Islanding detection method based on coordinated identification of harmonic dynamic characteristics and reactive slip

Abstract

A method and system for island detection based on the collaborative identification of harmonic dynamic characteristics and reactive slip, relating to the field of new energy island detection. To address the shortcomings of existing island detection, such as low feature extraction accuracy, rigid thresholds, and insufficient response speed, the method collects a microgrid system's operating data set based on its topological structure; implements dynamic feature data fusion processing to calculate harmonic voltages; uses harmonic voltage mutations as the initiation criterion for island detection; sequentially determines whether the harmonic voltage mutation, fundamental voltage amplitude, and each harmonic voltage offset exceed limits, thereby determining whether the microgrid is in an island state; constructs a virtual orthogonal signal to calculate the three-phase reactive slip rate, and determines whether the three-phase reactive slip rate exceeds limits. If not, the microgrid is in a non-island state; if so, the microgrid is in an island state. The present invention is primarily used to determine whether a microgrid is in an island state.

Description

Technical Field

[0001] The present invention relates to the field of new energy islanding detection, and in particular to an islanding detection method and a detection system thereof based on coordinated identification of harmonic dynamic characteristics and reactive slip. Background Art

[0002] Currently, traditional island detection methods (such as over / undervoltage and over / underfrequency methods) suffer from large detection blind spots, insufficient dynamic response, and poor disturbance compatibility in scenarios with high penetration of new energy. Detection methods based on harmonic characteristics and reactive power dynamic characteristics have gradually become a research hotspot. Harmonic components, due to their sensitivity to changes in grid impedance, can reflect the nonlinear characteristics of islanding conditions; while reactive slip (the dynamic rate of change of reactive power) can capture energy imbalances in transient processes. However, existing methods still rely solely on single harmonic or reactive parameters, making it difficult to deal with blind spots under complex operating conditions. Traditional FFT algorithms have spectral leakage in non-steady-state harmonic analysis. Although wavelet transforms have time-frequency localization capabilities, their resolution is insufficient and fixed thresholds cannot adapt to background harmonic fluctuations in the grid, which can easily lead to bottlenecks such as misjudgment or missed detection.

[0003] Currently implemented solutions include passive detection based on harmonic voltage mutation, hybrid harmonic-reactive power detection, and wavelet decomposition (WPD) with dynamic threshold adjustment. The passive detection method based on harmonic voltage mutation monitors the voltage amplitude of specific harmonics (such as the 3rd and 5th harmonics) and determines islanding when it exceeds a threshold. Advantages: No active perturbation is required, making it compatible with harmonically sensitive scenarios. Disadvantages: Relying on load harmonic characteristics, it fails when there is low distortion or background harmonic interference. It does not incorporate reactive power dynamic characteristics and cannot distinguish between islanding and normal load switching. The hybrid harmonic-reactive power detection method combines harmonic distortion and reactive power changes with a fixed threshold dual criterion. Advantages: Reduces blind spots compared to single-parameter methods. Disadvantages: Harmonic feature extraction uses FFT, resulting in insufficient transient resolution; reactive power change rate calculation does not incorporate time-domain tracking, resulting in significant dynamic response delay (>100ms); and the fixed threshold cannot adapt to grid impedance changes. The wavelet decomposition (WPD) with dynamic threshold adjustment method uses WPD to extract multi-band harmonic energy and dynamically adjusts the threshold using a sliding window. Advantages: Improves harmonic feature extraction accuracy. Disadvantages: Does not integrate reactive slip parameters, resulting in a single islanding criterion; the adaptive mechanism targets harmonics only, ignoring the spatiotemporal coupling characteristics of reactive power; and has a high protection action delay (>20ms), which cannot meet the rapid disconnection requirements of high-penetration scenarios.

[0004] Although the above solutions are effective in specific scenarios, none of them achieves spatiotemporal coordinated analysis of harmonic dynamic characteristics and reactive slip, and there are common problems such as low feature extraction accuracy, rigid thresholds, and insufficient response speed.

[0005] Therefore, there is a need for an island detection method and detection system based on the coordinated identification of harmonic dynamic characteristics and reactive slip, which can accurately judge the island state, have high detection accuracy, significantly reduce the detection blind area, and dynamically correct the characteristic threshold. Summary of the Invention

[0006] The present invention aims to solve the defects of existing island detection, such as low accuracy in feature extraction, rigid thresholds, and insufficient response speed. It provides an island detection method and a detection system based on the coordinated identification of harmonic dynamic characteristics and reactive slip, which can accurately judge the island state, have high detection accuracy, significantly reduce the detection blind area, and dynamically correct the feature threshold.

