Island power grid disturbance source first arrival check positioning method and system

Abstract

This invention discloses a method for first-arrival verification and location of disturbance sources in isolated power grids, relating to the field of broadband oscillation monitoring and control technology in new energy power grids. The method involves synchronously acquiring voltage and current data from multiple monitoring points, separating the target oscillation mode, extracting the first-arrival time sequence of the mode energy envelope, and generating candidate disturbance source regions by combining the topological distance between monitoring points, line impedance, and phase gradient. Subsequently, a low-amplitude, short-time diagnostic micro-disturbance signal is injected by a broadband oscillation suppression device to calculate the equivalent harmonic impedance change rate, damping sensitivity, and response phase shift of the candidate nodes. Finally, the equivalent harmonic impedance change rate, damping sensitivity, response phase shift, first-arrival location score, and mode energy transmission intensity are fused to calculate the disturbance source credibility of each candidate node. Based on the disturbance source credibility, the disturbance source location result is output, and a corresponding minimum control strategy is generated. This invention can improve the accuracy and control targeting of broadband oscillation disturbance source location in isolated new energy power grids.

Description

Technical Field

[0001] This invention relates to the field of broadband oscillation monitoring and control technology for new energy power grids, specifically a method and system for locating and verifying the first arrival of disturbance sources in an islanded power grid. Background Technology

[0002] With the high proportion of new energy sources, power electronic converters, and flexible loads integrated into islanded power grids, problems such as decreased system inertia, insufficient damping, and time-varying impedance characteristics are becoming increasingly prominent, easily inducing multi-band broadband oscillations, including subsynchronous, synchronous, and supersynchronous oscillations. Existing disturbance source location methods mostly rely on amplitude changes at a single monitoring point, steady-state energy flow direction, or offline impedance analysis, which are difficult to adapt to scenarios with multiple oscillation modes coupled, complex propagation paths, and rapidly changing source-load characteristics in islanded new energy power grids. Especially when multiple converters are operating in parallel, oscillation signals will be reflected, superimposed, and phase-shifted in weak grid topologies. Judging the disturbance source solely based on oscillation amplitude or energy direction is prone to misjudgment or location delays. Directly adopting large-scale disconnection or uniform damping control measures may result in excessive control costs and reduced power supply continuity. Therefore, there is an urgent need for a disturbance source location method that can combine the initial characteristics of oscillation propagation with impedance response characteristics to improve location accuracy and control targeting. Summary of the Invention

[0003] To address the problem that existing methods for locating broadband oscillation disturbance sources in isolated renewable energy power grids are susceptible to multi-source coupling, propagation reflection, and load fluctuations, resulting in insufficient location accuracy and weak control measures, this invention proposes a first-arrival verification and location method and system for disturbance sources in isolated power grids.

[0004] To achieve the above objectives, the present invention proposes the following technical solution:

[0005] A method for first-reach verification and location of disturbance sources in an isolated power grid includes the following steps: acquiring synchronous voltage and current sampling data from multiple monitoring points in an isolated renewable energy power grid, and performing time synchronization, anomaly removal, and broadband filtering on the sampling data to obtain preprocessed voltage and current complex signals; performing target oscillation mode separation on the preprocessed voltage and current complex signals, and extracting the frequency, amplitude, phase, damping ratio, and mode energy envelope of the target oscillation modes; generating candidate disturbance source regions based on the first time the mode energy envelope of each monitoring point exceeds the limit, combined with the topological distance of the monitoring points, line impedance, and phase gradient, and calculating the first-reach location score and mode energy transmission intensity of each candidate node; controlling a broadband oscillation suppression device to inject a low-amplitude short-time diagnostic micro-disturbance signal into the grid connection point corresponding to the candidate disturbance source region; and calculating the voltage before and after the injection of the diagnostic micro-disturbance signal. The system calculates the equivalent harmonic impedance change rate, damping sensitivity, and response phase shift for each candidate node based on the current response change. These parameters are then normalized and input into a weighted fusion model along with the first-reach location score and the modal energy transmission intensity to calculate the disturbance source confidence of each candidate node. The first-reach location score is determined based on the envelope first-reach timing, monitoring point topology distance, line impedance, and phase gradient. The equivalent harmonic impedance change rate characterizes the electrical response sensitivity of the candidate node to diagnostic perturbation signals, the damping sensitivity characterizes the negative damping contribution of the candidate node, and the response phase shift characterizes the perturbation response phase characteristics of the candidate node. The disturbance source location result is output based on the disturbance source confidence, and a corresponding minimum control strategy is generated.

[0006] Furthermore, the multiple monitoring points include the grid connection point of the new energy converter, the grid connection point of the energy storage converter, the load branch node, and the common connection point; the synchronous voltage and current sampling data cover a wide frequency range from 0.1Hz to 2500Hz, and the sampling rate is not less than 12.8kHz.

[0007] Furthermore, the target oscillation mode separation includes broadband mode search based on frequency adaptive tracking, sliding window spectral peak locking, and multi-mode decoupling processing; when multiple oscillation modes are detected to exist simultaneously, the target oscillation mode to be located is determined according to the mode energy ratio and damping attenuation trend.

