to meet protection speed requirements. This application employs a system identification method to construct a corresponding time-domain analytical expression, i.e., a digital twin model, based on precise data generated by the frequency-dependent model or field-measured data. The digital twin model reflects the physical laws described by the frequency-dependent model and boasts significantly faster computation speed.

[0065] After constructing the digital twin reference model of the F3C-HVDC system, fault detection of the F3C-HVDC system can be performed through the digital twin reference model. In step S100, parameters such as the actual head-end voltage, actual head-end current, and actual terminal voltage of the DC transmission line can be collected in real time through pre-deployed sensors. In step S200, the actual head-end current and actual terminal voltage of the DC transmission line are used as inputs to the digital twin reference model, and the predicted head-end voltage of the DC transmission line under the current actual head-end current and actual terminal voltage is output through the digital twin reference model, that is, the head-end voltage under the fault-free normal operation state of the system.

[0066] In step S300, determining whether a DC transmission line with a fully controlled current source converter is faulty is based on correlation analysis between the actual and predicted head-end voltages, including:

[0067] S310. Extract the multimodal features of the actual head-end voltage and the multimodal features of the predicted head-end voltage, and determine the initial weights of each multimodal feature. The multimodal features include at least time-domain features, frequency-domain features, and wavelet time-frequency features. Understandably, at each sampling point, parameters such as the actual head-end voltage can be sampled multiple times at a preset sampling frequency. In this application, time-domain features include, but are not limited to, mean, variance, and peak value; frequency-domain features include, but are not limited to, spectral energy and dominant frequency components; and wavelet time-frequency features are used to capture transient fault features and can be extracted through wavelet transform. The multimodal feature extraction process for the voltage signal is existing technology and is not limited here.

[0068] S320. Determine the current system operating status of the fully controlled current source converter DC transmission line based on the actual head-end voltage and actual head-end current of the fully controlled current source converter DC transmission line, specifically including:

[0069] S321. Obtain the rated voltage, rated power, and rated operating frequency of the grid-connected AC system of the fully controlled current source converter DC transmission line, and determine the actual operating frequency of the grid-connected side of the fully controlled current source converter DC transmission line. The rated voltage, rated power, and rated operating frequency of the grid-connected AC system of the transmission line can be predetermined. The actual operating frequency of the grid-connected side of the DC transmission line can be directly detected by a measuring device or calculated through phase tracking; this is not limited here.

[0070] S322. Determine the actual head-end voltage fluctuation of a fully controlled current source converter DC transmission line based on the difference between the actual head-end voltage and the rated voltage. The actual power of a fully controlled current source converter DC transmission line is determined based on the actual head-end voltage and actual head-end current. For example, the head-end voltage and current are sampled on the HVDC DC side, and the instantaneous power, i.e., the actual power, is calculated according to P=UI. The actual power fluctuation of the fully controlled current source converter DC transmission line is determined based on the difference between the actual power and the rated power. ,in, This represents the actual power fluctuation, where P represents the actual power. Indicates the rated power.

[0071] S323. Determine the voltage fluctuation rate of a fully controlled current source converter DC transmission line based on the ratio of the actual first-end voltage fluctuation to the rated voltage. ,Right now The load variation rate of a fully controlled current source converter DC transmission line is determined based on the ratio of actual power fluctuation to rated power. ,Right now The frequency offset Δ on the grid-connected side of the DC transmission line of the fully controlled current source converter is determined based on the difference between the actual operating frequency and the rated operating frequency. f The signal-to-noise ratio of a fully controlled current-source converter DC transmission line is determined based on the actual head-end voltage or actual head-end current of the line. SNR Understandably, the signal-to-noise ratio of DC transmission lines can be calculated using existing algorithms such as the residual signal method, the generalized integral method, and wavelet noise estimation, and this is not a limitation here.

[0072] S324. Determine the current system operating status of the DC transmission line with a fully controlled current source converter based on the voltage fluctuation rate, load change rate, frequency offset, and signal-to-noise ratio.

[0073] S330. Based on the current system operating status of the DC transmission line with a fully controlled current source converter, the initial weights of each multi-mode feature of the actual head-end voltage and the predicted head-end voltage are adjusted to obtain the real-time weights of each multi-mode feature, and the corresponding multi-mode features are weighted using the real-time weights of each multi-mode feature.