[0007] The present invention provides an island detection method based on coordinated identification of harmonic dynamic characteristics and reactive slip, comprising the following steps:

[0008] S1. Collect the operating data set of the microgrid system according to the topological structure of the microgrid system;

[0009] S2. Implement dynamic feature data fusion processing and calculate harmonic voltage;

[0010] A wavelet decomposition-fast Fourier transform hybrid algorithm is used to extract harmonic spectrum features; the harmonic spectrum features include harmonic components, harmonic voltage mutation amount and reactive slip rate;

[0011] S3, using the harmonic voltage mutation amount as the starting criterion for islanding detection; judging whether the harmonic voltage mutation amount exceeds the limit, if so, proceeding to the next step, if not, the microgrid system is in a non-islanding state;

[0012] S4, voltage diagnosis of overvoltage / undervoltage;

[0013] Determine whether the current fundamental voltage amplitude exceeds the limit. If not, proceed to the next step. If yes, the microgrid is in an island state.

[0014] Determine whether the voltage offset of each harmonic exceeds the limit. If not, proceed to the next step. If yes, the microgrid is in an island state.

[0015] S5, reactive slip time domain tracking;

[0016] A virtual orthogonal signal is constructed to calculate the three-phase reactive slip rate and determine whether the three-phase reactive slip rate exceeds the limit. If not, the microgrid is in a non-island state; if so, the microgrid is in an island state.

[0017] Further: in S1, the operating data set includes wind storage power generation system operating data, photovoltaic storage power generation system operating data, load real-time operating data, microgrid connection point data and bus voltage data; the bus voltage data is collected according to different wiring methods of different microgrid topologies.

[0018] Furthermore: in S2, the steps of extracting the harmonic spectrum features are as follows:

[0019] Decomposing the collected sampling signal using wavelet, dividing the signal into different spectrum bandwidths, wherein the spectrum bandwidths include spectrum bandwidths of fundamental wave, 3rd harmonic, 5th harmonic, 7th harmonic, 9th harmonic, and 11th harmonic;

[0020] Calculate the harmonic voltage mutation of different frequencies at the PCC point before and after the fault;

[0021] The reactive slip rate at the target moment is calculated based on the reactive value at the target moment obtained by sampling.

[0022] Furthermore: in S21, the spectrum bandwidth is obtained as follows:

[0023] S211. Define the sampling harmonic order, calculate the harmonic frequency according to the rated frequency, and obtain the target harmonic frequency;

[0024] S212. Calculate the sub-frequency bandwidth of the wavelet decomposition according to the target harmonic frequency to obtain the sub-frequency bandwidth of different sampling frequencies;

[0025] S213, extracting the sub-frequency bandwidth index of the target harmonic frequency and locating the sub-frequency bandwidth;

[0026] S214 . Sample the harmonic voltage of the corresponding sub-frequency bandwidth according to the sub-frequency bandwidth index, obtain a frequency band signal of the corresponding sub-frequency bandwidth, and calculate the corresponding spectrum bandwidth.

[0027] Furthermore: in S3, the step of calculating the amount of harmonic voltage mutation includes:

[0028] S31. Calculate the critical factor of harmonic voltage variation based on the load, the large power grid, the solar-storage power generation system, and the wind-storage power generation system;

[0029] S32. Calculate the harmonic voltage with increased amplitude according to the critical factor of harmonic voltage change, thereby obtaining the amount of harmonic voltage mutation.

[0030] Furthermore, in S3, the specific steps of determining whether the harmonic voltage mutation exceeds the limit include:

[0031] If the harmonic voltage mutation exceeds the limit, the protection mechanism is triggered; if not, it is determined that the microgrid system is in a non-island state.

[0032] Furthermore: in S4, the specific steps of determining whether the current fundamental voltage amplitude exceeds the limit include:

[0033] When it is detected that the fundamental voltage amplitude exceeds the preset threshold, it is determined that the microgrid system is in an islanded operation state and that the islanding effect has occurred; the switches of the photovoltaic storage power generation system and the wind power storage power generation system are cut off; if the fundamental voltage amplitude does not exceed the preset threshold, proceed to the next step;

[0034] When it is detected that the offset of each harmonic voltage exceeds the preset threshold, it is determined that the microgrid system is in an islanded operation state and that the islanding effect has occurred; the photovoltaic storage power generation system switch and the wind power storage power generation system switch are cut off; if the offset of each harmonic voltage exceeds the preset threshold, execute S5.

[0035] Furthermore, in S5, after the harmonic voltage mutation starts islanding detection, if the fundamental voltage amplitude and the harmonic voltage offsets do not exceed the limit, the reactive slip time domain tracking is started; the reactive slip time domain tracking specifically includes:

[0036] Construct virtual orthogonal signals to calculate the three-phase reactive power respectively; calculate the three-phase reactive slip rate in real time through the differential method;

[0037] Determine whether the reactive slip rate exceeds a limit. If so, determine that the microgrid system is in an island state, and cut off the switches of the photovoltaic storage power generation system and the wind power storage power generation system; if not, determine that the microgrid system is in a non-island state.