[0008] Furthermore, the first time the modal energy envelope exceeds the limit is the starting time when the energy envelope of the target oscillation mode continuously exceeds the background noise reference value by a preset multiple; the first arrival time series matrix is ​​constructed based on the first time difference of different monitoring points, and the oscillation propagation starting point is deduced by combining the grid topology adjacency matrix to form a disturbance source candidate region.

[0009] Furthermore, the diagnostic perturbation signal is a biased sinusoidal signal, a narrowband swept frequency signal, or a pseudo-random binary wideband signal located near the target oscillation frequency; the amplitude of the diagnostic perturbation signal is limited to a safe threshold that does not affect the normal power supply of the islanded power grid, and the injection duration is shorter than the preset diagnostic window.

[0010] Furthermore, the equivalent harmonic impedance change rate is determined as follows: the perturbation equivalent harmonic impedance is determined based on the ratio of the voltage phasor difference to the current phasor difference at the target frequency before and after the diagnostic perturbation signal injection; the equivalent harmonic impedance change rate is then determined based on the ratio of the difference between the perturbation equivalent harmonic impedance and the equivalent harmonic impedance before injection to the equivalent harmonic impedance before injection; the damping sensitivity is determined based on the change amplitude of the damping ratio of the target oscillation mode before and after perturbation injection.

[0011] Furthermore, the credibility of the disturbance source is calculated using a weighted fusion model. The inputs of the weighted fusion model include the first-reach location score, modal energy transmission intensity, equivalent harmonic impedance change rate, damping sensitivity, and response phase offset. The first-reach location score is determined based on the envelope first-reach time difference, phase gradient consistency, topological distance, and line impedance between the candidate node and its adjacent nodes. The modal energy transmission intensity is determined based on the average value of the real part of the conjugate product of the complex voltage signal of the target oscillation mode of the candidate node and the complex current signal of the target oscillation mode along the transmission direction of the node within a preset time window. When the credibility of the disturbance source of a candidate node is greater than a preset threshold, the candidate node is determined to be the location of the disturbance source.

[0012] Furthermore, the minimum control strategy includes target frequency band impedance reshaping, local damping injection, output limiting, power reduction operation, or selective cutoff; when impedance reshaping or local damping injection can reduce the oscillation amplitude to a preset threshold, non-cutoff control is preferentially executed; when the credibility of the disturbance source continues to increase and non-cutoff control is ineffective, selective cutoff control is executed.

[0013] A system for locating and verifying the first arrival of disturbance sources in an isolated power grid includes: a multi-point synchronous acquisition module connected to multiple monitoring points in an isolated renewable energy power grid, used to synchronously acquire voltage and current sampling data from each monitoring point, and to perform time synchronization, anomaly removal, and broadband filtering on the sampled data, outputting pre-processed voltage and current complex signals; a target oscillation mode separation module, used to perform target oscillation mode separation on the pre-processed voltage and current complex signals, extracting the frequency, amplitude, phase, damping ratio, and mode energy envelope of the target oscillation mode; a first arrival time series analysis module, used to generate candidate disturbance source regions based on the first time the mode energy envelope of each monitoring point exceeds the limit, combined with the topological distance between monitoring points, line impedance, and phase gradient, and to calculate the first arrival location score and mode energy transmission intensity of each candidate node; and a diagnostic micro-disturbance injection module, used to control broadband oscillation. The suppression device injects a low-amplitude, short-time diagnostic micro-disturbance signal into the grid connection point corresponding to the candidate disturbance source region; the impedance response verification module is used to calculate the equivalent harmonic impedance change rate, damping sensitivity, and response phase shift of each candidate node based on the voltage and current response changes before and after the injection of the diagnostic micro-disturbance signal; the disturbance source credibility determination module is used to normalize the equivalent harmonic impedance change rate, the damping sensitivity, and the response phase shift, and input them together with the first-reach location score and the modal energy transmission intensity into a weighted fusion model to calculate the disturbance source credibility of each candidate node; wherein, the first-reach location score is determined based on the envelope first-reach timing, monitoring point topology distance, line impedance, and phase gradient; the control strategy generation module is used to output the disturbance source location result based on the disturbance source credibility and generate the corresponding minimum control strategy.

[0014] Furthermore, the diagnostic perturbation injection module performs a safety check on the islanded power grid before injecting the diagnostic perturbation signal. When the voltage deviation at the point of common coupling is greater than a preset percentage of the rated value, the frequency deviation is greater than a preset value, or the energy storage system is in a power-limited state, the perturbation injection is suspended.

[0015] Compared with the prior art, the technical effects of the present invention are as follows:

[0016] First, this invention does not rely solely on oscillation amplitude or steady-state energy flow to determine the disturbance source, but instead uses the envelope arrival time sequence at the initial stage of oscillation to identify the starting point of disturbance propagation, which can reduce the interference of propagation reflection and multi-node coupling on the positioning results.

[0017] Second, the present invention introduces a diagnostic perturbation impedance response verification mechanism, which obtains the impedance response changes and damping sensitivity of candidate nodes through low-amplitude short-time perturbation signals, thereby achieving secondary confirmation of the candidate region of the perturbation source and improving the reliability of perturbation source location.