[0074] Specifically, based on the current system operating state of the fully controlled current source converter DC transmission line, the initial weights of each multi-mode feature of the actual and predicted head-end voltages are adjusted to obtain the real-time weights of each multi-mode feature. This includes: when the voltage fluctuation rate of the fully controlled current source converter DC transmission line is greater than the voltage fluctuation rate threshold, or when the load change rate of the fully controlled current source converter DC transmission line is greater than the load change rate threshold, the initial weights of each time-domain feature of the actual and predicted head-end voltages are increased to obtain the real-time weights of each time-domain feature of the actual and predicted head-end voltages. For example, when δu>5% or δp>10%, the weight w of the time-domain features (mean, variance) is increased. t Because the system is in a transient process, time-domain characteristics are more reflective of faults. When the signal-to-noise ratio (SNR) of a fully controlled current-source converter DC transmission line is less than the SNR threshold, the initial weights of each frequency domain characteristic of the actual and predicted head-end voltages are increased to obtain the real-time weights of each frequency domain characteristic of the actual and predicted head-end voltages. For example, when SNR < 20 dB, the weights of frequency domain characteristics (spectral energy, dominant frequency component) are increased. f To suppress noise interference; when the frequency offset on the grid-connected side of the DC transmission line of the fully controlled current source converter exceeds the frequency offset threshold, the initial weights of the wavelet time-frequency characteristics of the actual and predicted head-end voltages are increased to obtain the real-time weights of the wavelet time-frequency characteristics of the actual and predicted head-end voltages. For example, when Δf > 0.5Hz, the weight w of the wavelet time-frequency characteristics is increased. wtThis is to capture fault characteristics under frequency offset. For example, let the initial weight vector be W0=[w t ,w f ,w wt The weights are set to [0.4, 0.3, 0.3] and dynamically adjusted according to the above rules. For example, if δu > 5%, then W = [0.6, 0.2, 0.2]. Understandably, the weight adjustment range can be preset, for example, adjusted in a set step size each time, or adjusted as a percentage of the current weight.

[0075] In one specific embodiment of this application, the initial weights of each multimodal feature are determined through the following steps: calculating the fuzzy entropy value of each multimodal feature sequence, wherein a larger entropy value indicates higher feature discrimination and thus a larger weight; and obtaining the initial weight W of each multimodal feature sequence after normalization. entropy .

[0076] When adjusting the initial weights of each multimodal feature, the weight adjustment coefficients of each multimodal feature can be output through a pre-trained BP neural network model. The input to the BP neural network model is the system state parameters (δu, δp, Δf, SNR), and the output is the weight adjustment coefficients of each feature. A training sample set is pre-constructed based on historical fault data or fault simulation data of the DC transmission line. For example, system state parameters (δu, δp, Δf, SNR) are collected under different fault states of the DC transmission line, and the optimal weight combination is determined according to different fault states. The system state parameters under different fault states and the corresponding optimal weight combination are used as the training sample set, and the BP neural network is trained using the training sample set. When the system is running online, the BP neural network model can output a weight fine-tuning amount ΔW according to the real-time system state. The final weight can be expressed as: W = W entropy +ΔW.

[0077] S340. Based on the weighted multi-mode features, determine the correlation feature quantity between the actual head-end voltage and the predicted head-end voltage. Compare the current correlation feature quantity with the preset first correlation feature quantity threshold. If the current correlation feature quantity is greater than the first correlation feature quantity threshold, determine that the DC transmission line of the fully controlled current source converter is fault-free; otherwise, determine that the DC transmission line of the fully controlled current source converter is faulty.

[0078] Among them, the correlation features between the actual head-end voltage and the predicted head-end voltage are determined based on the weighted multimodal features, including:

[0079] S341. Determine the cosine similarity between the actual and predicted head-end voltages based on the weighted multimodal features. Specifically, calculate the cosine similarity between the actual and predicted head-end voltages using the following formula:

[0080]

[0081] in, Represents cosine similarity. This represents the real-time weight of the i-th multimodal feature. This represents the i-th multimode characteristic of the actual head-end voltage. This represents the i-th multimodal feature of the predicted head-end voltage.

[0082] S342. Obtain the historical multimodal characteristics of the actual head-end voltage of the fully controlled current source converter DC transmission line through a preset sliding window, construct the covariance matrix of the historical multimodal characteristics, and determine the Mahalanobis distance between the actual head-end voltage and the predicted head-end voltage of the fully controlled current source converter DC transmission line based on the covariance matrix of the actual head-end voltage, the predicted head-end voltage and the historical multimodal characteristics of the fully controlled current source converter DC transmission line.

[0083] Specifically, the Mahalanobis distance between the actual and predicted head-end voltages of a fully controlled current source converter DC transmission line is calculated using the following formula:

[0084]

[0085] in, Let X represent the Mahalanobis distance, Y represent the multimodal characteristics of the actual head-end voltage, Ct represent the multimodal characteristics of the predicted head-end voltage, and T represent the matrix transpose. This application uses a sliding window covariance matrix Ct instead of the global covariance matrix, thus better adapting to dynamic changes in the system.

[0086] Understandably, assuming the sliding window has m samples and each sample has a dimension of d (e.g., d=5, representing five features: mean, variance, peak value, spectral energy, and dominant frequency), we can construct an m×d data matrix D. Each row represents a feature vector at a given time step, and each column represents all samples of that feature at each time step. We calculate the arithmetic mean of each column to obtain its mean vector. Then, we subtract the corresponding mean vector from each row of the original data matrix D. The centered matrix is ​​obtained. The covariance is calculated using the formula: C t =(1 / m 1)D c T D c The covariance matrix Ct is calculated.