[0038] Furthermore: When the main grid fails and is disconnected, if the output power of the photovoltaic grid-connected inverter does not match the power required by the local load, the relationship between the grid active power and the microgrid grid connection point voltage is obtained during grid-connected operation and island constant power operation according to the power flow direction during actual operation of the microgrid system. ;

[0039] In the operating state where there is no power backflow from the microgrid to the large grid, When , the PCC point voltage decreases when the microgrid is islanded;

[0040] When the large power grid allows microgrid to have power reverse transmission, When , the microgrid will experience island operation and the PCC point voltage will decrease; when When , the microgrid will experience island operation and the PCC point voltage will increase.

[0041] The detection system disclosed in the present invention for implementing the islanding detection method based on the coordinated identification of harmonic dynamic characteristics and reactive slip includes a microgrid system, a data acquisition module, a data fusion module and an islanding judgment module.

[0042] The microgrid system includes a wind power generation system with storage, a photovoltaic power generation system with storage, a microgrid load and a large power grid. The intersection of the wind power generation system with storage, the photovoltaic power generation system with storage, the microgrid load and the large power grid is the microgrid connection point.

[0043] The data acquisition module is used to collect the operating data set of the microgrid system according to the topological structure of the microgrid system;

[0044] The data fusion module is used to perform fusion processing on the feature data to obtain a feature parameter space containing harmonic spectrum features;

[0045] The islanding judgment module is used to judge whether the microgrid is in an islanding state based on the harmonic voltage mutation amount, fundamental voltage amplitude, harmonic voltage offset and three-phase reactive slip rate.

[0046] The beneficial effects of the present invention are:

[0047] The present invention provides a microgrid islanding detection method and detection system based on the coordinated identification of harmonic dynamic characteristics and reactive slip. The WPD-FFT hybrid algorithm is used to improve the harmonic transient resolution, construct a harmonic mutation amount-reactive slip rate composite criterion, and introduce a parameter adaptive compensation mechanism. This fundamentally breaks through the technical bottleneck of traditional methods and provides a highly reliable and fast-response islanding detection solution for high-proportion new energy power grids.

[0048] The present invention adopts a method of constructing characteristic parameter space, and the threshold criteria can be flexibly adjusted according to the topology structure, supporting full coverage from industrial and commercial microgrids to remote off-grid systems. No additional hardware disturbance devices are required, effectively avoiding unnecessary startup of the active method, reducing the adverse impact on the system power quality, and achieving high-precision detection through algorithm upgrades, reducing the transformation cost by more than 60%.

[0049] The present invention integrates the spatiotemporal coupling analysis of harmonic dynamic characteristics and reactive slip to construct a composite characteristic parameter space of harmonic components, harmonic mutation and reactive slip rate, thus solving the failure problem of single parameter detection in power balance scenarios. The high-frequency sensitivity of harmonic voltage mutations and the transient tracking capability of reactive slip complement each other, significantly reducing detection blind spots. This makes it suitable for complex working conditions caused by a high proportion of new energy access. A 3ms-level voltage anomaly rapid protection strategy is designed to instantly cut off the grid connection when the voltage deviation exceeds the threshold, which is more than 30 times faster than the traditional 100-millisecond response, minimizing the risk of islanding. At the same time, for scenarios where the harmonics do not exceed the limit but exceed the limit, time-domain tracking of reactive slip is activated, and a delayed confirmation mechanism is used to avoid false operation, achieving the dual safety guarantee of "rapid cutoff + fine analysis". A parameter adaptive compensation mechanism is introduced to dynamically correct the characteristic threshold to overcome interference factors such as background harmonic fluctuations in the power grid and nonlinear changes in the load. Combined with the wavelet decomposition-FFT hybrid algorithm, high-resolution extraction of harmonic components is achieved in the time-frequency domain, and the transient harmonic detection accuracy is improved by more than 40%, ensuring the reliable capture of weak characteristic signals. Based on the initialization of the microgrid wiring method and the dynamic deployment of data collection points, it is adapted to various types of hybrid access scenarios of new energy such as solar storage and wind storage. The rapid protection mechanism reduces the impact of fault current on equipment, extends the life of key components, and avoids the deterioration of power quality, which meets the needs of safe and economical operation of the power grid under the "dual carbon" goals.

[0050] This invention aims to solve the common technical difficulties of island detection in scenarios with a high proportion of new energy access, and achieves the following core goals through technological innovation:

[0051] 1. Improve detection efficiency and response speed:

[0052] Through the spatiotemporal coordinated analysis of harmonic dynamic characteristics and reactive slip, the timing limitations of traditional single-parameter detection are overcome, and the island identification time is shortened from the traditional 100ms level to 3ms, significantly improving detection efficiency. This ensures millisecond-level rapid disconnection from the grid under extreme working conditions and prevents the spread of faults.