[0018] Third, the present invention uses a credibility fusion method to output the positioning results, which can not only give a single disturbance source judgment, but also form a candidate disturbance source ranking, making it more suitable for islanded new energy power grid scenarios with multiple source coupling or source-load interaction.

[0019] Fourth, this invention associates the disturbance source location results with the minimum control strategy, and prioritizes non-cut-off control methods such as local damping injection or target frequency band impedance reshaping. While ensuring the oscillation suppression effect, it reduces false cut-off and over-control, thereby improving the safety, economy and continuity of islanded power grid operation. Attached Figure Description

[0020] To more clearly illustrate the technical solution of the present invention, the accompanying drawings used in the specific embodiments are briefly described below. It should be understood that the following drawings are only used to illustrate some embodiments of the present invention and do not constitute a limitation on the scope of protection of the present invention.

[0021] Figure 1 This is a schematic diagram of the overall process of the method for first-reach verification and location of disturbance sources in an isolated power grid according to an embodiment of the present invention;

[0022] Figure 2 This is a schematic diagram of the generation of candidate regions for disturbance sources based on the first arrival time of modal energy envelope in an embodiment of the present invention;

[0023] Figure 3 This is a schematic diagram illustrating the fusion determination of diagnostic perturbation impedance response verification and perturbation source credibility in an embodiment of the present invention. Detailed Implementation

[0024] To make the objectives, technical solutions, and advantages of the embodiments of the present invention clearer, the technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, not all embodiments. 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.

[0025] This embodiment provides a method for locating the first arrival point of disturbances in an isolated power grid, using a specific island-type renewable energy isolated power grid as an application scenario. This isolated power grid includes a 35kV common connection bus, a photovoltaic inverter cluster, a wind power converter cluster, a grid-connected energy storage converter, industrial power electronic loads, and several 10kV feeder branches. Because this power grid operates independently from the main power grid, its equivalent short-circuit capacity is low. Changes in renewable energy output and power electronic loads easily lead to rapid changes in the impedance characteristics at the grid connection point, resulting in wideband oscillations in the subsynchronous, synchronous, and supersynchronous frequency bands.

[0026] In this embodiment, synchronous monitoring units are arranged at the common connection point, photovoltaic grid connection point, wind power grid connection point, energy storage grid connection point, and main load branch, forming a configuration as follows: Figure 1 The system shown is a first-arrival verification and location system for disturbance sources in an isolated power grid. The system includes a multi-point synchronous acquisition module, a target oscillation mode separation module, a first-arrival time series analysis module, a diagnostic perturbation injection module, an impedance response verification module, and a disturbance source credibility determination module. Steps one through three mainly complete broadband signal acquisition, target oscillation mode extraction, and disturbance source candidate region generation. The first-arrival time series analysis process is as follows: Figure 2 As shown. The method includes the following steps:

[0027] Step 1: Acquisition and preprocessing of synchronous data from multiple monitoring points

[0028] In this embodiment, five typical monitoring points are selected as the data source for locating disturbance sources. Each monitoring point is synchronized using BeiDou time synchronization or an IEEE 1588 precision clock to ensure that the time scale deviation between different monitoring points is less than 10μs. Each monitoring point synchronously acquires three-phase voltage and three-phase current signals, with a sampling rate set to 12.8kHz, an A / D conversion accuracy of no less than 16 bits, and a monitoring frequency band covering 0.1Hz to 2500Hz.

[0029] The configuration of each monitoring point is shown in the table below:

[0030] Table 1 Configuration information for each monitoring point

[0031]

[0032] The three-phase voltage and current signals collected at the i-th monitoring point are denoted as u. a,i (t), u b,i (t), u c,i (t) and i a,i (t), i b,i (t), i c,i (t). To facilitate subsequent modal analysis, the three-phase instantaneous quantities are first subjected to αβ transformation to obtain two-dimensional orthogonal signals:

[0033] ,

[0034] ,

[0035] ,

[0036] ,

[0037] In the formula, i represents the monitoring point number, and t represents the sampling time. , These represent the components of the voltage signal at the i-th monitoring point on the α-axis and β-axis, respectively. , These represent the components of the current signal at the i-th monitoring point on the α-axis and β-axis, respectively.

[0038] Subsequently, the signals from each monitoring point are processed through time alignment, outlier sampling removal, anti-aliasing filtering, and wideband pass filtering. The passband range of the wideband pass filter is set to 0.1Hz to 2500Hz to preserve subsynchronous, synchronous, and supersynchronous oscillation components while suppressing DC bias, high-frequency measurement noise, and transient glitches. The preprocessed voltage and current complex signals are expressed as follows:

[0039] ,

[0040] ,

[0041] In the formula, j is the imaginary unit. The voltage complex signal at the i-th monitoring point Let be the complex current signal at the i-th monitoring point.

[0042] In this embodiment, when the broadband voltage disturbance at any monitoring point exceeds 0.5% of the rated voltage, or when the modal energy of any target frequency band is higher than the background noise energy threshold for three consecutive sliding windows, the system initiates the disturbance source first arrival verification and location process, and records the start time as t0.

