An intelligent control system for power failure data acquisition and processing

Abstract

The application discloses an intelligent control system for power failure data acquisition and processing, and belongs to the technical field of power grid data processing. Through the cooperation among a CVT transient discharge tailing component reconstruction module, a real recovery voltage extraction and frequency change rate calculation module and an electromechanical-electromagnetic coupling oscillation identification and control module, the decay subsynchronous oscillation component generated by the discharge coil can be processed. The conventional Fourier algorithm regards the decay subsynchronous oscillation component as a real fundamental frequency offset, which leads to the serious distortion of the frequency change rate calculation, and further causes the misoperation of the low-frequency load shedding device to cut off the normal load. The application implements effective processing and solves the defect that the CVT transient "discharge tailing" induces false low-frequency oscillation in the prior art.

Description

Technical Field

[0001] This invention relates to the field of power grid data processing technology, and specifically to an intelligent control system for power fault data acquisition and processing. Background Technology

[0002] When a high-voltage line fault occurs, the sharp drop in fault voltage will excite the capacitor bank (main capacitor C1, voltage divider capacitor C2) inside the CVT (capacitive voltage transformer), generating a decaying subsynchronous oscillation component through the discharge coil, usually 5~15Hz. This physical tail component will be directly superimposed on the real power frequency recovery voltage. Conventional Fourier transform (DFT) algorithms treat this as the real fundamental frequency offset, resulting in serious distortion in the calculation of the frequency change rate, which in turn causes "false shedding of normal load" in low-frequency load shedding devices. This is particularly prominent on the grid-connected side of doubly-fed wind turbines, because the turbine rotor itself also contains subsynchronous modes. The superposition of the two makes it difficult for dispatchers to distinguish between system oscillation and CVT deception, resulting in the defect of CVT transient "discharge tail" inducing false low-frequency oscillations. Summary of the Invention

[0003] The purpose of this invention is to provide an intelligent control system for power fault data acquisition and processing, which solves the technical problems mentioned in the background art.

[0004] The objective of this invention can be achieved through the following technical solutions: An intelligent control system for power fault data acquisition and processing includes: CVT transient discharge tail component reconstruction module: Collects the instantaneous values ​​of three-phase voltage after a high-voltage line fault, constructs a fractional-order state-space observer based on the topology of the CVT internal main capacitor and voltage divider capacitor, reconstructs the transient discharge charge trajectory of the CVT internal capacitor bank by solving fractional-order differential equations, and separates the CVT discharge tail component. The module for extracting the true recovery voltage and calculating the rate of change of frequency is as follows: Based on the reconstructed CVT discharge tail component, a nonlinear phase cancellation operator is constructed. The CVT discharge tail component is used as an adaptive cancellation factor and is used to perform complex plane vector cancellation with the original acquired instantaneous values ​​of the three-phase voltage to extract the true recovery voltage on the grid side. Based on this true recovery voltage, the true rate of change of frequency is calculated by instantaneous phase differentiation. Electromechanical-electromagnetic coupling oscillation identification and control module: Based on the calculated real frequency change rate and combined with the dynamic response characteristics of the phase-locked loop of the doubly fed wind turbine rotor-side converter, an electromechanical-electromagnetic coupling oscillation identification matrix is ​​constructed. By calculating the eigenvalue trajectory of the identification matrix, the attributes of the low-frequency oscillation source are determined, and a blocking or enabling control command is output to the low-frequency load shedding device according to the identification result.

[0005] Furthermore, under normal operating conditions of the high-voltage line, a multi-frequency sweep excitation signal is injected into the secondary side of the CVT to collect the input voltage and output voltage response data of the CVT; based on the collected input and output response data, the physical topology parameters of the CVT are identified; at the same time, by analyzing the relationship curve between the hysteresis loop area and frequency of the discharge coil core material, the fractional order characterizing the dielectric memory effect of the capacitor is determined.

[0006] Furthermore, based on the identified physical topology parameters and fractional order of the CVT, a fractional-order state-space equation describing the transient discharge process of the capacitor bank inside the CVT is established; the charge at both ends of the main capacitor C1, the charge at both ends of the voltage divider capacitor C2, and the magnetic flux of the discharge coil are selected as state variables to construct a third-order fractional-order state-space observer model.

[0007] Furthermore, starting from the fault triggering moment, the constructed fractional-order state-space observer model is discretized and solved; at each sampling moment, the state vector at the current moment is calculated based on the instantaneous value of the three-phase voltage collected at the current moment and the state vector at the previous moment.

[0008] Furthermore, after calculating the state vector at the current moment, the CVT discharge tail component at the current moment is calculated through the output equation.

[0009] Furthermore, when implementing the adaptive calculation of the nonlinear phase cancellation gain factor, the energy ratio of the tail component in the original sampled signal is calculated based on the calculated complex plane analytic signal and the complex form tail component, and the nonlinear phase cancellation gain factor is dynamically adjusted accordingly.

[0010] Furthermore, when performing complex plane vector cancellation and orthogonal decomposition of the true recovery voltage, the obtained complex plane analytic signal and complex form tail component, as well as the calculated nonlinear phase cancellation gain factor, are used to perform complex plane vector cancellation operation, extract the true recovery voltage on the grid side, and decompose the true recovery voltage on the grid side into orthogonal components in the stationary coordinate system.