[0053] 2. Reduce detection blind spots and misjudgment risks:

[0054] A composite criterion of harmonic mutation and reactive slip rate is constructed, combined with a parameter adaptive compensation mechanism, and the threshold range is dynamically corrected, which reduces the detection blind area by more than 60% and the misjudgment rate to below 0.1%, significantly improving detection reliability and adapting to complex power grid environments.

[0055] 3. Improve system safety and equipment life:

[0056] Through the 3ms-level fast protection mechanism, the grid connection is instantly cut off when the voltage exceeds the limit, reducing the impact of fault current on key equipment such as photovoltaic inverters and energy storage converters, avoiding equipment overload damage, extending the equipment service life by more than 30%, and reducing the risk of electric shock.

[0057] 4. Reduce transformation costs and operation and maintenance complexity:

[0058] Software algorithm upgrades replace hardware disturbance devices, eliminating the need for additional harmonic filters or frequency disturbance equipment, reducing system modification costs by more than 50%. The topology adaptive design supports plug-and-play deployment, reducing on-site commissioning time and subsequent maintenance costs.

[0059] 5. Improve power quality and environmental benefits:

[0060] Avoid the negative impact of active disturbance injection on the harmonic content of the power grid, control the total harmonic distortion (THD) of the grid connection point voltage within 3%, and reduce electromagnetic pollution; at the same time, accurate detection reduces unnecessary power generation system shutdowns, increases the proportion of new energy consumption, and reduces carbon emissions by more than 15% annually.

[0061] 6. Enhance system compatibility and universality:

[0062] Through dynamic adaptation of microgrid topology and fusion of multi-source data, it is compatible with photovoltaic storage, wind storage and hybrid energy systems, supports full coverage from kilowatt-level household microgrids to megawatt-level industrial and commercial systems, provides standardized solutions for new power systems, and promotes the large-scale application of new energy.

[0063] This invention achieves breakthroughs in detection efficiency, safety, economy and environmental protection through the deep integration of multi-parameter collaborative identification, adaptive compensation algorithm and millisecond-level protection mechanism, providing highly reliable and low-cost island detection technology support for new energy high-penetration power grids, and contributing to the safe, green and efficient operation of the power system under the "dual carbon" goals. BRIEF DESCRIPTION OF THE DRAWINGS

[0064] Figure 1 It is a structural diagram of the microgrid system;

[0065] Figure 2 It is an optimized flow chart of an embodiment. DETAILED DESCRIPTION

[0066] The following are only preferred specific embodiments of the present invention, but the scope of protection of the present invention is not limited thereto. Any changes or replacements that can be easily thought of by a person skilled in the art within the technical scope disclosed by the present invention should be included in the scope of protection of the present invention. The embodiments described below are only used to explain the present invention and cannot be interpreted as limiting the present invention. The scope of protection of the present invention should be based on the scope of protection of the claims. The embodiments of the present invention are described in detail below. In order to facilitate the description of the present invention and simplify the description, the technical terms used in the description of the present invention should be interpreted broadly, including but not limited to conventional replacement schemes not mentioned in this application, and also including direct implementation and indirect implementation.

[0067] Example 1

[0068] Combine Figure 1 and Figure 2 This embodiment describes an islanding detection method based on coordinated identification of harmonic dynamic characteristics and reactive slip, including the following steps:

[0069] S1. Collecting an operating data set of the microgrid system according to the topological structure of the photovoltaic microgrid system, and performing precise processing on the operating data set:

[0070] like Figure 1 As shown in the figure, the overall operation data set of the microgrid system is collected. In addition to the operation data of the wind storage power generation system, the operation data of the photovoltaic storage power generation system, and the real-time operation data of the load, the voltage of the microgrid connection point (PCC) is also collected. , current ,frequency At the same time, according to the topology of different microgrid systems, the bus voltages of different wiring modes are collected. The high-frequency wavelet transform sampling method with double localization analysis is used for the 3rd harmonic voltage, 5th harmonic voltage, 7th harmonic voltage, 9th harmonic voltage and 11th harmonic voltage on the PCC side. The sampling frequency is .

[0071] Figure 1 In the example, PCC is the microgrid connection point. is the active power transmitted from the grid to the microgrid, It is the reactive power delivered by the grid to the microgrid, and the power flowing into the PCC is positive; The active power transmitted from the wind power generation system to the PCC is It is the reactive power delivered by the wind-storage power generation system to the PCC, and the power flowing into the PCC is positive; The active power transmitted from the photovoltaic power generation system to the PCC. It is the reactive power delivered by the wind-storage power generation system to the PCC, and the power flowing into the PCC is positive; To absorb the active power of PCC for microgrid load, The reactive power flowing out of PCC is positive because the microgrid load absorbs the reactive power of PCC.

[0072] The precision processing includes:

[0073] When acquiring sampled data, high-frequency sampling with a sampling interval of approximately 0.5ms and multi-point sampling are used to initially filter out noise other than the harmonic frequencies required for island detection. Amplification processing is also performed during the calculation process to improve sampling accuracy.