[0043] Step 2: Target Oscillation Mode Separation and Parameter Extraction

[0044] After completing synchronous acquisition and preprocessing at multiple monitoring points, the system performs target oscillation mode separation on the complex voltage and complex current signals at each monitoring point. This embodiment adopts a processing method of "frequency adaptive tracking - sliding window spectral peak locking - narrowband mode reconstruction" to adapt to the operating characteristics of fundamental frequency drift and multi-oscillation mode coupling in isolated renewable energy power grids.

[0045] First, in a length of T w Calculate the short-time spectrum of the complex voltage signal at the i-th monitoring point within the sliding time window:

[0046]

[0047] In the formula, Here, f represents the spectral result of the i-th monitoring point within the time window τ, and f is the analysis frequency. Let f be a sliding window function, where N is the window length and the number of sampling points. s The sampling frequency.

[0048] Then, candidate oscillation frequencies are searched based on the energy peak values ​​at each frequency point, and normal fundamental frequency fluctuations are eliminated by combining the results of adaptive tracking of the fundamental frequency. The target oscillation frequency is f. o Determine by the following formula:

[0049]

[0050] In the formula, f o Ω represents the target oscillation frequency, m represents the wideband search frequency, and m represents the number of monitoring points minus one. Let f be the modal energy of the i-th monitoring point at frequency f.

[0051] When multiple oscillation modes exist simultaneously in a system, the target oscillation mode for priority processing is determined according to the mode energy ratio, amplitude growth rate, and damping decay trend. Its comprehensive priority index P o,k Defined as:

[0052]

[0053] In the formula, P o,k E is the priority index for the k-th oscillation mode. k Let A be the energy of the k-th mode. k Let the amplitude of the k-th mode be . Let R be the damping ratio of the k-th mode, and R be the number of currently detected modes. , , Let be the weight coefficient, and satisfy... .

[0054] Determine the target oscillation frequency f o Then, at each monitoring point, a system based on f was constructed. o An adaptive narrowband filter centered on the target oscillation mode component x is used to extract the target oscillation mode component x. i (t). The target oscillation mode component can be a voltage mode component or a current mode component. In this embodiment, the voltage mode component is preferred for first-arrival analysis, and the current mode component is used to assist in determining the energy propagation direction.

[0055] The analytical signal of the target oscillation mode is obtained by Hilbert transform:

[0056]

[0057] In the formula, z i (t) represents the analytical signal of the target oscillation mode at the i-th monitoring point, x i (t) represents the time-domain component of the target oscillation mode. For x i Hilbert transform of (t).

[0058] The modal energy envelope, instantaneous phase, and instantaneous frequency can be further obtained from the analytic signal:

[0059]

[0060]

[0061]

[0062] In the formula, Let i be the modal energy envelope of the i-th monitoring point. For instantaneous phase, This is the instantaneous frequency. The above parameters are used for the envelope first-arrival timing analysis and phase gradient direction determination in subsequent step three.

[0063] Step 3: Generation of candidate regions for disturbance sources based on the first arrival time of the envelope

[0064] like Figure 2 As shown, after completing the separation of the target oscillation modes, the system uses the moment when the modal energy envelope of each monitoring point first exceeds the limit to determine the propagation starting point of the oscillation disturbance. Unlike the localization method based solely on the steady-state amplitude or energy flow direction of the oscillation, this embodiment focuses on extracting the first arrival time sequence characteristics at the initial stage of oscillation to reduce the impact of propagation reflection, node coupling, and multi-source superposition on the determination of the disturbance source.

[0065] First, adaptive thresholds for each monitoring point are established based on the background noise level before the oscillation starts. The mean background envelope value of the i-th monitoring point is expressed as:

[0066]

[0067] In the formula, T b This serves as the background statistical time window before the oscillation begins. Let be the mean of the background envelope at the i-th monitoring point.

[0068] The threshold for determining the first arrival at the i-th monitoring point is set as follows:

[0069]

[0070] In the formula, Let be the envelope threshold for the i-th monitoring point. The threshold coefficient is set to 3 to 5 in this embodiment.

[0071] When the modal energy envelope A of the i-th monitoring point i (t) continuously exceeds the threshold And the duration is not less than T c The moment when this condition is first met is defined as the first arrival time of the envelope at that monitoring point:

[0072]

[0073] In the formula, t a,i Let T be the first arrival time of the envelope of the i-th monitoring point. c This is the continuous over-limit confirmation time, used to avoid misjudgments caused by transient noise.

[0074] Subsequently, the first-arrival timing matrix T is constructed. a :

[0075]

[0076] If the arrival time of a certain monitoring point is significantly earlier than that of other monitoring points, and the phase gradient and modal energy flow direction on its adjacent branches both show characteristics of propagation outward from that node, then that monitoring point and its adjacent topological region are identified as candidate regions for disturbance sources. For any two adjacent monitoring points M... i and M j The first arrival time difference is defined as:

[0077]

[0078] when When, it indicates that the target oscillation mode reaches M earlier. i If the phase difference direction is consistent with the line topology direction, then the oscillation is likely caused by M. i To M j Propagation. The phase gradient consistency index is defined as:

[0079]

[0080] In the formula, D ij As a phase gradient consistency index, , M respectively i and M j The target mode phase near the first analysis time. When D ij When the value is greater than 0, the timing direction on this branch is considered to be consistent with the phase propagation direction.

