New energy-based centralized area power analysis and management method and system

Abstract

The invention discloses a new energy-based centralized area electric power analysis management method and system, and particularly relates to the technical field of photovoltaic power station fault analysis. Direct current side current, voltage and environment air pressure data are collected, an arc noise energy judgment threshold value is dynamically adjusted based on air pressure, and a current change rate difference correction coefficient is generated; a composite index of arc noise frequency band energy and voltage signal abrupt change is constructed, and an arc feature vector is generated in combination with waveform oscillation periodicity; comparing the feature vector with a preset template through a dynamic time warping matching algorithm, calculating a matching weight and a trend slope, and judging an arc reignition event in combination with a dynamic judgment threshold and a continuous increasing condition; and generating a priority breaking instruction sequence based on the event duration, controlling the breaking time sequence of the target string according to the grounding grid impedance distribution, and sending a blocking signal to the inverter in the coverage area.

Description

Technical Field

[0001] The present invention relates to the technical field of photovoltaic power station fault analysis, and more specifically, to a centralized regional power analysis and management method and system based on new energy. Background Art

[0002] The scale of centralized photovoltaic power station construction in high-altitude areas is constantly expanding. Such areas are rich in sunlight resources, but due to the special environmental influences such as low pressure and large temperature differences between day and night, the stability of the photovoltaic DC side electrical system faces severe challenges. In existing technologies, the detection and protection methods for DC side arc faults are mostly based on conventional environmental designs, and judgments are made by monitoring the sudden changes in current or voltage. However, in high-altitude and low-pressure environments, the insulation performance of air is significantly reduced, resulting in the dynamic changes in the generation and extinction of arcs. Traditional detection methods are difficult to adapt to such complex working conditions.

[0003] In the existing technology, the accuracy and timeliness of DC side arc fault detection are insufficient due to the lack of full consideration of the impact of high-altitude and low-pressure environments on arc characteristics. In particular, when the arc is temporarily extinguished and then reignited, traditional methods cannot effectively identify the superposition effect of transient multiple arcs, which can easily cause the protection device to malfunction or delay response, affecting not only the power supply continuity of the regional photovoltaic power station, but also potentially expanding the scope of the fault, threatening the overall safe operation of the power grid. Summary of the Invention

[0004] In order to overcome the above-mentioned defects of the prior art, embodiments of the present invention provide a centralized regional power analysis and management method and system based on new energy to solve the problems raised in the above-mentioned background technology.

[0005] To achieve the above object, the present invention provides the following technical solutions: A centralized regional power analysis and management method based on new energy includes the following steps: S1. Collect the current signal, voltage signal and ambient air pressure data of the DC side of the photovoltaic power station, perform high-frequency sampling on the current signal to generate a current waveform sequence, and calculate the difference in current change rate between adjacent sampling windows; S2. Adjust the arc noise energy judgment threshold based on the ambient air pressure data, and generate a dynamic judgment threshold based on the current change rate difference; S3, extracting the composite index of arc noise frequency band energy and voltage signal mutation in the current waveform sequence, and generating arc feature vectors based on the waveform oscillation periodicity; S4. Dynamically match the arc feature vector with a preset arc feature template, and calculate the matching weight and weight change trend; S5. When the weight change trend corresponding to the maximum matching weight exceeds the dynamic determination threshold and increases continuously, an arc restrike event is determined; S6. Generate a priority instruction sequence for string disconnection according to the duration of the arc restrike event; S7. Based on the impedance distribution of the grounding grid, the DC circuit breakers of the target strings are controlled to disconnect according to the priority according to the priority instruction sequence, and synchronous blocking instructions are sent to the inverters in the coverage area.

[0006] In a preferred embodiment, S1 includes: When collecting the current signal of the DC side of the photovoltaic power station, the voltage signal and ambient air pressure data are collected synchronously, and the collection timestamp of each signal is recorded; Sampling the current signal at a preset high-frequency sampling frequency to generate a current waveform sequence comprising a plurality of continuous sampling points; The current waveform sequence is divided into adjacent sampling windows according to a preset window length, and the sum of squares of the current mean differences of adjacent sampling windows is calculated as the current change rate difference.

[0007] In a preferred embodiment, S2 includes: According to the difference between the ambient air pressure data and the preset reference air pressure, the arc noise energy judgment threshold is adjusted in stages; generating a dynamic correction factor according to a ratio of the current change rate difference to a preset difference threshold, wherein the dynamic correction factor is a product of the ratio of the current change rate difference to the preset difference threshold and a preset gain coefficient; The arc noise energy judgment threshold value adjusted in stages is added to the dynamic correction factor to generate a dynamic judgment threshold value.

[0008] In a preferred embodiment, adjusting the arc noise energy determination threshold in stages includes: When the difference between the ambient air pressure data and the preset reference air pressure is greater than the first critical value, the arc noise energy judgment threshold is linearly reduced according to the difference ratio; when the difference between the ambient air pressure data and the preset reference air pressure is less than or equal to the first critical value, the arc noise energy judgment threshold is kept unchanged.

[0009] In a preferred embodiment, S3 includes: Perform frequency band filtering on the current waveform sequence according to the preset arc noise frequency band range, and calculate the square integral of the energy of the filtered signal in the time window as the arc noise frequency band energy; Detect the sudden change gradient of the voltage signal within the time window. When the sudden change gradient exceeds the preset gradient threshold, record the gradient amplitude as the voltage signal sudden change indicator. Multiplying the arc noise frequency band energy and the voltage signal mutation index to generate a composite index; The number of zero crossings of the waveform oscillation period of the current waveform sequence within the time window is counted, and the arc feature vector is generated by combining the composite index.