[0074] S2. Implement dynamic feature data fusion processing;

[0075] The wavelet decomposition-fast Fourier transform (WPD-FFT) hybrid algorithm is used to extract harmonic spectrum features and establish a characteristic parameter space including harmonic components, harmonic mutation amount, and reactive slip rate;

[0076] Wavelet decomposition (WPD) is used on the acquired sampling signal to divide the signal into different spectrum bandwidths of fundamental wave, 3rd harmonic, 5th harmonic, 7th harmonic, 9th harmonic and 11th harmonic through recursive filtering and down sampling.

[0077] Define the sampling harmonic order For rated frequency The frequencies of each harmonic are obtained as:

[0078] (1);

[0079] Where, For the Sub-target harmonic frequency.

[0080] According to the target harmonic frequency Calculate the sub-bandwidth of wavelet decomposition , decomposition level for:

[0081] (2);

[0082] Where, is the permissible frequency deviation, is the sampling frequency.

[0083] This results in different sampling frequencies Sub-bandwidth for:

[0084] (3);

[0085] Extract target harmonic frequencies Subband index And locate the subband, the calculation formula of the subband index is:

[0086] (4);

[0087] According to the calculated subband index , perform corresponding sub-band harmonic voltage sampling, thereby obtaining the frequency band signal of the corresponding sub-band , calculate its spectrum :

[0088] (5);

[0089] Where, is the permissible frequency deviation, is the sampling harmonic order, is the rated frequency.

[0090] Take the peak value in the spectrum , calculate the harmonic voltage amplitude of the corresponding frequency according to the spectrum peak and phase angle :

[0091] (6);

[0092] (7);

[0093] A time-scale-based mutation calculation method is introduced to calculate the voltage mutations of different frequency harmonics at the PCC before and after the fault. In most cases, the voltage harmonics will tend to the size during normal system operation after a period of time. Therefore, delayed mutation detection can avoid misjudgment.

[0094] (8);

[0095] Where, for Subharmonic voltage mutation amount, is the nth sampling moment, where .

[0096] Based on sampling Moment and The reactive value at the moment is calculated Reactive slip rate at any moment :

[0097] (9);

[0098] Where, and Respectively Moment and Reactive power value at the moment.

[0099] S3, Harmonic voltage mutation islanding detection start criterion;

[0100] Traditional overvoltage / undervoltage rapid diagnosis has a certain ability to identify islanding and non-islanding conditions. However, during actual system operation, the traditional harmonic voltage detection method may cause harmonic distortion. To further avoid the problem of inaccurate adjustment of the harmonic voltage distortion dead zone constant caused by harmonic distortion, this embodiment introduces a harmonic voltage mutation criterion as a starting method for islanding detection.

[0101] Assume that the magnitude of the harmonic current output from the grid side before and after the islanding occurs can be ignored. Due to the presence of filter inductors at the wind-solar inverter output, the harmonic current output from the wind-solar inverter cannot change suddenly at the moment of islanding, so the critical factor of harmonic voltage change of the microgrid is introduced. , It can be calculated by the following formula:

[0102] (12);

[0103] Where, 、 Respectively represent the equivalent inductance and equivalent capacitance of the load; 、 Respectively represent the equivalent inductance and equivalent capacitance of the large power grid; 、 They represent the equivalent inductance and equivalent capacitance of the photovoltaic power generation system respectively; 、 They represent the equivalent inductance and equivalent capacitance of the wind-storage power generation system respectively. The harmonic voltage order with increased amplitude can be expressed as follows:

[0104] (13);

[0105] Where, for The angular frequency of the system at the moment can be used to further screen the harmonic voltage amplitude of the required corresponding frequency, and the harmonic voltage mutation amount can be calculated through formula (8). This can reduce the repeated calculation and judgment of the harmonic voltage that has no obvious change after the islanding occurs, and can effectively improve the islanding algorithm detection speed.

[0106] Regarding the threshold of harmonic voltage mutation, it is important to consider avoiding fluctuations caused by normal switching operations. Normal operations such as load switching and capacitor switching in the microgrid system will cause short-term harmonic distortion. In order to avoid false start of microgrid island detection, the harmonic voltage mutation amount calculated in S2 is used. , introduce the threshold safety factor of harmonic voltage mutation (usually 0.03~0.1), the threshold for starting harmonic voltage mutation for system islanding state detection is:

[0107] (14);

[0108] Where, is the harmonic voltage mutation start threshold, unit (V); It is the rated voltage amplitude of the PCC point during normal operation of the microgrid.

[0109] S4, overvoltage / undervoltage quick diagnosis;

[0110] When a fault in the main grid causes an unexpected disconnection, if the output power of the photovoltaic grid-connected inverter does not match the power required by the local load, the voltage and frequency at the PCC will also change accordingly.