[0011] Furthermore, when constructing the electromechanical-electromagnetic coupled oscillation identification matrix, based on the output phase angle and output angular frequency obtained through processing, and combined with the actual frequency change rate and its change rate output by the preceding module, a fourth-order state-space identification matrix describing the characteristics of electromechanical-electromagnetic coupled oscillation is constructed.

[0012] Furthermore, when performing eigenvalue trajectory calculation of the identification matrix and oscillation source attribute discrimination, the constructed electromechanical-electromagnetic coupling oscillation identification matrix is ​​decomposed into eigenvalues, and its four eigenvalues ​​are calculated. Based on the real and imaginary parts of the eigenvalues, the attributes of the low-frequency oscillation source are determined.

[0013] Furthermore, based on the oscillation source attribute determination results, corresponding control commands are generated and output to the low-frequency load shedding device.

[0014] Compared to existing solutions, the beneficial effects achieved by this invention are: This invention, through the coordinated operation of a CVT transient discharge tail component reconstruction module, a true recovery voltage extraction and frequency change rate calculation module, and an electromechanical-electromagnetic coupling oscillation identification and control module, can effectively address the issue of attenuated subsynchronous oscillation components generated by the discharge coil being mistakenly treated as true fundamental frequency shifts by conventional Fourier algorithms, leading to severe distortion in frequency change rate calculations and consequently causing erroneous shearing of normal loads in low-frequency load shearing devices. This solves the defect of existing technologies where CVT transient "discharge tail" induces false low-frequency oscillations. Attached Figure Description

[0015] The invention will now be further described with reference to the accompanying drawings.

[0016] Figure 1 This is a flowchart illustrating the operation of an intelligent control system for power fault data acquisition and processing according to the present invention. Detailed Implementation

[0017] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0018] like Figure 1 As shown, the present invention is an intelligent control system for power fault data acquisition and processing, comprising: CVT Transient Discharge Tail Component Reconstruction Module: This module acquires the instantaneous three-phase voltage values ​​after a high-voltage line fault, constructs a fractional-order state-space observer based on the CVT's internal main capacitor and voltage divider capacitor topology, reconstructs the transient discharge charge trajectory of the CVT's internal capacitor bank by solving fractional-order differential equations, and separates the CVT discharge tail component. Specifically: Under normal operating conditions of the high-voltage line, a multi-frequency sweep excitation signal is injected into the secondary side of the CVT (capacitive voltage transformer), and the input and output voltage response data of the CVT are collected. Based on the collected input and output response data, a particle swarm optimization algorithm is used to identify the physical topology parameters of the CVT, including the main capacitor C1, the voltage divider capacitor C2, the equivalent inductance L of the discharge coil, and the equivalent resistance R. Simultaneously, by analyzing the relationship curve between the hysteresis loop area and frequency of the discharge coil core material, the fractional order characterizing the dielectric memory effect of the capacitor is determined. The specific steps include: Hysteresis loop tests were conducted on the core material of the CVT internal discharge coil under laboratory conditions at different frequency points. Next, the magnetic induction intensity of the iron core material is collected using a hysteresis loop measuring instrument. and magnetic field strength data, To obtain the hysteresis loop at each frequency point at the current sampling time, the relevant expression is: ; in: For the first One test frequency point; The starting frequency for the test frequency range is set to 0.1Hz; The frequency step interval is set to 0.5Hz; The frequency point number is a positive integer. ; is the total number of test frequency points, and is a dimensionless parameter.

[0019] It should be noted that by systematically collecting hysteresis loop data of the discharge coil core material over a wide frequency range, raw experimental data is provided to support the subsequent analysis of the relationship between hysteresis loss and frequency. In addition, in the embodiments of the present invention, all expressions involving calculations are standardized before calculation. Standardization includes, but is not limited to, extracting numerical values ​​from the data and normalization, etc., which are all existing conventional technical solutions. The specific implementation steps will not be elaborated here.

[0020] Based on the collected magnetic induction intensity and magnetic field strength The data is used to calculate the frequency at each point through numerical integration. The area enclosed by the lower hysteresis loop The expression involved is: ;in: For frequency The area of ​​the hysteresis loop at that frequency characterizes the energy density of the hysteresis loss per unit volume of the core material at that frequency. Indicates frequency A corresponding complete magnetization cycle Within, perform closed-loop integration; For frequency The corresponding magnetization period; The index of the sampling point; The number of sampling points within one magnetization cycle; For the first The magnetic field strength value at each sampling point; For the first The magnetic flux density value at each sampling point; This represents the difference in magnetic induction intensity between adjacent sampling points.

[0021] It should be noted that by quantifying the hysteresis loss of the core material at each frequency point, the area of ​​the hysteresis loop directly reflects the energy dissipation characteristics of the material in an alternating magnetic field. The relationship between this area and frequency contains key information about the material's memory effect.