[0010] In a preferred embodiment, S4 includes: Perform dynamic time warping alignment on the time series of arc feature vector and preset arc feature template to eliminate time axis distortion error; Calculate the point-by-point Euclidean distance of the aligned time series and normalize the Euclidean distance to generate the initial matching weight; According to the distribution of the initial matching weights in the time window, a linear regression model is fitted and the slope of the weight change trend is output; The initial matching weight is modified based on the ratio of the weight change trend slope to the preset slope threshold to generate the matching weight.

[0011] In a preferred embodiment, S5 includes: Traverse all matching weights and select the weight change trend corresponding to the maximum matching weight; Compare the weight change trend with the dynamic judgment threshold. When the weight change trend exceeds the dynamic judgment threshold, enter the incremental condition judgment; Check whether the weight change trend remains monotonically increasing within a preset number of consecutive time windows; If the weight change trend exceeds the dynamic judgment threshold and increases monotonically within a preset number of consecutive time windows, it is determined to be an arc restrike event.

[0012] In a preferred embodiment, S6 includes: Determine a preset duration interval based on the duration of the arc restrike event, and map the duration to a corresponding priority level; Count the duration of multiple arc restrike events within the same time window, and sort them from longest to shortest by duration to generate a string disconnection sequence list; Based on the string disconnection sequence list, a disconnection delay time is assigned to each string, and a priority instruction sequence including a string identifier and a disconnection delay time is generated.

[0013] In a preferred embodiment, S7 includes: Determine the boundaries of the blocking area based on the grounding grid impedance distribution data, and determine the coverage area based on the boundaries of the blocking area; According to the disconnection delay time in the priority instruction sequence, disconnection instructions are sent to the DC circuit breakers of the target strings in sequence, and the disconnection operations are controlled to be executed according to the priority. A blocking instruction is generated synchronously, which includes the identifier of the inverter within the coverage area and the blocking time window, and the blocking instruction is sent to the inverter through the communication protocol.

[0014] In another aspect, the present invention provides a centralized regional power analysis and management system based on new energy, comprising: Signal acquisition and processing module: collects the current signal, voltage signal and ambient air pressure data of the DC side of the photovoltaic power station, performs high-frequency sampling on the current signal to generate a current waveform sequence, and calculates the difference in current change rate between adjacent sampling windows; Threshold dynamic generation module: adjusts the arc noise energy judgment threshold based on ambient air pressure data, and generates a dynamic judgment threshold based on the current change rate difference; Feature extraction building block: extracts the composite index of arc noise frequency band energy and voltage signal mutation in the current waveform sequence, and generates arc feature vectors based on the waveform oscillation periodicity; Dynamic matching analysis module: dynamically matches the arc feature vector with the preset arc feature template, and calculates the matching weight and weight change trend; Event determination trigger module: When the weight change trend corresponding to the maximum matching weight exceeds the dynamic determination threshold and increases continuously, an arc restrike event is determined; Strategy generation and sequencing module: Generates a priority instruction sequence for string disconnection based on the duration of the arc restrike event; Execution control synchronization module: Based on the grounding grid impedance distribution, it controls the DC circuit breakers of the target strings to disconnect according to the priority instruction sequence, and sends synchronization blocking instructions to the inverters in the coverage area.

[0015] Compared with the prior art, the present invention has the following beneficial effects: 1. This invention significantly improves the detection accuracy and response efficiency of arc faults on the DC side of high-altitude photovoltaic power stations through the fusion analysis of multi-dimensional environmental parameters and electrical signals. It dynamically corrects the judgment threshold based on air pressure data and constructs composite features based on the difference in the current waveform change rate, effectively solving the misjudgment problem caused by the single environmental parameter in the existing technology. Through dynamic matching of arc feature vectors with preset templates and weight trend analysis, it can accurately capture the transient superposition effect during the arc restrike process, avoiding the risk of missed detection of multiple consecutive arc events by traditional mutation detection methods, thereby achieving reliable identification of fault events under complex working conditions. 2. Dynamically generate a string disconnection instruction sequence based on the duration of the arc restrike event, and intelligently delineate the lockout area based on the impedance distribution of the grounding grid, achieving accurate positioning and hierarchical isolation of the faulty strings. Through the timing coordination of the inverter lockout instruction and the circuit breaker disconnection operation, it not only ensures the rapid removal of high-priority faults, but also avoids the impact of simultaneous disconnection of multiple strings on the power grid, thereby maintaining power supply continuity while significantly improving the overall stability and safety margin of the regional power system. BRIEF DESCRIPTION OF THE DRAWINGS

[0016] Figure 1 This is a flow chart of a centralized regional power analysis and management method based on new energy in the present invention; Figure 2 This is a structural diagram of a centralized regional power analysis and management system based on new energy in the present invention. DETAILED DESCRIPTION

[0017] The following will provide a clear and complete description of the technical solutions in the embodiments of the present invention in conjunction with the accompanying drawings. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. All other embodiments obtained by ordinary technicians in this field based on the embodiments of the present invention without making any creative efforts shall fall within the scope of protection of the present invention.