[0111] like Figure 1 As shown in the figure, according to the actual power flow direction of the microgrid, it is assumed that the equivalent impedance of the load is for:

[0112] (15);

[0113] Where, is the system angular frequency, The equivalent resistance of the load can be obtained during grid-connected operation and island constant power operation, and the relationship between the grid active exchange value and the PCC point voltage can be obtained, that is, the active power transmitted by the grid to the microgrid. for:

[0114] (16);

[0115] Where, is the PCC point voltage after islanding occurs in the microgrid, The PCC point voltage is when the microgrid is operating normally. If the microgrid does not have power back to the large power grid in normal operation, When the microgrid system is islanded, the PCC voltage decreases. If the large power grid allows the microgrid to have power reverse transmission, when When , the microgrid system will experience island operation and the PCC point voltage will decrease; when When , the microgrid system will experience island operation and the PCC point voltage will increase.

[0116] The fundamental voltage threshold is directly related to equipment safety. The voltage range (±10%) specified in the national standard IEC 60038 is a hard requirement. In order to further improve the safety of equipment operation, the present invention introduces the fundamental voltage amplitude island judgment threshold safety factor. (Usually 0.1~0.15).

[0117] Due to the power reverse transmission in the microgrid, the islanding occurs and the voltage at the grid connection point decreases. That is, when the system is undervoltage, the fundamental voltage amplitude islanding judgment threshold The calculation formula is:

[0118] (17);

[0119] Where, The fundamental voltage amplitude islanding judgment threshold, unit (V).

[0120] When an island occurs and the system is overvoltage, the fundamental voltage amplitude island judgment threshold The calculation formula is:

[0121] (18);

[0122] When the microgrid system detects a voltage deviation that exceeds its normal value range (threshold), it triggers 3ms-level fast protection, cutting off the switches of the photovoltaic storage power generation system and the wind power storage power generation system. It can be considered that the microgrid system is in an island operation state and that the islanding effect has occurred.

[0123] However, when load matching occurs between the switches of the photovoltaic power generation system and the grid-connected inverters of the wind power generation system, and the local load, the changes in system voltage and frequency detected after the main power grid is disconnected often cannot exceed the detection threshold. In this case, the islanding state cannot be detected by the system, so this detection method has a large blind spot.

[0124] If the system has the above-mentioned voltage situation of non-large grid power supply matching load, the microgrid status is judged by the harmonic voltage variables that can be effectively determined by S3 and the 3rd harmonic voltage value, 5th harmonic voltage value, 7th harmonic voltage value, 9th harmonic voltage value, and 11th harmonic voltage value calculated by S2 and screened by S3.

[0125] Harmonics of different frequencies have different characteristics. When calculating the harmonic voltage threshold, it is necessary to consider them in a targeted manner according to the frequency. The third harmonic is mainly caused by the transformer saturation effect. When isolated, the LC resonance will amplify specific harmonics (for example, the third harmonic is easily amplified at the neutral point of the transformer). The threshold must be lower than the resonance danger value (generally 4%). In view of the sensitive characteristics of the third harmonic, the present invention introduces a third harmonic voltage offset threshold safety factor. (usually 0.03~0.04), then the calculation formula for the third harmonic voltage offset threshold is:

[0126] (19);

[0127] Where, is the harmonic voltage offset threshold and , unit (V).

[0128] Compared with the third harmonic, the fifth and above harmonics have lower sensitivity. Therefore, the present invention introduces a safety factor for the voltage offset threshold of the fifth and above harmonics. (usually 0.04~0.08), then the calculation formula for the fifth and above harmonic voltage offset threshold is:

[0129] (20);

[0130] Where, is the harmonic voltage offset threshold and , unit (V).

[0131] If the harmonic voltage exceeds the limit, the 3ms level fast protection is triggered, and the switches of the photovoltaic storage power generation system and the wind power storage power generation system are cut off, it can be considered that the system is in an island operation state and the islanding effect has occurred. Otherwise, execute S5.

[0132] S5, reactive slip time domain tracking;

[0133] If the harmonic characteristics of the load in an islanded state are very similar to those when connected to the grid, the change in harmonic voltage may not be obvious enough, causing the detection method to fail, thus forming a blind spot. For example, if the local load still maintains the same harmonic characteristics when it is islanded as when it is connected to the grid, then the harmonic voltage detection method may not be able to detect the island. This embodiment further introduces a reactive slip time-domain tracking method. After the harmonic mutation quantity initiates islanding detection, if the harmonic voltage load harmonic characteristics are not significantly different from those when it is connected to the grid, and the judgment condition for harmonic voltage exceeding the limit cannot be detected, then reactive slip time-domain tracking is entered.

[0134] During grid-connected operation, the grid, as a stable voltage source, can quickly balance fluctuations in the system's reactive power, resulting in a low rate of change in reactive power. During islanded operation, without the support of the grid, the dynamic reactive power balance between the distributed generation (DGs) and the local load is disrupted, significantly increasing the rate of change in reactive power.