[0087] S343. Determine the adjustable fusion coefficients of cosine similarity and Mahalanobis distance. Based on the adjustable fusion coefficients of cosine similarity and Mahalanobis distance, perform weighted fusion of cosine similarity and Mahalanobis distance to obtain the correlation characteristic quantity between the actual head-end voltage and the predicted head-end voltage of the DC transmission line of the fully controlled current source converter.

[0088] Specifically, the correlation characteristic between the actual and predicted head-end voltages of a fully controlled current source converter DC transmission line is calculated using the following formula:

[0089]

[0090] Where R represents the relevance feature, α represents the adjustable cosine similarity fusion coefficient, and β represents the adjustable Mahalanobis distance fusion coefficient, which can be preset according to system reliability requirements. Represents cosine similarity. This represents the Mahalanobis distance.

[0091] After obtaining the correlation feature between the real-time acquired actual head-end voltage and the predicted head-end voltage output by the digital twin reference model, the current correlation feature is compared with the preset first correlation feature threshold. If the current correlation feature is greater than the first correlation feature threshold, it is determined that the DC transmission line of the fully controlled current source converter is fault-free. If the current correlation feature is less than or equal to the first correlation feature threshold, it indicates that the similarity between the real-time acquired head-end voltage feature and the head-end voltage feature under the fault-free normal operation state simulated by the digital twin reference model is low, and the DC transmission line of the fully controlled current source converter is faulty.

[0092] In this application, if a fault is determined in a fully controlled current source converter DC transmission line, the method further includes: if the current correlation feature quantity is greater than a preset second correlation feature quantity threshold, determining that the fully controlled current source converter DC transmission line is an external fault; otherwise, determining that the fully controlled current source converter DC transmission line is an internal fault. The second correlation feature quantity threshold is determined through the following steps:

[0093] S351. Simulate external faults in DC transmission lines with fully controlled current source converters under different fault scenarios. For example, simulate the combined scenario of external faults and system disturbances in PSCAD / EMTDC and record the maximum correlation feature quantity under each scenario.

[0094] S352. Determine multiple correlation features between the actual and predicted head-end voltages of a fully controlled current source converter DC transmission line under each out-of-zone fault scenario within a specified sampling period, and obtain the maximum correlation feature within the specified sampling period for each out-of-zone fault scenario. For example, under a load of 80% rated power, an out-of-zone fault with a transition resistance of 200Ω occurs at the end of the line, and there is a small power fluctuation in the system. Run this scenario in PSCAD, assuming the fault duration is 100 milliseconds. Within these 100 milliseconds, collect data such as the head-end voltage at a set sampling frequency, calculate the correlation features between each collected head-end voltage and the predicted head-end voltage output by the digital twin reference model, and record the maximum correlation feature. For N fault scenarios, a set containing N maximum correlation features can be obtained: {ρ max1 ,ρ max2 ,ρ max3 ,...,ρ maxN}

[0095] S353. The maximum correlation feature quantity under all obtained out-of-area fault scenarios is fitted using the kernel density estimation method to obtain the probability density of the maximum correlation feature quantity under each out-of-area fault scenario. It can be understood that the N data points representing the maximum similarity degree under out-of-area faults constitute a statistical distribution. The distribution of ρ is fitted using the kernel density estimation (KDE) method. The calculation process of the kernel density estimation method is existing technology and is not limited here.

[0096] S354. Determine the maximum correlation feature quantity with a preset probability density value as the maximum correlation feature quantity for faults outside the region. For example, take the 95th percentile as ρ. max That is, in all simulated out-of-area fault scenarios, 95% of the scenarios have a maximum correlation feature value that is lower than or equal to this ρ value.

[0097] S355. Determine the threshold for the second correlation feature based on the maximum correlation feature of out-of-area faults, including: calculating the standard deviation of the maximum correlation feature under all out-of-area fault scenarios; determining the safety margin and reliability coefficient of the standard deviation of the maximum correlation feature of out-of-area faults; and using the sum of the maximum correlation feature, safety margin, and the standard deviation weighted by the reliability coefficient as the threshold for the second correlation feature. Specifically, the threshold optimization formula can be expressed as: Threshold = ρ max +k σ + Δmargin, where Δmargin is the safety margin, set according to system reliability requirements, k is the reliability coefficient, usually taken as 2~3, and σ is the standard deviation of the fault correlation characteristic quantity outside the zone. Understandably, the threshold needs to be set according to the correlation characteristic quantity between the measured value of the first-end voltage and the reference model voltage when avoiding the maximum transition resistance.

[0098] In this way, this application first uses the first correlation feature threshold to quickly distinguish whether the system has a fault, thereby quickly filtering out most normal operating conditions and reducing the computational burden; when a fault is determined to exist, the fault location is identified by the second correlation feature threshold to ensure the accuracy of the fault judgment criteria within the area and prevent false operation and failure to operate.