[0018] Example 1: Figure 1 The present invention provides a centralized regional power analysis and management method based on new energy, which includes the following steps: S1. Collect the current signal, voltage signal and ambient air pressure data of the DC side of the photovoltaic power station, perform high-frequency sampling on the current signal to generate a current waveform sequence, and calculate the difference in current change rate between adjacent sampling windows; S2. Adjust the arc noise energy judgment threshold based on the ambient air pressure data, and generate a dynamic judgment threshold based on the current change rate difference; S3, extracting the composite index of arc noise frequency band energy and voltage signal mutation in the current waveform sequence, and generating arc feature vectors based on the waveform oscillation periodicity; S4. Dynamically match the arc feature vector with a preset arc feature template, and calculate the matching weight and weight change trend; S5. When the weight change trend corresponding to the maximum matching weight exceeds the dynamic determination threshold and increases continuously, an arc restrike event is determined; S6. Generate a priority instruction sequence for string disconnection according to the duration of the arc restrike event; S7. Based on the impedance distribution of the grounding grid, the DC circuit breakers of the target strings are controlled to disconnect according to the priority according to the priority instruction sequence, and synchronous blocking instructions are sent to the inverters in the coverage area.

[0019] S1. Collect the current signal, voltage signal, and ambient air pressure data from the DC side of the photovoltaic power station, perform high-frequency sampling on the current signal to generate a current waveform sequence, and calculate the difference in current change rate between adjacent sampling windows, including: The system collects current and voltage signals from the DC side of the photovoltaic power station, as well as ambient air pressure data. Specifically, the DC side current signal is collected using a Hall effect current sensor, the DC side voltage signal is collected using a voltage sensor, and the ambient air pressure data is collected using a pressure sensor. During the collection process, the signal acquisition timestamps of the current, voltage, and pressure sensors are synchronized and recorded using the same clock source to ensure time alignment of the signals.

[0020] When high-frequency sampling of the current signal is performed, the preset high-frequency sampling frequency is set based on the frequency range of arc noise on the DC side of the photovoltaic power station. For example, when the arc noise frequency range covers 10kHz to 150kHz, the preset high-frequency sampling frequency is set to no less than 200kHz to ensure that the high-frequency components of the arc noise are fully captured. The current signal after high-frequency sampling generates a current waveform sequence consisting of multiple consecutive sampling points, which is then stored in a buffer.

[0021] When dividing the current waveform sequence into adjacent sampling windows according to a preset window length, the preset window length is set based on the response time of the PV power plant's DC-side protection system. For example, if the preset window length is set to 10ms and the interval between adjacent sampling windows is 5ms, this creates a partially overlapping window division. To calculate the sum of the squares of the current mean differences between adjacent sampling windows as the current rate of change difference, the current average value of all sampling points within each sampling window is first calculated. The difference between the current average values of two adjacent sampling windows is then calculated. Finally, the square of this difference is used as the current rate of change difference for that adjacent window.

[0022] In the above steps, the sum of squares of the current mean differences is used to quantify the degree of abrupt changes in the current waveform within adjacent windows. This sum-of-squares calculation can amplify abrupt changes and improve sensitivity to transient current fluctuations. For example, when the current mean difference between adjacent windows is 0.5A, the sum-of-squares is 0.25A². When the difference increases to 1A, the sum-of-squares increases accordingly to 1A², significantly differentiating between different abrupt changes.

[0023] S2. Adjust the arc noise energy judgment threshold based on the ambient air pressure data, and generate a dynamic judgment threshold based on the current change rate difference, including: During specific implementation, the preset reference air pressure is the standard atmospheric pressure value, that is, 101.325kPa, which is the internationally accepted average air pressure value at sea level and is used to compare the actual air pressure data in high altitude areas. The first critical value is set to 10% of the reference air pressure based on the experimental data of the impact of air pressure changes on arc characteristics in high altitude areas. For example, when the ambient air pressure in high altitude areas is 80% of the reference air pressure, the difference is 20%, which exceeds the first critical value of 10%. At this time, the judgment threshold is linearly reduced according to the difference ratio. The proportional coefficient of the linear reduction is determined by experimental calibration, specifically, the judgment threshold is reduced by 0.5% for every 1% difference. For example, when the difference is 20%, the judgment threshold is reduced by 10%. If the difference is 8% (does not exceed the first critical value), the judgment threshold remains at the initial value.

[0024] The preset difference threshold is determined by statistically analyzing the current rate of change differences during normal operation of the PV power plant's DC side. For example, if 30 consecutive days of operating data are collected, the 95th percentile of the current rate of change differences is calculated as the preset difference threshold. If the current rate of change difference is 0.15A² and the preset difference threshold is 0.1A², the ratio is 1.5.

[0025] The preset gain factor is set based on the arc detection sensitivity requirement. Specifically, by adjusting the factor, the dynamic correction factor can be adjusted within the range of 0.5 to 2.0. For example, when the gain factor is set to 1.2, the dynamic correction factor is 1.5×1.2=1.8.

[0026] When adding the staged adjustment threshold to the dynamic correction factor to generate the dynamic threshold, for example, if the adjusted threshold is 90% of the initial value (reduced by 10% for a 20% difference) and the dynamic correction factor is 1.8, the dynamic threshold is 90 + 1.8 = 91.8 (assuming the initial threshold is 100 units). If the adjusted threshold is 100 units (the difference does not exceed the critical value) and the dynamic correction factor is 0.96 (the current change rate difference is 0.08A², the ratio is 0.8 × the gain is 1.2), the dynamic threshold is 100 + 0.96 = 100.96.