[0022] When fitting the hysteresis loop area-frequency relationship curve, the calculated hysteresis loop area at each frequency point will be... With corresponding frequency Curve fitting is performed using a fractional power-law model, and the fitting coefficients are obtained. The corresponding expression is: ;in: For the calculated frequency The area of ​​the hysteresis loop below; is the hysteresis loss coefficient, representing the baseline value of the hysteresis loop area per unit frequency; is the frequency exponent of hysteresis loss, which characterizes the nonlinearity of the hysteresis loop area as a function of frequency, and is a dimensionless parameter. Taking the logarithm of both sides of the above formula transforms it into a linear regression problem: ; The fitting parameters were calculated using the least squares method. and : ; ; ; in: It is the natural logarithm function; It is a natural exponential function.

[0023] It should be noted that by using a fractional power-law model, the nonlinear relationship between the hysteresis loop area and frequency is accurately fitted, and the hysteresis loss frequency index is extracted. This index reflects the frequency dependence of the hysteresis loss of the core material, providing intermediate parameters for the subsequent derivation of the fractional order of the capacitor dielectric memory effect.

[0024] Based on the magnetic-electric analogy principle of ferromagnetic materials, when establishing the mapping relationship between the hysteresis loss characteristics of the discharge coil core and the dielectric memory effect of the CVT capacitor, according to electromagnetic field theory, the hysteresis memory effect of the core material and the dielectric memory effect of the capacitor are mathematically described by a unified fractional calculus. Their frequency dependence characteristics satisfy a conjugate complementary relationship, and the relevant expressions are as follows: ';in: The fractional order characterizing the dielectric memory effect of CVT capacitors is a dimensionless parameter with a value range of [value range missing]. ; It originates from the integer-order upper bound of the magneto-electric duality relation in fractional calculus theory.

[0025] When constructing the fractional-order state-space observer model, based on the identified CVT physical topology parameters and fractional-order α, a fractional-order state-space equation describing the transient discharge process of the capacitor bank inside the CVT is established; the charge quantity across the main capacitor C1 is selected. The charge across the voltage divider capacitor C2 and discharge coil magnetic flux As the state variable, a third-order fractional-order state-space observer model is constructed, with the expression: ; ;in: Indicates order is The Caputo fractional differential operator is used to characterize the non-integer order memory decay characteristics between the discharge coil and the capacitor bank; For state vectors, ; The input vector is the instantaneous value of the three-phase voltage after a high-voltage line fault. To reconstruct the pure CVT discharge tail component of the output; The system matrix is ​​a 3×3 matrix constructed based on the physical topology parameters of the CVT, and its elements are determined by the functional relationship between C1, C2, L, and R. The input matrix is ​​3×1, representing the weights of the influence of external voltage excitation on the system state; The output matrix is ​​1×3, representing the contribution coefficient of each state variable to the discharge tail component.

[0026] The purpose of this step is to establish a mathematical model that can accurately describe the physical process of transient discharge of the capacitor bank inside the CVT. By using fractional differential equations, the non-integer order memory decay characteristics between the discharge coil and the capacitor bank are captured, providing a theoretical basis for subsequent real-time reconstruction of the discharge tail component.

[0027] During fault triggering and observer initialization, the three-phase voltage amplitude of the high-voltage line is monitored in real time. When the voltage amplitude of any phase drops by more than 30% of the rated voltage within a single power frequency cycle, a fault is determined to have occurred in the high-voltage line, and the CVT transient discharge tail component reconstruction process is immediately triggered. At the fault triggering moment... The instantaneous values ​​of the three-phase voltage on the secondary side of the CVT are collected as the initial input of the observer. Based on the capacitance charge and discharge coil flux linkage during steady-state operation before the fault, the initial state vector of the fractional-order state-space observer is set. ; The purpose of this step is to accurately identify the time of the fault occurrence and promptly initiate the reconstruction process. At the same time, by setting a reasonable initial state, it avoids numerical oscillations in the observer during the initial stage of startup, thus ensuring the numerical stability and convergence speed of the reconstruction process.

[0028] When performing numerical solutions and real-time reconstruction of tailing components for fractional differential equations, from the fault triggering time... Initially, an improved Grünwald-Letnikov numerical method was used to discretize and solve the constructed fractional-order state-space observer model; at each sampling time... ; The current discrete time step number is a positive integer. The sampling interval is based on the instantaneous values ​​of the three-phase voltages collected at the current moment. and the state vector of the previous time step The current state vector is calculated using the following discretized iterative formula: ;in: The current sampling time The state vector; It is a 3×3 identity matrix; The sampling interval; This serves as a step number for retracing historical memory. The number of memory steps represents the memory depth of the fractional derivative, and its value is a positive integer. The coefficients of the generalized binomial are calculated using the following formula: ; For a historic moment The state vector; Calculate the state vector at the current time. Then, output the equation. Calculate the CVT discharge tail component at the current moment.

[0029] It should be noted that, through an efficient numerical solution algorithm, the transient discharge charge trajectory of the CVT internal capacitor bank is calculated in real time, and the pure discharge tail component is separated from it, so as to achieve accurate reconstruction of the 5~15Hz subsynchronous oscillation component, and provide an accurate compensation signal for the subsequent extraction of the real recovery voltage.

[0030] The reconstructed output CVT discharge tail component The spectrum of the reconstructed tail component is compared with the original three-phase voltage instantaneous value to calculate the energy ratio of the reconstructed tail component in the 5~15Hz frequency band. If the energy ratio is lower than the preset threshold (set to 85%), the reconstruction accuracy is deemed insufficient, and an adaptive correction mechanism is activated: the equivalent resistance R parameter value is dynamically adjusted using the gradient descent method, and the fractional-order state space observer model is updated. The tail component reconstruction is then re-executed until the energy ratio of the reconstructed tail component in the 5~15Hz frequency band reaches or exceeds the preset threshold.