[0081] To comprehensively consider first-reach timing, topology distance, and line impedance, this embodiment calculates a first-reach positioning score for each candidate node:

[0082]

[0083] In the formula, S i The first-reach positioning score is given to the i-th node. Z is the set of nodes adjacent to node i. ij Let be the equivalent impedance of the line between node i and node j. , , Let be the weight coefficient, and satisfy... .

[0084] In this embodiment, if the modal energy envelope exceeding the target frequency of 187Hz first occurs at the M1 photovoltaic grid connection point, followed by simultaneous envelope elevation at the M0 common connection bus, the M3 energy storage grid connection point, and the M4 load branch, and the phase gradient from M1 to M0 is consistent with the energy propagation direction, then the system marks the photovoltaic cluster where M1 is located and its adjacent grid-connected lines as the first candidate region for disturbance source; if the first arrival score of the M4 load branch is the second highest, then the M4 load branch is marked as the second candidate region. These candidate regions are not directly used as the final disturbance source conclusion, but rather as subsequent... Figure 3 The input shown is used for diagnostic perturbation impedance response verification and reliability fusion judgment, thereby avoiding mislocation caused by relying solely on the first arrival timing.

[0085] Step 4: Diagnostic perturbation signal injection and candidate node response acquisition

[0086] After obtaining the candidate region of the disturbance source in step three, the system does not immediately perform cut-off or power limiting control, but instead enters a process such as... Figure 3 The diagnostic perturbation verification stage is shown. In this embodiment, assuming that step three initially determines that the photovoltaic grid connection point M1 is the first candidate perturbation source and the power electronic load branch M4 is the second candidate perturbation source, the broadband oscillation suppression device sequentially injects low-amplitude, short-duration diagnostic perturbation signals into the grid connection points corresponding to M1 and M4 to obtain the impedance response characteristics of the candidate nodes near the target oscillation frequency.

[0087] The diagnostic perturbation signal is preferably selected from the target oscillation frequency f. o The expression for the nearby biased sinusoidal signal is:

[0088]

[0089] In the formula, s d (t) represents the diagnostic perturbation signal, A d f is the amplitude of the perturbation. o Let φ be the target oscillation frequency, Δf be the frequency deviation, and φ be the frequency offset. d To inject the initial phase.

[0090] In this embodiment, the target oscillation frequency f o The frequency is 187Hz, the frequency offset Δf ranges from 3Hz to 8Hz, and the duration of the perturbation signal is T. d Take a time interval of 40ms to 80ms, with a perturbation amplitude A. dThe perturbation injection is limited to 0.2% to 1.0% of the rated voltage or rated current. To avoid the impact of perturbation injection on the normal operation of the islanded power grid, the system performs safety checks on the bus voltage deviation, frequency deviation, and energy storage state of charge before injection. When the voltage deviation at the point of common coupling is greater than 5% of the rated value, the frequency deviation is greater than 0.5Hz, or the energy storage system is in a power-limited state, perturbation injection is suspended, and only the passive first-reach positioning result is retained.

[0091] During the diagnostic perturbation injection, each monitoring point continues to synchronously record voltage and current data according to the sampling method in step one, and the time window before injection, the time window during injection, and the recovery time window after injection are respectively denoted as T. pre T in and T post By using the above segmented sampling, the system can distinguish between natural oscillation responses and artificial perturbation responses, avoiding misjudging the increase in the original oscillation amplitude as the impedance-sensitive characteristic of candidate nodes.

[0092] Step 5: Calculation of Harmonic Impedance Response and Damping Sensitivity of Candidate Nodes

[0093] After completing the diagnostic perturbation signal injection, the system performs frequency domain analysis on the voltage and current responses of the candidate node and its adjacent nodes, and calculates the equivalent harmonic impedance change rate, response phase shift, and damping sensitivity near the target frequency.

[0094] For the i-th candidate node, at the diagnostic perturbation frequency f d =f o At +Δf, the voltage phasor and current phasor before and after injection are extracted respectively, denoted as... , , , The perturbation equivalent harmonic impedance of the candidate node is expressed as:

[0095]

[0096] In the formula, For the i-th candidate node at the diagnostic perturbation frequency f d The equivalent harmonic impedance at the location.

[0097] To determine whether candidate nodes exhibit abnormal sensitivity to the target oscillation frequency band, the equivalent harmonic impedance change rate η is further defined. Z,i :

[0098]

[0099] In the formula, Let ε be the equivalent harmonic impedance of the i-th candidate node before injection, and let ε be a very small positive number to prevent the denominator from being zero.