[0027] In the above steps, the preset baseline air pressure is selected according to international standards. The first critical value is determined by experimentally verifying the impact of sudden changes in high-altitude air pressure on arc characteristics. The linear reduction coefficient is calibrated through arc simulation experiments. For example, the trend of arc noise energy threshold changes under different air pressure differences is statistically analyzed and a linear relationship is fitted. The statistical method for the preset difference threshold is the industry-standard quantile method. The gain coefficient is set by adjusting the detection sensitivity requirements, for example, by testing the effect of different gain values on the false positive rate of arc detection in a laboratory environment.

[0028] S3. Extract the composite index of arc noise frequency band energy and voltage signal mutation in the current waveform sequence, and generate arc feature vectors based on the waveform oscillation periodicity, including: In specific implementations, the preset arc noise frequency range is set to 10kHz to 150kHz based on experimental data from DC-side arc discharges in photovoltaic power plants. A digital bandpass filter is used to band-filter the current waveform sequence. The squared integral of the energy of the filtered signal within the time window is calculated by calculating the square of the amplitude at each sampling point and then accumulating it. For example, with a 10ms time window, the filtered current signal has 2000 sampling points within the window (sampling frequency 200kHz). The squared current values at each sampling point are summed to obtain the arc noise frequency energy.

[0029] When detecting a sudden change in the voltage signal gradient within a time window, the absolute value of the difference between adjacent sampling points of the voltage signal is calculated. A valid sudden change is determined when the absolute value of three consecutive differences exceeds the preset gradient threshold. The preset gradient threshold is set to 5V / ms based on the voltage fluctuation range during normal operation of the PV power station's DC side. For example, if the voltage signal drops from 600V to 590V within 1ms, the sudden change gradient is 10V / ms, exceeding the preset gradient threshold. This gradient amplitude of 10V / ms is recorded as a voltage signal sudden change indicator. If the voltage sudden change gradient is 4V / ms, which does not exceed the threshold, the sudden change indicator is not recorded.

[0030] When multiplying the arc noise band energy by the voltage signal mutation index to generate a composite index, if the arc noise band energy is 500 units and the voltage signal mutation index is 10V / ms, the composite index is 500 × 10 = 5000. If the voltage signal mutation index is not triggered (i.e., does not exceed the threshold), the composite index only retains the arc noise band energy value.

[0031] When counting the number of zero crossings during the waveform oscillation cycle of a current waveform sequence within a time window, the zero crossing count is obtained by detecting the number of zero crossings from positive to negative or negative to positive. For example, if 50 zero crossings are detected within a 10ms time window, the corresponding frequency of the oscillation cycle is 50 / (0.01×2) = 2500Hz (two zero crossings per cycle). When combining composite indicators to generate arc feature vectors, the composite indicator and the number of zero crossings are used as two dimensions to construct a two-dimensional vector. For example, if the composite indicator is 5000 and the number of zero crossings is 50, the arc feature vector is represented as (5000, 50).

[0032] In the above steps, the preset arc noise frequency range was determined based on actual arc discharge experimental data to ensure coverage of the main arc noise energy distribution. The digital bandpass filter design utilizes a finite impulse response (FIR) filter with cutoff frequencies set at 10kHz and 150kHz, a stopband attenuation greater than 40dB, and a passband ripple less than 1dB. The squared energy integral is used to quantify the arc noise energy intensity and eliminate the influence of low-frequency interference signals. The voltage signal gradient is detected using a sliding window difference method with a window length of three sampling points to avoid noise interference at a single sampling point.

[0033] The composite index is generated by multiplying the noise energy and voltage mutation characteristics, amplifying abnormal signals that occur simultaneously. For example, when only high noise energy is present but no voltage mutation occurs (possibly due to environmental interference), the composite index is low. However, when both noise energy and voltage mutation occur simultaneously (possibly a true arc), the composite index increases significantly. Zero-crossing statistics are used to reflect the oscillation frequency of the current waveform. High-frequency oscillations are often associated with arc characteristics.

[0034] S4. Dynamically match the arc feature vector with the preset arc feature template and calculate the matching weight and weight change trend, including: In specific implementation, the preset arc feature template is constructed by collecting the feature vectors of typical arc events on the DC side of a photovoltaic power station. For example, arc discharges under different air pressure conditions are simulated in a laboratory environment, and the corresponding arc feature vectors are extracted as templates. During the dynamic time warping alignment process, a dynamic time warping algorithm is used to nonlinearly align the time series of the arc feature vector and the template, allowing the time axis to be locally stretched or compressed to eliminate waveform distortion caused by differences in arc duration. For example, when the time series length of the arc feature vector is 50 points and the template sequence is 45 points, dynamic time warping inserts duplicate points or merges adjacent points to make the lengths of the two consistent, and calculates the minimum cumulative distance path.