[0031] It should be noted that by verifying the accuracy of the reconstruction results in real time and performing dynamic correction when necessary, the output CVT discharge tail component can accurately reflect the subsynchronous oscillation characteristics in the actual physical process, thereby improving the robustness and adaptability of the entire reconstruction module.

[0032] In this embodiment of the invention, by constructing a fractional-order state-space observer based on the CVT physical topology, the non-integer-order memory decay characteristics between the discharge coil and the capacitor bank can be accurately captured, achieving high-precision reconstruction of the 5~15Hz subsynchronous oscillation component, which can effectively improve the reconstruction accuracy. The improved Grünwald-Letnikov numerical method is used for discretization, which has low computational complexity and can meet the millisecond-level real-time requirements of the power system. Through the reconstruction result verification and dynamic correction mechanism, the observer model parameters can be automatically adjusted when the CVT parameters drift due to factors such as aging and temperature changes, ensuring the accuracy and reliability of long-term operation. The pure CVT discharge tail component output by reconstruction provides an accurate compensation signal for the subsequent real recovery voltage extraction and frequency change rate calculation module, fundamentally solving the technical problem that the conventional Fourier algorithm misjudges the tail component as the fundamental frequency offset, and effectively preventing the low-frequency load shedding device from erroneously cutting off normal loads.

[0033] The module for extracting the true recovery voltage and calculating the rate of frequency change is as follows: Based on the reconstructed CVT discharge tail component, a nonlinear phase cancellation operator is constructed. The CVT discharge tail component is used as an adaptive cancellation factor and subjected to complex plane vector cancellation with the originally acquired instantaneous three-phase voltage values ​​to extract the true recovery voltage on the grid side. Based on this true recovery voltage, the true rate of frequency change is calculated through instantaneous phase differentiation. Specifically: Collect instantaneous values ​​of three-phase voltage after a high-voltage line fault , , Furthermore, its analytic signal representation in the complex plane is constructed using the Hilbert transform. Simultaneously, the pure CVT discharge tail component of the preceding output is... Similarly, by mapping to the complex plane through the Hilbert transform, the complex form of the trailing component is obtained. The expression involved is: ; ;in: , , These are the instantaneous voltage samples of phases A, B, and C after a high-voltage line fault. The instantaneous values ​​of the three-phase voltages after Clarke transformation are at rest. coordinate system Axial components; The imaginary unit; This represents the Hilbert transform operator, used to extract the orthogonal components of a signal; This is the analytical signal representation of the original acquired instantaneous three-phase voltage values ​​in the complex plane; The pure CVT discharge tail component, which is reconstructed from the output of the preceding module, corresponds to a real number signal; For the complex tail component mapped to the complex plane.

[0034] It should be noted that converting the voltage sampling signal and tail component in the real domain into a complex plane analytical signal containing amplitude and phase information can provide a data basis for subsequent accurate vector cancellation operations in the complex plane, ensuring that the cancellation process can simultaneously handle the amplitude attenuation and phase shift of the signal.

[0035] When performing adaptive calculation of the nonlinear phase cancellation gain factor, it is based on the obtained complex plane analytic signal. and complex form trailing components The energy proportion of the trailing component in the original sampled signal is calculated, and the nonlinear phase cancellation gain factor is dynamically adjusted accordingly. The corresponding expression is: ; ;in: The current sampling time The tail component, relative to the energy proportion of the original sampled signal, is a dimensionless parameter; The trailing component in complex form The magnitude of the trailing component represents its instantaneous amplitude. For complex plane analytic signals The magnitude of the original sampled signal represents its instantaneous amplitude. To prevent extremely small positive constants with a denominator of zero, for example, take... ; The current sampling time The nonlinear phase cancellation gain factor is a dimensionless parameter with a value range of [0,1], used to control the cancellation intensity. This is a preset upper limit value for the cancellation gain factor, for example, 0.98, used to prevent excessive cancellation from damaging the real signal. It is a dimensionless parameter. is a nonlinear adjustment coefficient with a value range of [5,50], used to control the response sensitivity of the gain factor as the energy ratio changes. It is a dimensionless parameter, and the specific value can be determined based on existing offline simulation calibration. It is a natural exponential function.

[0036] It should be noted that the cancellation strength is adaptively adjusted according to the actual degree of contamination of the CVT discharge tail component in the current signal. When the tail component has a high energy ratio, such as in the early stage of a fault, the gain factor increases rapidly to enhance the cancellation effect. When the tail component gradually decays, the gain factor decreases smoothly to avoid false cancellation of the pure power frequency recovery voltage, thereby ensuring the dynamic stability of the extraction process.