[0099] Under normal system operation, the actual voltage characteristics and the predicted voltage characteristics of the digital twin reference model have a high degree of similarity. However, when the system is in an external fault state, the fault point is located outside the protected line. The traveling wave generated by the fault usually needs to propagate and be reflected by adjacent lines before reaching the protection installation point of this line, i.e., the head end. Therefore, although the voltage characteristics detected at the head end, such as the waveform, will be distorted, this distortion is indirect and attenuated. The actual waveform still maintains a high degree of similarity with the fault-free predicted waveform. Therefore, if the obtained correlation feature quantity is greater than the second correlation feature quantity threshold, it is determined to be an external fault. If the fault point is located inside the protected line, the fault will immediately and directly affect the voltage waveform at the head end. The voltage waveform will produce a fundamental and unique distortion, and the similarity with the waveform of the fault-free prediction model will drop sharply. Therefore, if the obtained correlation feature quantity is less than the second correlation feature quantity threshold, it is determined to be an internal fault.

[0100] Compared to traveling wave protection, the method of this application is not affected by line distributed parameters, has low requirements for sampling frequency, and does not require rapid capture of traveling wave front characteristics; compared to differential undervoltage protection, it has a higher tolerance to transition resistance and high reliability; compared to current differential protection, it is not affected by distributed capacitance current and has high speed; compared to undervoltage protection, it has a higher tolerance to transition resistance and high sensitivity.

[0101] A second aspect of this application provides a control method for a fully controlled current source converter DC transmission line, comprising:

[0102] If a fault is detected in the DC transmission line of the fully controlled current source converter using the above-mentioned fault detection method, the fully controlled current source converter is locked, and / or the DC transmission line of the fully controlled current source converter is disconnected.

[0103] Specifically, if a fault is detected in the DC transmission line of the fully controlled current source converter, the converter is locked, that is, the switching devices in the converter are turned off and no longer respond to the control commands of the system, so that the DC current on the current transmission line is zero; and / or the DC circuit breaker on the current transmission line is opened to disconnect the current transmission line, thereby isolating the fault in the current transmission line and protecting the converter and other equipment.

[0104] If a fault is detected in the DC transmission line of the fully controlled current source converter outside the zone, the converter is controlled to provide reactive power to the AC side to help stabilize the AC voltage.

[0105] like Figure 4 As shown, in a third aspect of this application, a fault detection device for a fully controlled current source converter DC transmission line is provided, comprising:

[0106] The data acquisition module is configured to acquire the actual starting voltage, actual starting current, and actual ending voltage of the DC transmission line of the fully controlled current source converter.

[0107] The first-end voltage prediction module is configured to take the actual first-end current and actual end voltage as inputs, and output the predicted first-end voltage of the fully controlled current source converter DC transmission line through a preset digital twin reference model.

[0108] The digital twin reference model is used to represent the time-domain mapping relationship between the start-up current, the end-up voltage, and the start-up voltage.

[0109] The fault detection module is configured to determine whether a DC transmission line of a fully controlled current source converter is faulty based on a correlation analysis of the actual head-end voltage and the predicted head-end voltage.

[0110] It is understood that those skilled in the art will clearly recognize that, for the sake of convenience and brevity, the above-described division of functional units and modules is used as an example. In practical applications, the above functions can be assigned to different functional units and modules as needed, that is, the internal structure of the device can be divided into different functional units or modules to complete all or part of the functions described above. The functional units and modules in the embodiments can be integrated into one processing unit, or each unit can exist physically separately, or two or more units can be integrated into one unit. The integrated unit can be implemented in hardware or as a software functional unit. Furthermore, the specific names of the functional units and modules are only for easy differentiation and are not intended to limit the scope of protection of this application. The specific working process of the units and modules in the above system can be referred to the corresponding process in the foregoing method embodiments, and will not be repeated here.

[0111] A fourth aspect of this application provides a control device for a fully controlled current source converter DC transmission line, comprising:

[0112] The control module is configured to, if a fault is determined in the DC transmission line of the fully controlled current source converter using the aforementioned fault detection method, control the fully controlled current source converter to lock out, and / or control the DC transmission line of the fully controlled current source converter to disconnect.

[0113] A fifth aspect of this application provides a fault detection system for DC transmission lines using a fully controlled current source converter, comprising:

[0114] A fully controlled current source converter; and a controller communicatively connected to the fully controlled current source converter;

[0115] The controller is configured to perform either the fault detection method for a fully controlled current source converter DC transmission line as described above, or the control method for a fully controlled current source converter DC transmission line as described above.

[0116] In a sixth aspect, this application provides a machine-readable storage medium storing instructions that, when executed by a processor, configure the processor to perform either the fully controlled current source converter DC transmission line fault detection method described above, or the fully controlled current source converter DC transmission line control method described above.

[0117] In a seventh aspect, this application provides a terminal device, including a memory, a processor, and a computer program stored in the memory and executable on the processor. When the processor executes the computer program, it implements the steps of the fault detection method for a fully controlled current source converter DC transmission line as described above, or executes the control method for a fully controlled current source converter DC transmission line as described above.