[0035] To calculate the point-by-point Euclidean distance of the aligned time series, the square root of the difference between the eigenvector components of each aligned point is taken. For example, if the arc eigenvector component is (5000, 50) and the template component is (4800, 55), the Euclidean distance is the square root of the sum of the squared differences between the two components, which is approximately 200.06. After summing the distances of all points, the total distance is mapped to the range of 0 to 1 using the min-max normalization method to generate the initial matching weight. For example, if the minimum distance is 100, the maximum distance is 500, and the current total distance is 300, the initial matching weight is (300-100) / (500-100) = 0.5.

[0036] When fitting a linear regression model based on the distribution of initial matching weights within a time window, the sampling point numbers within the time window are used as the independent variable, the initial matching weights are used as the dependent variable, and the slope of the regression line is calculated as the slope of the weight change trend. For example, if the time window contains 10 sampling points and the initial matching weight sequence is [0.3, 0.4, 0.5, 0.6, 0.7, 0.8, 0.9, 1.0, 1.1, 1.2], the calculated regression slope is 0.1, indicating that the weights increase linearly over time. The preset slope threshold is set to 0.08 based on the typical amplitude of weight changes in historical arc events.

[0037] When correcting the initial matching weight based on the ratio of the weight change trend slope to the preset slope threshold, if the slope value is 0.1 and the threshold is 0.08, the ratio is 1.25. The correction factor is set as the product of the ratio and the preset gain factor, which is set to 1.2 based on the detection sensitivity requirements. For example, if the correction factor is 1.25 × 1.2 = 1.5, the initial matching weight of 0.5 is corrected to 0.5 × 1.5 = 0.75, generating the matching weight. If the slope value is 0.06 (less than the threshold of 0.08), the correction factor is 0.6 × 1.2 = 0.72, and the initial weight of 0.5 is corrected to 0.5 × 0.72 = 0.36.

[0038] In the above steps, the dynamic time warping algorithm is implemented using open-source libraries (such as Python's fastdtw library). Its core logic involves dynamic programming to solve for the optimal alignment path. Euclidean distance calculation and normalization are standard mathematical operations. The linear regression model is fitted using the least squares method, where the slope equals the covariance divided by the variance. However, this implementation does not require explicit calls to a mathematical library; existing linear regression tools can be used directly. The preset slope threshold is determined based on a statistical analysis of historical arc event data. For example, the weighted slope of 100 arc events is calculated and the 95th percentile is used as the threshold. The gain parameter of the correction coefficient is determined through laboratory testing. For example, the gain value is adjusted within a test set to select the coefficient that optimally balances the detection rate and false alarm rate.

[0039] S5. When the weight change trend corresponding to the maximum matching weight exceeds the dynamic determination threshold and increases continuously, an arc restrike event is determined, including: In specific implementation, all matching weights are traversed, and the weight change trend corresponding to the maximum matching weight is selected. For example, if three template matches generate matching weights of 0.8, 0.9, and 1.2, the weight change trend corresponding to the maximum value of 1.2 is selected. When comparing the weight change trend with the dynamic judgment threshold, the dynamic judgment threshold is set to 0.1 based on historical arc event data. For example, if the weight change trend is 0.15, exceeding the dynamic judgment threshold of 0.1, the incremental condition judgment is entered.

[0040] The weight change trend is checked to see if it remains monotonically increasing within a preset number of consecutive time windows. The preset number of time windows is set to three windows based on the minimum duration of an arc restrike event. For example, if the time window length is 10ms, three consecutive windows correspond to a 30ms continuous monitoring period. If the weight change trend values in three consecutive time windows are 0.12, 0.15, and 0.18, respectively, it is determined to be monotonically increasing. If the trend value in one window is 0.15, 0.13, and 0.17, the continuous increase condition is not met due to the decreasing trend value in the intermediate windows.

[0041] To determine an arc restrike event, the weight trend must exceed the dynamic determination threshold and be continuously increasing. For example, if the weight trend is 0.15 (exceeding the threshold of 0.1) and is 0.10, 0.15, and 0.20 (monotonically increasing) in three consecutive windows, it is determined to be an arc restrike event. If the weight trend is 0.09 (not exceeding the threshold) or if the weight trend is 0.15 but the window sequence is 0.15, 0.12, and 0.18 (non-continuously increasing), no determination is triggered.

[0042] In the above steps, the dynamic judgment threshold is set based on a statistical analysis of weight change trends in historical arc events. For example, the trend value distribution of 100 real arc events is calculated and the 95th percentile is used as the threshold. The preset number of time windows is set based on the minimum duration of an arc restrike event. For example, laboratory testing shows that arc restrike requires at least 30ms of continuous energy accumulation, so three 10ms windows are set. Monotonic increase detection is achieved by comparing the trend values of adjacent windows. If the trend value of the subsequent window is strictly greater than that of the previous window, it is judged as increasing.

[0043] S6. Generate a priority instruction sequence for string disconnection based on the duration of the arc restrike event, including: In specific implementation, the preset duration range is set into three intervals based on the response time of the PV power plant's DC-side protection system and the arc energy accumulation characteristics: short duration (0-30ms), medium duration (30-60ms), and long duration (60ms and above). For example, if the arc restrike event lasts for 45ms, it is mapped to the medium duration interval, corresponding to the medium priority level. If it lasts for 75ms, it is mapped to the long duration interval, corresponding to the high priority level.

[0044] When counting the duration of multiple arc restrike events within the same time window, the string identifier and its corresponding duration are recorded for each event. For example, if three arc restrike events are detected within the same 10ms time window, with durations of 25ms, 50ms, and 70ms, respectively, the corresponding string identifiers are A01, B02, and C03. When generating a string interruption order list by sorting duration from longest to shortest, the resulting order is C03 (70ms), B02 (50ms), and A01 (25ms).