[0037] When performing complex plane vector cancellation and true recovery voltage orthogonal decomposition, the obtained complex plane analytic signal is used. and complex form trailing components and the calculated nonlinear phase cancellation gain factor Perform complex plane vector cancellation operation to extract the true recovered voltage on the grid side. and restore the actual voltage on the grid side. Decomposed into orthogonal components in a stationary coordinate system, the corresponding expression is: ; ; ;in: This is an analytical signal representation of the extracted true recovered voltage on the grid side in the complex plane. To truly restore voltage at rest coordinate system Orthogonal components of the axes, i.e., the real part; To truly restore voltage at rest coordinate system Orthogonal components of the axes, i.e., the imaginary part; This indicates taking the real part of a complex number; This indicates taking the imaginary part of a complex number.

[0038] It should be noted that by using complex plane vector subtraction, the CVT discharge tail component (5~15Hz subsynchronous oscillation component) superimposed on the original signal is precisely stripped away in terms of both amplitude and phase, thereby obtaining a pure power frequency that is not contaminated by the CVT transient "discharge tail", recovering the voltage signal, and decomposing it into orthogonal components, which can provide a standard input for subsequent instantaneous frequency calculation without delay.

[0039] When performing instantaneous phase derivative and true frequency rate of change smoothing calculations, the true recovery voltage quadrature component is used as the basis. and The instantaneous phase is calculated using the arctangent function, and the actual rate of frequency change is obtained by differentiating the instantaneous phase over time. To suppress high-frequency noise from differentiation operations, a first-order low-pass filter is introduced for smoothing. The corresponding expression is: ; ; ;in: The instantaneous phase of the true recovery voltage; It is the arctangent function in the four quadrants; The rate of change of the original frequency without filtering; Indicates time First derivative operations; This represents the final true frequency change rate after smoothing. This is the time constant of the first-order low-pass filter, used to balance the smoothness of the frequency change rate calculation with the response delay.

[0040] It should be noted that the traditional Fourier algorithm, which relies on a fixed time window, suffers from computational delay and spectral leakage. Instead, it uses the instantaneous phase differentiation method to achieve delay-free and high-precision calculation of the frequency change rate. At the same time, low-pass filtering effectively suppresses the high-frequency noise amplified by the differentiation operation, ensuring that the output frequency change rate can truly reflect the electromechanical transient characteristics of the power grid system and completely eliminates the frequency calculation distortion caused by the CVT transient "discharge tail".

[0041] In this embodiment of the invention, by constructing a nonlinear phase cancellation operator, the CVT discharge tail component is dynamically and accurately removed from the original sampled values ​​in the complex plane. This fundamentally solves the measurement distortion problem caused by the superposition of the 5-15Hz subsynchronous oscillation component on the power frequency recovery voltage, ensuring that the purity of the extracted true recovery voltage reaches over 98%. A nonlinear phase cancellation gain factor based on energy proportion is introduced, which can adaptively adjust the cancellation strength in real time according to the attenuation characteristics of the tail component during the fault transient process. This ensures effective cancellation in the early stage of the fault while avoiding accidental damage to the true power frequency signal at the end of the transient, significantly improving the robustness of the system. The instantaneous phase differential method based on orthogonal components replaces the traditional Fourier algorithm, breaking the time window limitation on frequency calculation speed and achieving millisecond-level zero-delay response of frequency change rate. Combined with low-pass filtering, it effectively suppresses noise interference while ensuring high dynamic response speed, and the calculation accuracy meets the stringent requirements of low-frequency load shedding devices. The true frequency change rate output by this module truly reflects the physical frequency change of the power grid system, completely eliminating the false frequency offset induced by CVT transient "discharge tail". This effectively prevents the low-frequency load shedding device from erroneously shedding normal loads in complex scenarios such as the grid-connected side of doubly-fed wind turbines, ensuring the safe and stable operation of the power system. Electromechanical-electromagnetic coupled oscillation identification and control module: Based on the calculated true frequency change rate and combined with the dynamic response characteristics of the phase-locked loop of the doubly-fed induction generator rotor-side converter, an electromechanical-electromagnetic coupled oscillation identification matrix is ​​constructed. By calculating the eigenvalue locus of the identification matrix, the attributes of the low-frequency oscillation source are determined, and based on the determination result, a blocking or enabling control command is output to the low-frequency load shedding device. Specifically: Real-time acquisition of the output phase angle of the phase-locked loop (PLL) of the rotor-side converter (RSC) of the doubly-fed wind turbine. and output angular frequency And combined with the actual frequency change rate calculated by the preceding module The state deviation vector of the phase-locked loop dynamic response is constructed, and the corresponding expression is: ; ;in: This refers to the deviation between the phase angle output by the phase-locked loop and the actual phase angle of the power grid. The phase angle of the phase-locked loop output of the rotor-side converter of the doubly-fed wind turbine; The phase angle of the actual voltage on the grid side is calculated from the orthogonal components of the actual recovered voltage extracted by the preceding module; This is the angular frequency of the phase-locked loop output. The actual angular frequency on the power grid side is obtained by superimposing the actual frequency change rate calculated by the preceding module onto the rated angular frequency. , The rated frequency of the power grid. For integration variables; This represents the deviation between the output angular frequency of the phase-locked loop and the actual angular frequency of the power grid.

[0042] It should be noted that by quantifying the dynamic tracking deviation of the doubly fed wind turbine phase-locked loop during the transient process, we can provide state input variables on the phase-locked loop side for the subsequent construction of the electromechanical-electromagnetic coupled oscillation identification matrix, and reveal the coupling relationship between the electromagnetic transients of power electronic equipment and the electromechanical transients of the power grid.