[0100] Simultaneously, the system calculates the damping sensitivity K based on the change in the damping ratio of the target oscillation mode before and after injection. i :

[0101]

[0102] In the formula, The target modal damping ratio before perturbation injection. K represents the target modal damping ratio during perturbation injection. i Let be the damping sensitivity of the i-th candidate node.

[0103] Response phase offset δφ i Calculate using the following formula:

[0104]

[0105] When the impedance change rate, damping sensitivity, and response phase shift of a candidate node are significantly higher than those of other candidate nodes, it indicates that the node has a stronger negative damping contribution or impedance anomaly characteristics near the target oscillation frequency band, which can serve as an important basis for improving the credibility of the disturbance source.

[0106] Step Six: Determining the Credibility of Disturbance Sources

[0107] After completing the first-reach timing analysis and perturbation impedance response verification, the system performs a credibility fusion judgment on each candidate node. The credibility fusion uses the first-reach positioning score S obtained in step three. i and modal energy transmission intensity P i And the equivalent harmonic impedance change rate η obtained in step five. Z,i Damping sensitivity K i and response phase offset δφ i A unified normalization process is performed. Among these, the phase gradient consistency D... ij It has been used for the first-time location scoring S i The calculation is no longer used as an independent input for credibility fusion.

[0108] The modal energy transmission intensity P i Calculated based on the complex voltage signal and complex current signal of the target oscillation mode of the candidate node:

[0109]

[0110] In the formula, P i Let T be the modal energy transmission intensity of the i-th candidate node. p For the modal energy statistical time window, t a,i Let u be the first arrival time of the envelope of the i-th candidate node. o,i (t) and i o,i(t) represents the complex voltage signal and the complex current signal of the target oscillation mode of the i-th candidate node, respectively, and * denotes complex conjugate. The real part is taken; the positive direction of the current is defined as flowing from the candidate node to the grid, P i The larger the value, the greater the intensity of the target oscillation energy propagated outward by the candidate node.

[0111] The credibility C of the comprehensive disturbance source of the i-th candidate node i Defined as:

[0112]

[0113] In the formula, C i Let represent the confidence level of the disturbance source for the i-th candidate node. To achieve a normalized first-response positioning score, To normalize the modal energy transfer intensity, To normalize the rate of change of equivalent harmonic impedance, To normalize the damping sensitivity, Let w1, w2, w3, w4, and w5 be the normalized response phase shift, and let w1, w2, w3, w4, and w5 be weighting coefficients that satisfy the following:

[0114] w1+w2+w3+w4+w5=1

[0115] In this embodiment, if the system emphasizes rapid localization, w1 and w2 are increased; if the system is in a multi-source coupling or significant propagation and reflection condition, w3 and w4 are increased to enhance the weight of the diagnostic perturbation verification results. By default, w1, w2, w3, w4, and w5 are set to 0.25, 0.20, 0.25, 0.20, and 0.10, respectively.

[0116] When the confidence level C of a candidate node i Greater than the preset confidence threshold C th And the confidence difference between it and the second candidate node is greater than ΔC. th When the node is identified as the primary disturbance source, the system outputs the disturbance source ranking result and enters the multi-point collaborative control mode when the credibility of multiple candidate nodes is close.

[0117]

[0118] In the formula, i * This is the final identification of the main disturbance source node.

[0119] Step 7: Generate the minimum control strategy based on the positioning results

[0120] The system generates a minimum control strategy based on the reliability and ranking results of the disturbance sources output in step six. The minimum control strategy follows the principles of "local first, then global; non-cut-off first, then cut-off; target frequency band control first, then broadband control" to avoid affecting the power supply continuity of the islanded power grid due to misjudgment or over-control.

[0121] In this embodiment, the control strategies for different confidence levels and oscillation severity are shown in the table below:

[0122] Table 2. Control strategies for different confidence levels and oscillation severity.

[0123]

[0124] The impedance reshaping control quantity for the target frequency band can be expressed as:

[0125]

[0126] In the formula, u c (s) represents the modified control command, u ref (s) represents the original reference instruction, Z v (s) represents the virtual impedance of the target frequency band, i h (s) represents the current component of the target oscillation frequency band.

[0127] When performing local damping injection, the broadband oscillation suppression device injects a damping current that is opposite to the target oscillation velocity component into the disturbance source node or its adjacent grid connection point:

[0128]

[0129] In the formula, i d (t) represents the damping injection current, K d For damping gain, x o (t) represents the target oscillation mode component.

[0130] In this embodiment, when the overall reliability of the M1 photovoltaic grid-connected point reaches 0.86 and the reliability of the M4 load branch is 0.61, the system identifies M1 as the main disturbance source and first performs virtual impedance reshaping on the 187Hz target frequency band of the photovoltaic cluster where M1 is located, while the energy storage converter provides local damping injection. If the target oscillation amplitude decreases by more than 60% within 200ms after control, non-cut-off control is maintained and normal operation is gradually restored; if the oscillation amplitude still does not decrease below the safety threshold within 300ms, graded power reduction control is performed on the M1 photovoltaic cluster; if the reliability of the disturbance source continues to increase after power reduction, selective cut-off of the corresponding grid-connected branch is performed.