[0045] When assigning a disconnection delay time to each string based on the string disconnection order list, the disconnection delay time is set based on the string priority level and grid stability requirements. For example, high-priority string C03 is assigned a disconnection delay of 0ms (immediate disconnection), medium-priority string B02 is assigned a delay of 50ms, and low-priority string A01 is assigned a delay of 100ms. When generating a priority command sequence, the command sequence includes the mapping between string identifiers and disconnection delay times, for example, C03: 0ms, B02: 50ms, A01: 100ms.

[0046] In the above steps, the preset duration intervals are based on experimental data on arc energy accumulation and equipment tolerance. For example, laboratory tests have shown that arcs lasting longer than 60ms can cause irreversible damage to fuses, so they are assigned a high priority. The rules for allocating disconnect delays are determined through grid transient stability simulations. For example, adjacent strings should be disconnected at least 50ms apart to prevent voltage sags from accumulating. The string identifier is a unique predefined code in the PV power plant monitoring system that is used to locate the physical location of the target string.

[0047] S7. Based on the grounding grid impedance distribution, the DC circuit breakers of the target strings are controlled to disconnect according to the priority instruction sequence, and synchronous blocking instructions are sent to the inverters in the coverage area, including: In practice, ground grid impedance distribution data is obtained through impedance testing of the PV power plant's grounding system. This testing method involves injecting a current signal of a specific frequency into the ground grid, measuring the phase difference and amplitude ratio between the voltage and current at each node, and calculating the impedance value for each area of the ground grid. For example, if 10 measurement points are selected in the ground grid, the measured impedance values range from 0.5Ω to 5Ω. Areas with impedance values less than 2Ω are defined as low-impedance areas, while areas with impedance values greater than or equal to 2Ω are defined as high-impedance areas.

[0048] When defining the blocking zone boundary based on grounding grid impedance distribution data, the blocking zone boundary is defined as the boundary between the low-impedance zone and the high-impedance zone. For example, if the impedance value of one zone is 1.8Ω (low-impedance zone) and the impedance value of the adjacent zone is 2.3Ω (high-impedance zone), the blocking zone boundary is defined as the connecting line between the two zones. The coverage area is determined by the physical location of the low-impedance zone. For example, if the low-impedance zone includes strings A01, B02, and C03, the coverage area is the electrical circuit containing these three strings.

[0049] When disconnect commands are sent to the DC circuit breakers of the target strings in sequence according to the disconnect delay times in the priority command sequence, the disconnect delay times are set in a stepped interval based on the priority level. For example, the disconnect delay time for high-priority string C03 is 0ms (immediate execution), the delay time for medium-priority string B02 is 50ms, and the delay time for low-priority string A01 is 100ms. After receiving the disconnect commands, the DC circuit breaker executes the disconnect operations in sequence according to the delay times, for example, disconnecting C03 at 0ms, B02 after 50ms, and A01 after 100ms.

[0050] When a blocking instruction is generated synchronously, it includes the identifiers of the inverters within the coverage area and the blocking time window. The inverter identifiers are determined by the configuration of the PV power plant monitoring system. For example, the inverter identifiers within the coverage area are INV-01, INV-02, and INV-03. The blocking time window is set based on the duration of the disconnection operation. For example, if the total disconnection duration is 100ms, the blocking time window is set to 50ms from the start to the end of the disconnection (i.e., 0ms to 150ms). When sending the blocking instruction to the inverter via the communication protocol, the Modbus TCP protocol is used, encapsulating the blocking instruction into a data frame with the function code "Write Multiple Registers". The target register address corresponds to the inverter's blocking control bit.

[0051] In the above steps, the grounding grid impedance test method is a frequency response analysis method known in the field of power systems. The test frequency is selected as 1kHz to avoid power frequency interference. The boundary of the blocking area is divided based on the impact of the impedance value on the fault current diffusion path. The low impedance area is more likely to conduct fault current and needs to be blocked first. The step-by-step interval of the disconnection delay time is determined by the transient stability simulation of the power grid. For example, the simulation results show that a 50ms disconnection interval between adjacent strings can avoid a voltage drop of more than 10%. The setting of the blocking time window takes into account the dissipation time of the residual energy of the arc after the disconnection operation.

[0052] Example 2: Figure 2 The present invention provides a structural diagram of a centralized regional power analysis and management system based on new energy, comprising: Signal acquisition and processing module: collects the current signal, voltage signal and ambient air pressure data of the DC side of the photovoltaic power station, performs high-frequency sampling on the current signal to generate a current waveform sequence, and calculates the difference in current change rate between adjacent sampling windows; Threshold dynamic generation module: adjusts the arc noise energy judgment threshold based on ambient air pressure data, and generates a dynamic judgment threshold based on the current change rate difference; Feature extraction building block: extracts the composite index of arc noise frequency band energy and voltage signal mutation in the current waveform sequence, and generates arc feature vectors based on the waveform oscillation periodicity; Dynamic matching analysis module: dynamically matches the arc feature vector with the preset arc feature template, and calculates the matching weight and weight change trend; Event determination trigger module: When the weight change trend corresponding to the maximum matching weight exceeds the dynamic determination threshold and increases continuously, an arc restrike event is determined; Strategy generation and sequencing module: Generates a priority instruction sequence for string disconnection based on the duration of the arc restrike event; Execution control synchronization module: Based on the grounding grid impedance distribution, it controls the DC circuit breakers of the target strings to disconnect according to the priority instruction sequence, and sends synchronization blocking instructions to the inverters in the coverage area.