[0118] like Figure 5 The diagram shown is a schematic representation of a terminal device provided in an embodiment of this application. Figure 5 As shown, the terminal device 10 of this embodiment includes a processor 100, a memory 101, and a computer program 102 stored in the memory 101 and executable on the processor 100. When the processor 100 executes the computer program 102, it implements the steps in the above method embodiments. Alternatively, when the processor 100 executes the computer program 102, it implements the functions of each module / unit in the above device embodiments.

[0119] For example, computer program 102 may be divided into one or more modules / units, one or more of which are stored in memory 101 and executed by processor 100 to complete this application. One or more modules / units may be a series of computer program instruction segments capable of performing a specific function, which describe the execution process of computer program 102 in terminal device 10.

[0120] Terminal device 10 may be a computing device such as a desktop computer, laptop, handheld computer, or cloud server. Terminal device 10 may include, but is not limited to, a processor 100 and a memory 101. Those skilled in the art will understand that... Figure 5 This is merely an example of terminal device 10 and does not constitute a limitation on terminal device 10. It may include more or fewer components than shown, or combine certain components, or different components. For example, terminal device may also include input / output devices, network access devices, buses, etc.

[0121] Processor 100 can be a Central Processing Unit (CPU), or other general-purpose processors, digital signal processors (DSPs), application-specific integrated circuits (ASICs), field-programmable gate arrays (FPGAs), or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, etc. The general-purpose processor can be a microprocessor or any conventional processor.

[0122] The memory 101 can be an internal storage unit of the terminal device 10, such as a hard disk or RAM of the terminal device 10. The memory 101 can also be an external storage device of the terminal device 10, such as a plug-in hard disk, Smart Media Card (SMC), Secure Digital (SD) card, or FlashCard equipped on the terminal device 10. Furthermore, the memory 101 can include both internal and external storage units of the terminal device 10. The memory 101 is used to store computer programs and other programs and data required by the terminal device 10. The memory 101 can also be used to temporarily store data that has been output or will be output.

[0123] Those skilled in the art will understand that embodiments of this application can be provided as methods, systems, or computer program products. Therefore, this application can take the form of a completely hardware embodiment, a completely software embodiment, or an embodiment combining software and hardware aspects. Furthermore, this application can take the form of a computer program product embodied on one or more computer-usable storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code.

[0124] It should also be noted that the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such process, method, article, or apparatus. Unless otherwise specified, an element defined by the phrase "comprising one..." does not exclude the presence of other identical elements in the process, method, article, or apparatus that includes that element.

[0125] The above are merely embodiments of this application and are not intended to limit the scope of this application. Various modifications and variations can be made to this application by those skilled in the art. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of this application should be included within the scope of the claims of this application.

Claims

1. A method for fault detection in DC transmission lines using a fully controlled current source converter, characterized in that, include: Obtain the actual starting voltage, actual starting current, and actual ending voltage of a DC transmission line with a fully controlled current source converter. Using the actual head current and actual tail voltage as inputs, the predicted head voltage of the fully controlled current source converter DC transmission line is output through a preset digital twin reference model. The correlation analysis between the actual and predicted head-end voltages is used to determine whether the DC transmission line of the fully controlled current source converter is faulty. The digital twin reference model is used to represent the time-domain mapping relationship between the start-up current, the end-up voltage, and the start-up voltage. The digital twin reference model is constructed through the following steps: Based on the frequency-dependent model, a two-terminal frequency domain relationship is established to represent the linear mapping relationship between the starting voltage and starting current and the ending voltage and ending current of the DC transmission line of the fully controlled current source converter. The two-terminal frequency domain relationship is then converted into a single-input output relationship with the starting current and ending voltage as inputs and the starting voltage as the output. Based on the first-end current, the last-end voltage, and the first-end voltage under fault-free operation, the single-input-output relationship is fitted into a digital twin reference model in the time domain. The single-input-output relationship includes: in, , Where tanh represents the ratio of the hyperbolic sine function to the hyperbolic cosine function, and sech represents the reciprocal of the hyperbolic cosine function. Let be the propagation constant. R(ω) is the characteristic impedance, in Ω; L(ω) is the resistance per unit length, in rad / m; G(ω) is the inductance per unit length; C(ω) is the conductance per unit length; and l is the length of the transmission line, in meters. This is the voltage at the beginning of the transmission line. This refers to the voltage at the end of the transmission line. Let be the current at the beginning of the transmission line, j be an imaginary number, ω be the angular frequency, cosh be the hyperbolic cosine function, and sinh be the hyperbolic sine function.