[0131] Through the above steps, this invention achieves a closed-loop processing flow from "first-arrival location—perturbation verification—reliability fusion—minimum control". Compared with the method of directly judging the disturbance source based on the oscillation amplitude or steady-state energy flow direction, this embodiment can improve the reliability of disturbance source location and reduce unnecessary cut-off control under the conditions of multi-source coupling, impedance time-varying and significant propagation reflection in islanded new energy power grids.

[0132] This invention also provides a system for locating and verifying the first arrival of disturbance sources in an islanded power grid, corresponding to the above method, comprising:

[0133] The multi-point synchronous acquisition module is connected to multiple monitoring points in the isolated new energy power grid. It is used to synchronously acquire voltage and current sampling data of each monitoring point, and to perform time synchronization, anomaly removal and wideband filtering on the sampling data, and output the pre-processed voltage and current complex signals.

[0134] The target oscillation mode separation module is used to perform target oscillation mode separation on the preprocessed voltage and current complex signals, and extract the frequency, amplitude, phase, damping ratio and mode energy envelope of the target oscillation mode;

[0135] The first-reach time series analysis module is used to generate candidate regions of disturbance sources based on the first time the modal energy envelope of each monitoring point exceeds the limit, combined with the topological distance of the monitoring points, line impedance and phase gradient, and to calculate the first-reach location score and modal energy transmission intensity of each candidate node.

[0136] The diagnostic perturbation injection module is used to control the broadband oscillation suppression device to inject a low-amplitude, short-time diagnostic perturbation signal into the grid connection point corresponding to the candidate region of the perturbation source;

[0137] The impedance response verification module is used to calculate the equivalent harmonic impedance change rate, damping sensitivity and response phase offset of each candidate node based on the changes in voltage and current response before and after the injection of the diagnostic perturbation signal.

[0138] The disturbance source credibility determination module is used to normalize the equivalent harmonic impedance change rate, the damping sensitivity, and the response phase offset, and input them together with the first arrival location score and the modal energy transmission intensity into a weighted fusion model to calculate the disturbance source credibility of each candidate node; wherein, the first arrival location score is determined based on the envelope first arrival time series, the topological distance of the monitoring point, the line impedance, and the phase gradient;

[0139] The control strategy generation module is used to output the disturbance source location result based on the credibility of the disturbance source and generate the corresponding minimum control strategy.

[0140] The above description is merely a specific embodiment of the present invention, but the scope of protection of the present invention is not limited thereto. Any variations or substitutions that can be easily conceived by those skilled in the art within the technical scope disclosed in the present invention should be included within the scope of protection of the present invention. Therefore, the scope of protection of the present invention should be determined by the scope of the claims.

Claims

1. A method for first-arrival verification and location of disturbance sources in an isolated power grid, characterized in that, Includes the following steps: Acquire synchronous voltage and current sampling data from multiple monitoring points in an isolated new energy power grid, and perform time synchronization, anomaly removal, and wideband filtering on the sampling data to obtain preprocessed voltage and current complex signals; The preprocessed voltage and current complex signals are subjected to target oscillation mode separation to extract the frequency, amplitude, phase, damping ratio and mode energy envelope of the target oscillation mode; Based on the first time the modal energy envelope of each monitoring point exceeds the limit, combined with the topological distance of the monitoring points, line impedance and phase gradient, a candidate region of disturbance source is generated, and the first arrival location score and modal energy transmission intensity of each candidate node are calculated. The broadband oscillation suppression device is controlled to inject a low-amplitude, short-time diagnostic perturbation signal into the grid connection point corresponding to the candidate region of the disturbance source; Based on the changes in voltage and current response before and after the injection of the diagnostic perturbation signal, the equivalent harmonic impedance change rate, damping sensitivity, and response phase offset of each candidate node are calculated. The equivalent harmonic impedance change rate, the damping sensitivity, and the response phase offset are normalized and then input into a weighted fusion model along with the first-reach location score and the modal energy transmission intensity to calculate the perturbation source confidence of each candidate node. The first-reach location score is determined based on the envelope first-reach timing, monitoring point topology distance, line impedance, and phase gradient. The equivalent harmonic impedance change rate characterizes the electrical response sensitivity of the candidate node to diagnostic perturbation signals. The damping sensitivity characterizes the negative damping contribution of the candidate node, and the response phase offset characterizes the perturbation response phase characteristics of the candidate node. Based on the credibility of the disturbance source, the disturbance source location result is output, and the corresponding minimum control strategy is generated.

2. The method for first-arrival verification and location of disturbance sources in an islanded power grid according to claim 1, characterized in that, The multiple monitoring points include grid connection points of new energy converters, grid connection points of energy storage converters, load branch nodes, and common connection points; the synchronous voltage and current sampling data cover a wide frequency range from 0.1Hz to 2500Hz, with a sampling rate of not less than 12.8kHz.

3. The method for first-arrival verification and location of disturbance sources in an islanded power grid according to claim 1, characterized in that, The target oscillation mode separation includes broadband mode search based on frequency adaptive tracking, sliding window spectral peak locking, and multi-mode decoupling processing; when multiple oscillation modes are detected to exist simultaneously, the target oscillation mode to be located is determined according to the mode energy ratio and damping attenuation trend.