[0043] When constructing the electromechanical-electromagnetic coupled oscillation identification matrix, the extracted output phase angle is used as the basis. and output angular frequency Combined with the actual frequency change rate output by the preceding module and its rate of change A fourth-order state-space identification matrix describing the electromechanical-electromagnetic coupling oscillation characteristics is constructed, and the system state vector is defined as follows. Then the state equation expression of the identification matrix is: Among them, the electromechanical-electromagnetic coupled oscillation identification matrix It is a 4×4 matrix, and its expression is: In the formula, A 4×4 electromechanical-electromagnetic coupled oscillation identification matrix is ​​used to describe the coupled oscillation characteristics between the dynamic state of the doubly-fed wind turbine phase-locked loop and the electromechanical transient state of the power grid; , where is the proportional controller gain of the phase-locked loop of the rotor-side converter of the doubly fed wind turbine, and is a dimensionless parameter; The integral controller gain for the phase-locked loop of the rotor-side converter of the doubly-fed wind turbine; The voltage-phase coupling coefficient has a value range of [0.001, 0.1], which characterizes the influence of grid voltage amplitude changes on phase-locked loop phase tracking. The specific value can be calculated analytically through the phase-locked loop small-signal linearization model or obtained by actual measurement. The frequency deviation-electromagnetic torque coupling coefficient, with a value range of [0.01, 2], characterizes the influence of the phase-locked loop frequency tracking deviation on the output electromagnetic torque of the doubly-fed wind turbine. It can be derived from the small-signal model of the active power control loop of the existing rotor-side converter. The equivalent inertial time constant of the doubly fed wind turbine unit has a value range of [2,8] and a default value of 5. is the equivalent damping coefficient of the doubly fed wind turbine unit, which is a dimensionless parameter with a value range of [0.01, 0.15]. The specific value can be determined based on existing offline simulation calibration.

[0044] It should be noted that constructing a fourth-order coupled model that can simultaneously characterize the electromagnetic transients (phase-locked loop dynamics) of power electronic devices and the electromechanical transients (frequency dynamics) of power systems, and unifying the two types of dynamic processes with different time scales within the same state space framework through the identification matrix, can provide a mathematical basis for subsequent identification of the attributes of low-frequency oscillation sources.

[0045] When performing eigenvalue locus calculation and oscillation source attribute discrimination of the identification matrix, the constructed electromechanical-electromagnetic coupled oscillation identification matrix is ​​used. Perform eigenvalue decomposition and calculate its four eigenvalues. , Based on the real and imaginary parts of the eigenvalues, the properties of the low-frequency oscillation source are determined. The corresponding expressions and discrimination rules are as follows: ;in: For identification matrix The ' eigenvalues; For the first The real part of each eigenvalue represents the decay or divergence rate of the corresponding oscillation mode; The imaginary unit; For the first The imaginary part of each eigenvalue represents the angular frequency of the corresponding oscillation mode; The specific discrimination rule is as follows: Rule 1: Determination of electromechanical oscillation sources: If there exists a pair of conjugate complex characteristic roots If the following three conditions are met simultaneously, then a low-frequency oscillation source is determined to be an electromechanical oscillation source: Condition 1: That is, the oscillation frequency is in the typical frequency range of electromechanical oscillation (0.5~2.5Hz). Condition 2: That is, the real part of the characteristic root is positive, indicating that the oscillation has a divergent trend; Condition 3: That is, the absolute value of the real part of the eigenvalue of the electromechanical mode is significantly greater than the sum of the absolute values ​​of the real parts of the other eigenvalues. The modal separation threshold is set to 3, and it is a dimensionless parameter. Rule 2: Determination of electromagnetic oscillation source: If there exists a pair of conjugate complex characteristic roots If all three of the following conditions are met, the low-frequency oscillation source is determined to be an electromagnetic oscillation source, that is, a subsynchronous oscillation dynamically dominated by a phase-locked loop: Condition 1: That is, the oscillation frequency is in the typical frequency band of electromagnetic oscillation; Condition 2: That is, the real part of the characteristic root is positive, indicating that the oscillation has a divergent trend; Condition 3: That is, the absolute value of the real part of the characteristic root of the electromagnetic mode is significantly greater than the sum of the absolute values ​​of the real parts of the other characteristic roots; Rule 3: Determination of Coupled Oscillation Source: If the conditions of Rule 1 and Rule 2 are partially satisfied at the same time, that is, the real parts of the characteristic roots of both the electromechanical mode and the electromagnetic mode are positive, and the mode separation degree does not meet condition 3, then the low-frequency oscillation source is determined to be an electromechanical-electromagnetic coupled oscillation source.

[0046] It should be noted that, through eigenvalue locus analysis, the physical nature of low-frequency oscillations can be precisely distinguished mathematically: whether it is electromechanical oscillation of traditional power systems (0.5~2.5Hz), electromagnetic oscillation caused by power electronic equipment (5~50Hz), or a mixed oscillation resulting from the coupling of the two. This determination directly determines the subsequent control strategy for low-frequency load shedding devices, avoiding misguided one-size-fits-all control.