[0053] The above formulas are all dimensionless and numerical calculations. The formula is a formula that is closest to the actual situation obtained by collecting a large amount of data and performing software simulation. The preset parameters and thresholds in the formula are set by technicians in this field according to actual conditions.

[0054] It should be noted that the present invention can be deployed on the device itself to implement embedded applications, and can also be run on a PC or other terminal with a user interface, thereby meeting various hardware environments and usage requirements.

[0055] The above embodiments can be implemented in whole or in part via software, hardware, firmware, or any other combination. When implemented using software, the above embodiments can be implemented in whole or in part in the form of a computer program product. The computer program product comprises one or more computer instructions or computer programs. When loaded or executed on a computer, the processes or functions described in the embodiments of this application are fully or partially performed. The computer can be a general-purpose computer, a special-purpose computer, a computer network, or other programmable device. The computer instructions can be stored in a computer-readable storage medium or transmitted from one computer-readable storage medium to another. For example, the computer instructions can be transmitted from one website, computer, server, or data center to another website, computer, server, or data center via wired means (e.g., infrared, wireless, microwave, etc.). The computer-readable storage medium can be any available medium accessible by a computer or a data storage device such as a server or data center that contains a collection of one or more available media. The available medium can be magnetic media (e.g., floppy disks, hard disks, tapes), optical media (e.g., DVDs), or semiconductor media. The semiconductor media can be a solid-state drive.

[0056] Those skilled in the art will clearly understand that, for the convenience and brevity of description, the specific working processes of the systems, devices and modules described above can refer to the corresponding processes in the aforementioned method embodiments and will not be repeated here.

[0057] In the several embodiments provided in this application, it should be understood that the disclosed systems, devices and methods can be implemented in other ways. For example, the device embodiments described above are merely schematic. For example, the division of the modules is only a logical function division. In actual implementation, there may be other division methods, such as multiple modules or components can be combined or integrated into another system, or some features can be ignored or not executed. Another point is that the mutual coupling or direct coupling or communication connection shown or discussed can be through some interfaces, indirect coupling or communication connection of devices or modules, which can be electrical, mechanical or other forms.

[0058] The modules described as separate components may or may not be physically separate, and the components shown as modules may or may not be physical modules, and may be located in one place or distributed across multiple network modules. Some or all of the modules may be selected to achieve the purpose of this embodiment according to actual needs.

[0059] In addition, each functional module in each embodiment of the present application may be integrated into one processing module, or each module may exist physically separately, or two or more modules may be integrated into one module.

[0060] If the functions are implemented in the form of software function modules and sold or used as independent products, they can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present application, or the part that contributes to the prior art, or the part of the technical solution, can be embodied in the form of a software product. The computer software product is stored in a storage medium and includes several instructions for enabling a computer device (which can be a personal computer, server, or network device, etc.) to execute all or part of the steps of the method described in each embodiment of the present application. The aforementioned storage medium includes various media that can store program codes, such as a USB flash drive, a mobile hard disk, a read-only memory (ROM), a random access memory (RAM), a magnetic disk, or an optical disk.

[0061] The above description is merely a specific embodiment of the present application, but the scope of protection of the present application is not limited thereto. Any changes or substitutions that can be easily conceived by a person skilled in the art within the technical scope disclosed in this application should be included in the scope of protection of the present application. Therefore, the scope of protection of the present application should be based on the scope of protection of the claims.

[0062] Finally: The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, improvements, etc. made within the spirit and principles of the present invention should be included in the scope of protection of the present invention.

Claims

1. A centralized regional power analysis and management method based on new energy, characterized in that: The steps include: S1. Collect the current signal, voltage signal and ambient air pressure data of the DC side of the photovoltaic power station, perform high-frequency sampling on the current signal to generate a current waveform sequence, and calculate the difference in current change rate between adjacent sampling windows; S2. Adjust the arc noise energy judgment threshold based on the ambient air pressure data, and generate a dynamic judgment threshold based on the current change rate difference; S3, extracting the composite index of arc noise frequency band energy and voltage signal mutation in the current waveform sequence, and generating arc feature vectors based on the waveform oscillation periodicity; S4. Dynamically match the arc feature vector with a preset arc feature template, and calculate the matching weight and weight change trend; S5. When the weight change trend corresponding to the maximum matching weight exceeds the dynamic determination threshold and increases continuously, an arc restrike event is determined; S6. Generate a priority instruction sequence for string disconnection according to the duration of the arc restrike event; S7. Based on the impedance distribution of the grounding grid, the DC circuit breakers of the target strings are controlled to disconnect according to the priority according to the priority instruction sequence, and synchronous blocking instructions are sent to the inverters in the coverage area.

2. A centralized regional power analysis and management method based on new energy according to claim 1, characterized in that: S1 includes: When collecting the current signal of the DC side of the photovoltaic power station, the voltage signal and ambient air pressure data are collected synchronously, and the collection timestamp of each signal is recorded; Sampling the current signal at a preset high-frequency sampling frequency to generate a current waveform sequence comprising a plurality of continuous sampling points; The current waveform sequence is divided into adjacent sampling windows according to a preset window length, and the sum of squares of the current mean differences of adjacent sampling windows is calculated as the current change rate difference.