4. The method for first-arrival verification and location of disturbance sources in an isolated power grid according to claim 1, characterized in that, The first time the modal energy envelope exceeds the limit is the starting time when the energy envelope of the target oscillation mode continuously exceeds the background noise reference value by a preset multiple. The first arrival time series matrix is ​​constructed based on the first time difference of different monitoring points, and the oscillation propagation starting point is deduced by combining the grid topology adjacency matrix to form a disturbance source candidate region.

5. The method for first-arrival verification and location of disturbance sources in an isolated power grid according to claim 1, characterized in that, The diagnostic perturbation signal is a biased sinusoidal signal, a narrowband swept frequency signal, or a pseudo-random binary wideband signal located near the target oscillation frequency; the amplitude of the diagnostic perturbation signal is limited to a safe threshold that does not affect the normal power supply of the islanded power grid, and the injection duration is shorter than the preset diagnostic window.

6. The method for first-arrival verification and location of disturbance sources in an isolated power grid according to claim 1, characterized in that, The equivalent harmonic impedance change rate is determined as follows: the perturbation equivalent harmonic impedance is determined based on the ratio of the voltage phasor difference to the current phasor difference at the target frequency before and after the perturbation signal injection; the equivalent harmonic impedance change rate is then determined based on the ratio of the difference between the perturbation equivalent harmonic impedance and the equivalent harmonic impedance before injection to the equivalent harmonic impedance before injection. The damping sensitivity is determined based on the change in the damping ratio of the target oscillation mode before and after perturbation injection.

7. The method for first-arrival verification and location of disturbance sources in an islanded power grid according to claim 1, characterized in that, The credibility of the disturbance source is calculated using a weighted fusion model. The inputs of the weighted fusion model include the first-to-arrive location score, modal energy transmission intensity, equivalent harmonic impedance change rate, damping sensitivity, and response phase offset. The first-to-arrive location score is determined based on the envelope first-to-arrive time difference, phase gradient consistency, topological distance, and line impedance between the candidate node and its neighboring nodes. The modal energy transmission intensity is determined based on the average value of the real part of the conjugate product of the complex voltage signal of the target oscillation mode of the candidate node and the complex current signal of the target oscillation mode along the transmission direction of the node within a preset time window. When the credibility of the disturbance source of a candidate node is greater than a preset threshold, the candidate node is determined to be the location of the disturbance source.

8. The method for first-arrival verification and location of disturbance sources in an islanded power grid according to claim 1, characterized in that, The minimum control strategy includes target frequency band impedance reshaping, local damping injection, output limiting, power reduction operation, or selective cutoff; when impedance reshaping or local damping injection can reduce the oscillation amplitude to a preset threshold, non-cutoff control is preferentially executed; when the credibility of the disturbance source continues to increase and non-cutoff control is ineffective, selective cutoff control is executed.

9. A system for locating and verifying the first arrival of disturbance sources in an isolated power grid, characterized in that, include: The multi-point synchronous acquisition module is connected to multiple monitoring points in the isolated new energy power grid. It is used to synchronously acquire voltage and current sampling data of each monitoring point, and to perform time synchronization, anomaly removal and wideband filtering on the sampling data, and output the pre-processed voltage and current complex signals. The target oscillation mode separation module is used to perform target oscillation mode separation on the preprocessed voltage and current complex signals, and extract the frequency, amplitude, phase, damping ratio and mode energy envelope of the target oscillation mode; The first-reach time series analysis module is used to generate candidate regions of disturbance sources based on the first time the modal energy envelope of each monitoring point exceeds the limit, combined with the topological distance of the monitoring points, line impedance and phase gradient, and to calculate the first-reach location score and modal energy transmission intensity of each candidate node. The diagnostic perturbation injection module is used to control the broadband oscillation suppression device to inject a low-amplitude, short-time diagnostic perturbation signal into the grid connection point corresponding to the candidate region of the perturbation source; The impedance response verification module is used to calculate the equivalent harmonic impedance change rate, damping sensitivity and response phase offset of each candidate node based on the changes in voltage and current response before and after the injection of the diagnostic perturbation signal. The disturbance source credibility determination module is used to normalize the equivalent harmonic impedance change rate, the damping sensitivity, and the response phase offset, and input them together with the first arrival location score and the modal energy transmission intensity into a weighted fusion model to calculate the disturbance source credibility of each candidate node; wherein, the first arrival location score is determined based on the envelope first arrival time series, the topological distance of the monitoring point, the line impedance, and the phase gradient; The control strategy generation module is used to output the disturbance source location result based on the credibility of the disturbance source and generate the corresponding minimum control strategy.

10. The islanded power grid disturbance source first-reach verification and location system according to claim 9, characterized in that, Before injecting the diagnostic perturbation signal, the diagnostic perturbation injection module performs a safety check on the islanded power grid. When the voltage deviation at the point of common coupling is greater than a preset percentage of the rated value, the frequency deviation is greater than a preset value, or the energy storage system is in a power-limited state, the perturbation injection is suspended.

Images