[0047] Based on the oscillation source attribute determination result, the corresponding control command is generated and output to the low-frequency load shedding device; the specific control logic is as follows: Scenario 1: When the judgment result is an electromechanical oscillation source: The current low-frequency oscillation is determined to be caused by a genuine active power deficit in the power system, resulting in a frequency drop. The low-frequency load shedding device should operate normally to disconnect the load, restore the system frequency, and generate a permissive control command. The expression is: ; At this time, after receiving the permission command, the low-frequency load shedding device will perform the load shedding operation normally according to the preset load shedding cycle and delay logic. Scenario 2: When the determination result is an electromagnetic oscillation source: The current low-frequency oscillation is determined to be a subsynchronous oscillation caused by the dynamics of the doubly-fed induction generator's phase-locked loop, and not a true active power deficit in the power system. Therefore, the low-frequency load shedding device should not activate and a blocking control command should not be generated. The expression is: ; Simultaneously, a phase-locked loop parameter adjustment suggestion signal is sent to the rotor-side converter of the doubly-fed wind turbine. It is recommended to reduce the proportional gain of the phase-locked loop. To enhance system damping, the corresponding expression is: ;in: This is a recommended value for adjusting the proportional gain of the phase-locked loop (PLL), and it is a dimensionless parameter. This is the phase-locked loop gain adjustment coefficient, with a value range of [value range missing]. , is a dimensionless parameter used to control the magnitude of gain downsampling.

[0048] Case 3: When the judgment result is an electromechanical-electromagnetic coupled oscillation source: The current low-frequency oscillation is determined to be the result of the combined effects of electromechanical and electromagnetic transients. Therefore, it is necessary to simultaneously implement load reduction and parameter adjustment measures, and generate condition-permitted control commands. The expression is: ;in: This is the control command output to the low-frequency load shedding device. A value of 1 indicates that the action is allowed, and a value of 0 indicates that the action is blocked. It is a dimensionless logic variable. Let be the real part of the characteristic roots of the electromechanical modes; Let be the real part of the characteristic roots of the electromagnetic mode; When the dominance of the electromechanical mode is stronger than that of the electromagnetic mode, the dominance of the electromechanical mode is measured by the absolute value of the real part of the characteristic root, allowing the low-frequency load shearing device to operate; conversely, it is locked, and electromagnetic oscillation components are suppressed preferentially by adjusting the phase-locked loop parameters. At the same time, regardless of whether it is allowed or locked, a phase-locked loop parameter adjustment suggestion signal is sent to the rotor-side converter of the doubly-fed wind turbine. .

[0049] It should be noted that, based on the accurate identification results of the oscillation source attributes, the system intelligently decides whether to allow the low-frequency load shedding device to operate and simultaneously optimizes the phase-locked loop parameters of the doubly-fed wind turbine. Through differentiated control strategies, the system not only ensures the reliable operation of the low-frequency load shedding device in real active power deficit scenarios, but also effectively prevents false load shedding caused by false low-frequency oscillations induced by power electronic equipment. At the same time, the system enhances its overall damping capability through adaptive adjustment of the phase-locked loop parameters.

[0050] In this embodiment of the invention, by constructing a fourth-order electromechanical-electromagnetic coupled oscillation identification matrix, the electromagnetic transients (phase-locked loop dynamics) of power electronic equipment and the electromechanical transients (frequency dynamics) of the power system are modeled in a unified manner. Using eigenvalue locus analysis, the oscillation source attributes are accurately identified, effectively improving the identification accuracy. Simultaneously, it can effectively distinguish between three scenarios: electromechanical oscillations (0.5~2.5Hz), electromagnetic oscillations (5~50Hz), and coupled oscillations. Based on the oscillation source attributes, differentiated blocking or enabling control commands are output, fundamentally solving the technical problem of false low-frequency oscillations caused by phase-locked loop dynamics in doubly-fed wind turbine grid-connected scenarios, which leads to the erroneous shedding of normal loads. In the blocking... Simultaneously with the low-frequency load shedding device, a phase-locked loop (PLL) parameter adjustment suggestion signal is sent to the rotor-side converter of the doubly-fed induction generator (DFIG). By reducing the PLL proportional gain, the system damping is actively enhanced, suppressing electromagnetic oscillations at the source and preventing the oscillations from continuously worsening and threatening grid safety. For the complex scenario of electromechanical-electromagnetic coupling oscillations, a conditional allowable control strategy based on the strength of mode dominance is proposed. Under the premise of ensuring grid frequency safety, unnecessary load shedding is avoided to the greatest extent, achieving an optimal balance between safety and economy. The entire identification and control process is based on matrix eigenvalue decomposition and control logic judgment, with low computational complexity, meeting the real-time requirements of power system relay protection and automatic control devices.

[0051] In the several embodiments provided by this invention, it should be understood that the disclosed system can be implemented in other ways. For example, the embodiments of the invention described above are merely illustrative; for example, the division of modules is only a logical functional division, and there may be other division methods in actual implementation.

[0052] The modules described as separate components may or may not be physically separate. The components shown as modules may or may not be physical modules; they may be located in one place or distributed across multiple network modules. Some or all of the modules can be selected to achieve the purpose of this embodiment according to actual needs.

[0053] Furthermore, the functional modules in the various embodiments of the present invention can be integrated into one processing module, or each module can exist physically separately, or two or more modules can be integrated into one module. The integrated module described above can be implemented in hardware or in the form of hardware plus software functional modules.