[0135] Construct a virtual orthogonal signal (such as Hilbert transform or all-pass filter) to calculate the three-phase reactive power separately:

[0136] (twenty one);

[0137] Where, is the M-phase reactive power of PCC point at time t; and are the voltage and current of phase M at the PCC point at time t, respectively. T is the fundamental period (e.g., for a 50 Hz system, T = 0.02 s).

[0138] Start reactive slip time domain tracking to determine whether the reactive slip exceeds the limit. If so, cut off the switches of the photovoltaic storage power generation system and the wind power storage power generation system after a delay; if not, return to the non-islanding state.

[0139] Calculate the three-phase reactive slip in real time using the differential method:

[0140] (twenty two);

[0141] Where Δt is the sampling time interval.

[0142] According to the maximum reactive slip rate of the system under normal operating conditions (unit: ), the system islanding state threshold can be set to (unit: ),in is the safety factor (usually 1.5~3).

[0143] If the harmonic voltage exceeds the limit, the long-delay protection is triggered, and the reactive slip rate is calculated and monitored in real time. If the reactive slip is still higher than the threshold after the delay, the switches of the photovoltaic storage power generation system and the wind power storage power generation system are cut off. It can be considered that the system is in an island operation state and the island effect has occurred.

[0144] Example 2

[0145] Combine Figure 1 and Figure 2 This embodiment describes a detection system for implementing the islanding detection method based on the coordinated identification of harmonic dynamic characteristics and reactive slip, including a microgrid system, a data acquisition module, a data fusion module, and an islanding judgment module.

[0146] The microgrid system includes a wind power generation system with energy storage, a photovoltaic power generation system with energy storage, a microgrid load and a large power grid. The intersection of the wind power generation system with energy storage, the photovoltaic power generation system with energy storage, the microgrid load and the large power grid is the microgrid connection point.

[0147] The data acquisition module is used to collect the operating data set of the microgrid system according to the topological structure of the microgrid system;

[0148] The data fusion module is used to perform fusion processing on the feature data to obtain a feature parameter space containing harmonic spectrum features;

[0149] The islanding judgment module is used to judge whether the microgrid is in an islanding state based on the harmonic voltage mutation amount, fundamental voltage amplitude, harmonic voltage offset and three-phase reactive slip rate.

Claims

1. An island detection method based on coordinated identification of harmonic dynamic characteristics and reactive slip, characterized in that: The steps include: S1. Collect the operating data set of the microgrid system according to the topological structure of the microgrid system; S2. Implement dynamic feature data fusion processing and calculate harmonic voltage; A wavelet decomposition-fast Fourier transform hybrid algorithm is used to extract harmonic spectrum features; the harmonic spectrum features include harmonic components, harmonic voltage mutation amount and reactive slip rate; Wavelet decomposition is used to decompose the collected sampled signal, dividing the signal into different spectral bandwidths, including the spectral bandwidths of the fundamental wave, the third harmonic, the fifth harmonic, the seventh harmonic, the ninth harmonic, and the eleventh harmonic. This allows the acquisition of various harmonic components. By integrating the spatiotemporal coupling analysis of harmonic dynamic characteristics and reactive slip, a composite characteristic parameter space of harmonic components, harmonic mutation, and reactive slip rate is constructed. S3, using the harmonic voltage mutation amount as the starting criterion for islanding detection; judging whether the harmonic voltage mutation amount exceeds the limit, if so, proceeding to the next step, if not, the microgrid system is in a non-islanding state; S4, voltage diagnosis of overvoltage / undervoltage; Determine whether the current fundamental voltage amplitude exceeds the limit. If not, proceed to the next step. If yes, the microgrid is in an island state. Determine whether the voltage offset of each harmonic exceeds the limit. If not, proceed to the next step. If yes, the microgrid is in an island state. S5, reactive slip time domain tracking; A virtual orthogonal signal is constructed to calculate the three-phase reactive slip rate and determine whether the three-phase reactive slip rate exceeds the limit. If not, the microgrid is in a non-island state; if so, the microgrid is in an island state.

2. The island detection method based on coordinated identification of harmonic dynamic characteristics and reactive slip according to claim 1 is characterized in that: In S1, the operating data set includes wind-storage power generation system operating data, photovoltaic-storage power generation system operating data, load real-time operating data, microgrid connection point data and bus voltage data; the bus voltage data is collected according to different wiring methods of different microgrid topologies.

3. The island detection method based on coordinated identification of harmonic dynamic characteristics and reactive slip according to claim 1 is characterized in that: In S2, the steps of extracting the harmonic spectrum features are as follows: Calculate the harmonic voltage mutation of different frequencies at the PCC point before and after the fault; The reactive slip rate at the target moment is calculated based on the reactive value at the target moment obtained by sampling.