2. The fault detection method for DC transmission lines using a fully controlled current source converter according to claim 1, characterized in that, Based on the frequency-dependent model, a two-terminal frequency domain relationship is established to represent the linear mapping relationship between the starting-end voltage and starting-end current and the ending voltage and ending current of the fully controlled current source converter DC transmission line, including: The frequency-dependent characteristics of resistance, inductance, conductance and capacitance per unit length in the DC transmission line of the fully controlled current source converter are determined based on the frequency-dependent model. Based on the frequency-varying characteristics of resistance, inductance, conductance and capacitance per unit length, the relationship between the traveling wave, reverse traveling wave, terminal voltage and terminal current at both ends of the DC transmission line of the fully controlled current source converter is determined in the frequency domain. Based on the relationship between the traveling wave, reverse traveling wave, terminal voltage, and terminal current at both ends of the fully controlled current source converter DC transmission line in the frequency domain, a first mapping relationship between the first terminal voltage and the terminal voltage and terminal current of the fully controlled current source converter DC transmission line is established, as well as a second mapping relationship between the first terminal current and the terminal voltage and terminal current of the fully controlled current source converter DC transmission line. Based on the first mapping relationship and the second mapping relationship, a two-terminal frequency domain relationship is established to represent the linear mapping relationship between the starting voltage and starting current of the fully controlled current source converter DC transmission line and the ending voltage and ending current.

3. The fault detection method for DC transmission lines using a fully controlled current source converter according to claim 1, characterized in that, The dual-end frequency domain relationship includes: in, Let be the propagation constant. R(ω) is the characteristic impedance, in Ω; R(ω) is the resistance per unit length, in rad / m; L(ω) is the inductance per unit length; G(ω) is the conductance per unit length; C(ω) is the capacitance per unit length; and l is the length of the transmission line, in meters. This is the voltage at the beginning of the transmission line. This refers to the voltage at the end of the transmission line. This refers to the current at the beginning of the transmission line. Let ω represent the end current of the transmission line, j be an imaginary number, ω represent the angular frequency, cosh represent the hyperbolic cosine function, and sinh represent the hyperbolic sine function.

4. The fault detection method for DC transmission lines using a fully controlled current source converter according to claim 2, characterized in that, Converting the aforementioned two-terminal frequency domain relationship into a single-input output relationship with the first-terminal current and the last-terminal voltage as inputs and the first-terminal voltage as the output includes: Based on the second mapping relationship, with the terminal current of the fully controlled current source converter DC transmission line as the intermediate quantity, the first mapping relationship is converted into a third mapping relationship between the starting voltage and the terminal voltage and starting current of the fully controlled current source converter DC transmission line. Based on the third mapping relationship, the starting current and ending voltage of the fully controlled current source converter DC transmission line are mapped to the current frequency domain response and voltage frequency domain response of the starting voltage, respectively, to obtain a single-input output relationship with the starting voltage as the output.

5. The fault detection method for DC transmission lines using a fully controlled current source converter according to claim 1, characterized in that, Based on the start-up current, end-up voltage, and start-up voltage under fault-free operation, the single-input-output relationship is fitted into a digital twin reference model in the time domain, including: Continuously acquire the starting voltage, starting current and ending voltage of the DC transmission line of the fully controlled current source converter under fault-free operation; The current frequency domain response and voltage frequency domain response of the first-terminal voltage are fitted by a parameter identification algorithm so that the current frequency domain response and voltage frequency domain response of the first-terminal voltage are expressed as a complex exponential sum, and the fitting result is obtained. The fitting result is inversely transformed to obtain the current time-domain impulse response and voltage time-domain impulse response of the first-terminal voltage, thus obtaining the digital twin reference model of the single-input-output relationship in the time domain.

6. The fault detection method for DC transmission lines using a fully controlled current source converter according to claim 1, characterized in that, Digital twin reference models include: in, This represents the predicted head-end voltage output by the digital twin reference model. This represents the actual current at the beginning of the circuit. Let represent the actual terminal voltage, t represent time t, and e be the natural constant. , , , , , , , and All are model coefficients, among which, , Let represent the amplitude coefficients of the m / n-th oscillation component, respectively. , Let represent the attenuation coefficients of the m / n-th oscillation component, respectively. , These represent the angular frequencies of the m / n-th oscillation component, respectively. , Let M and N represent the initial phases of the m / nth oscillation components, respectively, and M and N represent the fitting order, respectively.

7. The fault detection method for DC transmission lines using a fully controlled current source converter according to claim 1, characterized in that, Determining whether the DC transmission line of the fully controlled current source converter is faulty based on correlation analysis between the actual and predicted head-end voltages includes: Extract the multimodal features of the actual head-end voltage and the multimodal features of the predicted head-end voltage, and determine the initial weight of each multimodal feature. The multimodal features include at least time-domain features, frequency-domain features and wavelet time-frequency features. The current system operating status of the fully controlled current source converter DC transmission line is determined based on the actual head-end voltage and actual head-end current of the fully controlled current source converter DC transmission line. Based on the current system operating status of the DC transmission line with the fully controlled current source converter, the initial weights of each multi-mode feature of the actual head-end voltage and the predicted head-end voltage are adjusted to obtain the real-time weights of each multi-mode feature, and the corresponding multi-mode features are weighted using the real-time weights of each multi-mode feature. Based on the weighted multimodal features, the correlation feature quantity between the actual head-end voltage and the predicted head-end voltage is determined. The current correlation feature quantity is compared with the preset first correlation feature quantity threshold. If the current correlation feature quantity is greater than the first correlation feature quantity threshold, it is determined that the DC transmission line of the fully controlled current source converter is fault-free; otherwise, it is determined that the DC transmission line of the fully controlled current source converter is faulty.