[0054] It will be apparent to those skilled in the art that the present invention is not limited to the details of the exemplary embodiments described above, and that the present invention can be implemented in other specific forms without departing from the essential characteristics of the present invention.

[0055] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and are not intended to limit it. Although the present invention has been described in detail with reference to preferred embodiments, those skilled in the art should understand that modifications or equivalent substitutions can be made to the technical solutions of the present invention without departing from the spirit and scope of the technical solutions of the present invention.

Claims

1. An intelligent control system for power fault data acquisition and processing, characterized in that, include: CVT transient discharge tail component reconstruction module: Collects the instantaneous values ​​of three-phase voltage after a high-voltage line fault, constructs a fractional-order state-space observer based on the topology of the CVT internal main capacitor and voltage divider capacitor, reconstructs the transient discharge charge trajectory of the CVT internal capacitor bank by solving fractional-order differential equations, and separates the CVT discharge tail component. The module for extracting the true recovery voltage and calculating the rate of change of frequency is as follows: Based on the reconstructed CVT discharge tail component, a nonlinear phase cancellation operator is constructed. The CVT discharge tail component is used as an adaptive cancellation factor and is used to perform complex plane vector cancellation with the original acquired instantaneous values ​​of the three-phase voltage to extract the true recovery voltage on the grid side. Based on this true recovery voltage, the true rate of change of frequency is calculated by instantaneous phase differentiation. Electromechanical-electromagnetic coupling oscillation identification and control module: Based on the calculated real frequency change rate and combined with the dynamic response characteristics of the phase-locked loop of the doubly fed wind turbine rotor-side converter, an electromechanical-electromagnetic coupling oscillation identification matrix is ​​constructed. By calculating the eigenvalue trajectory of the identification matrix, the attributes of the low-frequency oscillation source are determined, and a blocking or enabling control command is output to the low-frequency load shedding device according to the identification result.

2. The intelligent control system for power fault data acquisition and processing according to claim 1, characterized in that, Under normal operating conditions of the high-voltage line, a multi-frequency sweep excitation signal is injected into the secondary side of the CVT, and the input voltage and output voltage response data of the CVT are collected. Based on the collected input and output response data, the physical topology parameters of the CVT are identified. At the same time, by analyzing the relationship curve between the hysteresis loop area and frequency of the discharge coil core material, the fractional order characterizing the dielectric memory effect of the capacitor is determined.

3. The intelligent control system for power fault data acquisition and processing according to claim 2, characterized in that, Based on the identified physical topology parameters and fractional order of the CVT, a fractional-order state-space equation describing the transient discharge process of the capacitor bank inside the CVT is established. The charge on both ends of the main capacitor C1, the charge on both ends of the voltage divider capacitor C2, and the magnetic flux of the discharge coil are selected as state variables to construct a third-order fractional-order state-space observer model.

4. The intelligent control system for power fault data acquisition and processing according to claim 3, characterized in that, Starting from the fault triggering moment, the constructed fractional-order state-space observer model is discretized and solved; at each sampling moment, the state vector at the current moment is calculated based on the instantaneous value of the three-phase voltage collected at the current moment and the state vector at the previous moment.

5. The intelligent control system for power fault data acquisition and processing according to claim 4, characterized in that, After calculating the state vector at the current moment, the CVT discharge tail component at the current moment is calculated through the output equation.

6. The intelligent control system for power fault data acquisition and processing according to claim 5, characterized in that, When implementing the adaptive calculation of the nonlinear phase cancellation gain factor, the energy ratio of the tail component in the original sampled signal is calculated based on the calculated complex plane analytic signal and the complex form tail component, and the nonlinear phase cancellation gain factor is dynamically adjusted accordingly.

7. The intelligent control system for power fault data acquisition and processing according to claim 6, characterized in that, When performing complex plane vector cancellation and true recovery voltage orthogonal decomposition, the obtained complex plane analytic signal and complex form tail component, as well as the calculated nonlinear phase cancellation gain factor, are used to perform complex plane vector cancellation operation, extract the grid-side true recovery voltage, and decompose the grid-side true recovery voltage into orthogonal components in the stationary coordinate system.

8. The intelligent control system for power fault data acquisition and processing according to claim 7, characterized in that, When constructing the electromechanical-electromagnetic coupled oscillation identification matrix, a fourth-order state-space identification matrix describing the characteristics of electromechanical-electromagnetic coupled oscillation is constructed based on the output phase angle and output angular frequency obtained through processing, combined with the actual frequency change rate and its change rate output by the preceding module.

9. The intelligent control system for power fault data acquisition and processing according to claim 8, characterized in that, When performing eigenvalue locus calculation and oscillation source attribute discrimination of the identification matrix, the constructed electromechanical-electromagnetic coupled oscillation identification matrix is ​​subjected to eigenvalue decomposition, and its four eigenvalues ​​are calculated. The roots are used to determine the properties of low-frequency oscillation sources based on their real and imaginary parts.

10. The intelligent control system for power fault data acquisition and processing according to claim 9, characterized in that, Based on the oscillation source attribute determination result, the corresponding control command is generated and output to the low-frequency load shedding device.

Images