4. The island detection method based on coordinated identification of harmonic dynamic characteristics and reactive slip according to claim 3 is characterized in that: In S21, the steps of acquiring the spectrum bandwidth are as follows: S211. Define the sampling harmonic order, calculate the harmonic frequency according to the rated frequency, and obtain the target harmonic frequency; S212. Calculate the sub-frequency bandwidth of the wavelet decomposition according to the target harmonic frequency to obtain the sub-frequency bandwidth of different sampling frequencies; S213, extracting the sub-frequency bandwidth index of the target harmonic frequency and locating the sub-frequency bandwidth; S214 . Sample the harmonic voltage of the corresponding sub-frequency bandwidth according to the sub-frequency bandwidth index, obtain a frequency band signal of the corresponding sub-frequency bandwidth, and calculate the corresponding spectrum bandwidth.

5. The island detection method based on coordinated identification of harmonic dynamic characteristics and reactive slip according to claim 1 is characterized in that: In S3, the step of calculating the amount of harmonic voltage mutation includes: S31. Calculate the critical factor of harmonic voltage variation based on the load, the large power grid, the solar-storage power generation system, and the wind-storage power generation system; S32. Calculate the harmonic voltage with increased amplitude according to the critical factor of harmonic voltage change, thereby obtaining the amount of harmonic voltage mutation.

6. The island detection method based on coordinated identification of harmonic dynamic characteristics and reactive slip according to claim 1 is characterized in that: In S3, the specific steps of determining whether the harmonic voltage mutation exceeds the limit include: If the harmonic voltage mutation exceeds the limit, the protection mechanism is triggered; if not, it is determined that the microgrid system is in a non-island state.

7. The island detection method based on coordinated identification of harmonic dynamic characteristics and reactive slip according to claim 1 is characterized in that: In S4, the specific steps of determining whether the current fundamental voltage amplitude exceeds the limit include: When it is detected that the fundamental voltage amplitude exceeds the preset threshold, it is determined that the microgrid system is in an islanded operation state and that the islanding effect has occurred; the switches of the photovoltaic storage power generation system and the wind power storage power generation system are cut off; if the fundamental voltage amplitude does not exceed the preset threshold, proceed to the next step; When it is detected that the offset of each harmonic voltage exceeds the preset threshold, it is determined that the microgrid system is in an islanded operation state and that the islanding effect has occurred; the photovoltaic storage power generation system switch and the wind power storage power generation system switch are cut off; if the offset of each harmonic voltage exceeds the preset threshold, execute S5.

8. The island detection method based on coordinated identification of harmonic dynamic characteristics and reactive slip according to claim 1 is characterized in that: In S5, after the harmonic voltage mutation starts islanding detection, if the fundamental voltage amplitude and each harmonic voltage offset do not exceed the limit, the reactive slip time domain tracking is started. The reactive slip time domain tracking specifically includes: Construct virtual orthogonal signals to calculate the three-phase reactive power respectively; calculate the three-phase reactive slip rate in real time through the differential method; Determine whether the reactive slip rate exceeds a limit. If so, determine that the microgrid system is in an island state, and cut off the switches of the photovoltaic storage power generation system and the wind power storage power generation system; if not, determine that the microgrid system is in a non-island state.

9. The island detection method based on coordinated identification of harmonic dynamic characteristics and reactive slip according to claim 1 is characterized in that: When the main grid fails and is disconnected, if the output power of the photovoltaic grid-connected inverter does not match the power required by the local load, the active power transmitted by the grid to the microgrid is calculated according to the power flow direction during the actual operation of the microgrid system during grid-connected operation and island constant power operation. ; In the operating state where there is no power backflow from the microgrid system to the large power grid, When , the PCC point voltage decreases when the microgrid is islanded; When the large power grid allows the microgrid system to have power reverse transmission, When , the microgrid will experience island operation and the PCC point voltage will decrease; when When , the microgrid will experience island operation and the PCC point voltage will increase.

10. A detection system for implementing an island detection method based on coordinated identification of harmonic dynamic characteristics and reactive slip as described in any one of claims 1 to 9, characterized in that: Including microgrid system, data acquisition module, data fusion module and island judgment module, The microgrid system includes a wind power generation system with storage, a photovoltaic power generation system with storage, a microgrid load and a large power grid. The intersection of the wind power generation system with storage, the photovoltaic power generation system with storage, the microgrid load and the large power grid is the microgrid connection point. The data acquisition module is used to collect the operating data set of the microgrid system according to the topological structure of the microgrid system; The data fusion module is used to perform fusion processing on the feature data to obtain a feature parameter space containing harmonic spectrum features; The islanding judgment module is used to judge whether the microgrid is in an islanding state based on the harmonic voltage mutation amount, fundamental voltage amplitude, harmonic voltage offset and three-phase reactive slip rate.

Images