8. The fault detection method for DC transmission lines using a fully controlled current source converter according to claim 7, characterized in that, The current system operating status of the fully controlled current source converter DC transmission line is determined based on the actual head-end voltage and actual head-end current of the fully controlled current source converter DC transmission line, including: Obtain the rated voltage, rated power, and rated operating frequency on the grid-connected side of the DC transmission line of the fully controlled current source converter, and determine the actual operating frequency on the grid-connected side of the DC transmission line of the fully controlled current source converter. The actual voltage fluctuation of the fully controlled current source converter DC transmission line is determined based on the difference between the actual head-end voltage and the rated voltage. The actual power of the fully controlled current source converter DC transmission line is determined based on the actual head-end voltage and the actual head-end current. The actual power fluctuation of the fully controlled current source converter DC transmission line is determined based on the difference between the actual power and the rated power. The voltage fluctuation rate of the fully controlled current source converter DC transmission line is determined based on the ratio of the actual head-end voltage fluctuation to the rated voltage; the load change rate of the fully controlled current source converter DC transmission line is determined based on the ratio of the actual power fluctuation to the rated power; the frequency offset of the grid-connected side of the fully controlled current source converter DC transmission line is determined based on the difference between the actual operating frequency and the rated operating frequency; and the signal-to-noise ratio of the fully controlled current source converter DC transmission line is determined based on the actual head-end voltage or the actual head-end current. The current system operating status of the DC transmission line with the fully controlled current source converter is determined based on the voltage fluctuation rate, load change rate, frequency offset, and signal-to-noise ratio.

9. The fault detection method for DC transmission lines using a fully controlled current source converter according to claim 8, characterized in that, Based on the current system operating state of the fully controlled current source converter DC transmission line, the initial weights of each multimode feature of the actual head-end voltage and the predicted head-end voltage are adjusted to obtain the real-time weights of each multimode feature, including: When the voltage fluctuation rate of the fully controlled current source converter DC transmission line is greater than the voltage fluctuation rate threshold, or the load change rate of the fully controlled current source converter DC transmission line is greater than the load change rate threshold, the initial weights of each time-domain feature of the actual head-end voltage and the predicted head-end voltage are increased to obtain the real-time weights of each time-domain feature of the actual head-end voltage and the predicted head-end voltage. When the signal-to-noise ratio of the DC transmission line of the fully controlled current source converter is less than the signal-to-noise ratio threshold, the initial weights of each frequency domain feature of the actual head-end voltage and the predicted head-end voltage are increased to obtain the real-time weights of each frequency domain feature of the actual head-end voltage and the predicted head-end voltage. When the frequency offset on the grid-connected side of the DC transmission line of the fully controlled current source converter is greater than the frequency offset threshold, the initial weights of the wavelet time-frequency characteristics of the actual head-end voltage and the predicted head-end voltage are increased to obtain the real-time weights of the wavelet time-frequency characteristics of the actual head-end voltage and the predicted head-end voltage.

10. The fault detection method for DC transmission lines using a fully controlled current source converter according to claim 7, characterized in that, Based on the weighted multimodal features, the correlation features between the actual and predicted head-end voltages are determined, including: The cosine similarity between the actual head-end voltage and the predicted head-end voltage is determined based on the weighted multimodal features. The historical multimodal characteristics of the actual head-end voltage of the fully controlled current source converter DC transmission line are obtained by a preset sliding window, and the covariance matrix of the historical multimodal characteristics is constructed. Based on the actual head-end voltage, the predicted head-end voltage and the covariance matrix of the historical multimodal characteristics of the fully controlled current source converter DC transmission line, the Mahalanobis distance between the actual head-end voltage and the predicted head-end voltage of the fully controlled current source converter DC transmission line is determined. Determine the adjustable fusion coefficients of cosine similarity and Mahalanobis distance, and perform weighted fusion of the cosine similarity and Mahalanobis distance based on the adjustable fusion coefficients of cosine similarity and Mahalanobis distance to obtain the correlation characteristic quantity between the actual head-end voltage and the predicted head-end voltage of the DC transmission line of the fully controlled current source converter.

11. The fault detection method for DC transmission lines using a fully controlled current source converter according to claim 10, characterized in that, The cosine similarity between the actual and predicted head-end voltages is determined based on the weighted multimodal features, including: The cosine similarity between the actual and predicted head-end voltages is calculated using the following formula: in, Represents cosine similarity. This represents the real-time weight of the i-th multimodal feature. This represents the i-th multimode characteristic of the actual head-end voltage. This represents the i-th multimodal feature of the predicted head-end voltage.