3. A centralized regional power analysis and management method based on new energy according to claim 1, characterized in that S2 include: According to the difference between the ambient air pressure data and the preset reference air pressure, the arc noise energy judgment threshold is adjusted in stages; generating a dynamic correction factor according to a ratio of the current change rate difference to a preset difference threshold, wherein the dynamic correction factor is a product of the ratio of the current change rate difference to the preset difference threshold and a preset gain coefficient; The arc noise energy judgment threshold value adjusted in stages is added to the dynamic correction factor to generate a dynamic judgment threshold value.

4. A centralized regional power analysis and management method based on new energy according to claim 3, characterized in that: Adjusting the arc noise energy threshold in stages includes: When the difference between the ambient air pressure data and the preset reference air pressure is greater than the first critical value, the arc noise energy judgment threshold is linearly reduced according to the difference ratio; when the difference between the ambient air pressure data and the preset reference air pressure is less than or equal to the first critical value, the arc noise energy judgment threshold is kept unchanged.

5. A centralized regional power analysis and management method based on new energy according to claim 1, characterized in that: S3 includes: Perform frequency band filtering on the current waveform sequence according to the preset arc noise frequency band range, and calculate the square integral of the energy of the filtered signal in the time window as the arc noise frequency band energy; Detect the sudden change gradient of the voltage signal within the time window. When the sudden change gradient exceeds the preset gradient threshold, record the gradient amplitude as the voltage signal sudden change indicator. Multiplying the arc noise frequency band energy and the voltage signal mutation index to generate a composite index; The number of zero crossings of the waveform oscillation period of the current waveform sequence within the time window is counted, and the arc feature vector is generated by combining the composite index.

6. A centralized regional power analysis and management method based on new energy according to claim 1, characterized in that S4 include: Perform dynamic time warping alignment on the time series of arc feature vector and preset arc feature template to eliminate time axis distortion error; Calculate the point-by-point Euclidean distance of the aligned time series and normalize the Euclidean distance to generate the initial matching weight; According to the distribution of the initial matching weights in the time window, a linear regression model is fitted and the slope of the weight change trend is output; The initial matching weight is modified based on the ratio of the weight change trend slope to the preset slope threshold to generate the matching weight.

7. A centralized regional power analysis and management method based on new energy according to claim 1, characterized in that: S5 includes: Traverse all matching weights and select the weight change trend corresponding to the maximum matching weight; Compare the weight change trend with the dynamic judgment threshold. When the weight change trend exceeds the dynamic judgment threshold, enter the incremental condition judgment; Check whether the weight change trend remains monotonically increasing within a preset number of consecutive time windows; If the weight change trend exceeds the dynamic judgment threshold and increases monotonically within a preset number of consecutive time windows, it is determined to be an arc restrike event.

8. A centralized regional power analysis and management method based on new energy according to claim 1, characterized in that S6 include: Determine a preset duration interval based on the duration of the arc restrike event, and map the duration to a corresponding priority level; Count the duration of multiple arc restrike events within the same time window, and sort them from longest to shortest by duration to generate a string disconnection sequence list; Based on the string disconnection sequence list, a disconnection delay time is assigned to each string, and a priority instruction sequence including a string identifier and a disconnection delay time is generated.

9. A centralized regional power analysis and management method based on new energy according to claim 1, characterized in that: S7 includes: Determine the boundaries of the blocking area based on the grounding grid impedance distribution data, and determine the coverage area based on the boundaries of the blocking area; According to the disconnection delay time in the priority instruction sequence, disconnection instructions are sent to the DC circuit breakers of the target strings in sequence, and the disconnection operations are controlled to be executed according to the priority. A blocking instruction is generated synchronously, which includes the identifier of the inverter within the coverage area and the blocking time window, and the blocking instruction is sent to the inverter through the communication protocol.

10. A centralized regional power analysis and management system based on new energy, used to implement a centralized regional power analysis and management method based on new energy according to any one of claims 1 to 9, characterized in that: include: Signal acquisition and processing module: collects the current signal, voltage signal and ambient air pressure data of the DC side of the photovoltaic power station, performs high-frequency sampling on the current signal to generate a current waveform sequence, and calculates the difference in current change rate between adjacent sampling windows; Threshold dynamic generation module: adjusts the arc noise energy judgment threshold based on ambient air pressure data, and generates a dynamic judgment threshold based on the current change rate difference; Feature extraction building block: extracts the composite index of arc noise frequency band energy and voltage signal mutation in the current waveform sequence, and generates arc feature vectors based on the waveform oscillation periodicity; Dynamic matching analysis module: dynamically matches the arc feature vector with the preset arc feature template, and calculates the matching weight and weight change trend; Event determination trigger module: When the weight change trend corresponding to the maximum matching weight exceeds the dynamic determination threshold and increases continuously, an arc restrike event is determined; Strategy generation and sequencing module: Generates a priority instruction sequence for string disconnection based on the duration of the arc restrike event; Execution control synchronization module: Based on the grounding grid impedance distribution, it controls the DC circuit breakers of the target strings to disconnect according to the priority instruction sequence, and sends synchronization blocking instructions to the inverters in the coverage area.

Images