12. The fault detection method for DC transmission lines using a fully controlled current source converter according to claim 10, characterized in that, The Mahalanobis distance between the actual and predicted head-end voltages of the fully controlled current source converter DC transmission line is calculated using the following formula: in, Let X represent the Mahalanobis distance, Y represent the multimode characteristics of the actual head-end voltage, and C represent the multimode characteristics of the predicted head-end voltage. t Let T represent the covariance matrix of the historical multimodal features, and let T denote the matrix transpose.

13. The fault detection method for DC transmission lines using a fully controlled current source converter according to claim 10, characterized in that, The correlation characteristic between the actual and predicted head-end voltages of the fully controlled current source converter DC transmission line is calculated using the following formula: Where R represents the relevance feature quantity, α represents the adjustable cosine similarity fusion coefficient, and β represents the adjustable Mahalanobis distance fusion coefficient. Represents cosine similarity. This represents the Mahalanobis distance.

14. The fault detection method for DC transmission lines using a fully controlled current source converter according to claim 7, characterized in that, If a fault is determined in the DC transmission line of the fully controlled current source converter, the method further includes: If the current correlation feature quantity is greater than the preset second correlation feature quantity threshold, the DC transmission line of the fully controlled current source converter is determined to be an external fault; otherwise, the DC transmission line of the fully controlled current source converter is determined to be an internal fault. The second correlation feature threshold is determined through the following steps: Simulate the external faults of the fully controlled current source converter DC transmission line under different fault scenarios; Within a specified sampling period, determine multiple correlation features between the actual head-end voltage and the predicted head-end voltage of the fully controlled current source converter DC transmission line under each out-of-area fault scenario, and obtain the maximum correlation feature within the specified sampling period under each out-of-area fault scenario; The probability density of the maximum correlation feature under each out-of-area fault scenario is obtained by fitting the kernel density estimation method to the maximum correlation feature under all out-of-area fault scenarios. The maximum correlation feature with a probability density of a preset probability density value is determined to be the maximum correlation feature of faults outside the region. The second correlation feature threshold is determined based on the maximum correlation feature of the fault outside the region.

15. The fault detection method for DC transmission lines using a fully controlled current source converter according to claim 14, characterized in that, Determining the threshold of the second correlation feature based on the maximum correlation feature of the fault outside the region includes: Calculate the standard deviation of the maximum correlation feature across all out-of-area fault scenarios; Determine the safety margin of the maximum correlation characteristic of the out-of-area fault and the reliability coefficient of the standard deviation; The sum of the maximum correlation feature of the fault outside the zone, the safety margin, and the standard deviation after weighting by the reliability coefficient is used as the threshold of the second correlation feature.

16. A control method for a DC transmission line using a fully controlled current source converter, characterized in that, include: If a fault in the DC transmission line of the fully controlled current source converter is determined by the fault detection method of the fully controlled current source converter according to any one of claims 1-15, the fully controlled current source converter is locked, and / or the DC transmission line of the fully controlled current source converter is disconnected.

17. A fault detection device for DC transmission lines using a fully controlled current source converter, characterized in that, The fault detection method for DC transmission lines using a fully controlled current source converter as described in any one of claims 1-15, wherein the apparatus comprises: The data acquisition module is configured to acquire the actual starting voltage, actual starting current, and actual ending voltage of the DC transmission line of the fully controlled current source converter. The first-end voltage prediction module is configured to take the actual first-end current and actual end voltage as inputs, and output the predicted first-end voltage of the fully controlled current source converter DC transmission line through a preset digital twin reference model. The digital twin reference model is used to represent the time-domain mapping relationship between the start-up current, the end-up voltage, and the start-up voltage. The fault detection module is configured to determine whether the DC transmission line of the fully controlled current source converter is faulty based on the correlation analysis between the actual head-end voltage and the predicted head-end voltage.

18. A control device for a fully controlled current source type converter DC transmission line, characterized in that, include: The control module is configured to, if a fault is determined in the DC transmission line of the fully controlled current source converter according to the fault detection method of the fully controlled current source converter according to any one of claims 1-15, control the fully controlled current source converter to lock up, and / or control the fully controlled current source converter DC transmission line to disconnect.

19. A converter system, characterized in that, include: Fully controlled current source converter; as well as The controller is communicatively connected to the fully controlled current source converter; The controller is configured to perform the fault detection method for a fully controlled current source converter DC transmission line as described in any one of claims 1-15, or to perform the control method for a fully controlled current source converter DC transmission line as described in claim 16.

20. A machine-readable storage medium storing instructions thereon, characterized in that, When executed by a processor, this instruction causes the processor to be configured to perform the fault detection method for a fully controlled current source converter DC transmission line as described in any one of claims 1-15, or to perform the control method for a fully controlled current source converter DC transmission line as described in claim 16.

21. A terminal device, comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, characterized in that, When the processor executes the computer program, it implements the steps of the fault detection method for a fully controlled current source converter DC transmission line as described in any one of claims 1-15, or executes the control method for a fully controlled current source converter DC transmission line as described in claim 16.

Images