An X-ray solar flare automatic detection method and system based on dual-mode triggering and nested interruption

Abstract

The application discloses a kind of X-ray solar flare automatic detection method and system based on dual-mode trigger and nested interruption, which comprises: obtaining satellite X-ray flux data, and pre-processing to obtain standard flux sequence;Standard flux sequence is simultaneously monitored by dynamic parallel dual-mode trigger of ultrafast mode and conventional mode;When any mode meets trigger criterion, parallel double end determination mechanism of conventional attenuation determination and nested interruption truncation is used, event end time is obtained, and flare event closed loop interval is determined;For each flare event closed loop interval, the maximum flux value and the corresponding peak time are extracted, the peak time is reversed backtracking, the start time is corrected, and the flare event and the corresponding start and end time are output.The present application greatly improves the detection rate of strong flare, effectively controls the error of time parameter determination, and eliminates the uncertainty of subjective research and judgment;Lightweight algorithm and high physical interpretability are suitable for engineering low-power deployment.

Description

Technical Field

[0001] This invention relates to the field of space weather monitoring and forecasting technology, specifically to an automatic detection method and system for X-ray solar flares based on dual-mode triggering and nested interruption. Background Technology

[0002] Solar flares are among the most violent solar activities. Their eruptions cause sudden disturbances in the ionosphere and are often accompanied by coronal mass ejections (CMEs) and solar proton events (SEPs), posing serious threats to spacecraft safety, shortwave communications, and ground power grids. Therefore, establishing a high-precision, low-latency automatic flare detection system is a core component of operational space weather forecasting.

[0003] Currently, flare detection methods using X-ray sensor data from geostationary orbit environmental satellites suffer from the following shortcomings: 1. Statistical analysis methods based on active region features: The performance of these methods is highly dependent on the completeness of feature selection. When solar activity reaches its peak, the active region evolves dramatically, and features are cross-coupled, making it easy for single statistical correlations to fail and difficult to effectively distinguish between real flare eruptions and background disturbances. 2. Rule-based expert system methods: These methods transform empirical knowledge into IF-THEN logical rules (such as the WOLF system). These methods lack dynamic adaptability; when faced with complex scenarios (such as consecutive flares), the expert system cannot recognize new rules, easily leading to missed flare detections. 3. Machine learning and deep learning methods (such as one-dimensional convolutional neural networks): Although they have strong feature extraction capabilities, their "black box" nature results in poor physical interpretability. Furthermore, model performance is highly dependent on the distribution of the training set, and their generalization ability across solar activity cycles is unknown. Additionally, deep learning models typically rely on high-performance computing power such as GPUs, which does not meet the low-power engineering deployment requirements of onboard processors or station edge devices.

[0004] In summary, during periods of maximum solar activity, existing technologies often misclassify highly overlapping "series of flares" as a single event with an extremely long duration, leading to missed flare detections, undercounting, and severe distortion in the determination of start and end times. Furthermore, existing high-precision models often involve high computational costs and cannot meet the stringent requirements of operational forecasting for lightweight design and high robustness. Summary of the Invention

[0005] To address the shortcomings of existing technologies in segmenting complex nested events such as "consecutive flares" during periods of maximum solar activity, including inaccurate start-end time determination, poor physical interpretability, and high computational consumption of existing deep learning models, this invention provides an automatic X-ray solar flare detection method and system based on dual-mode triggering and nested interruption. This invention aims to achieve dynamic decoupling and accurate segmentation of consecutive flare events through pure time-series morphological analysis, while also ensuring accurate capture of both pulsed and slowly varying flares. It outputs a standardized, high-precision list of flare events under low-power conditions without requiring GPU computing power.

[0006] In view of this, the present invention proposes an automatic detection method for X-ray solar flares based on dual-mode triggering and nested interruption, comprising: Step 1: Acquire satellite X-ray flux data and preprocess it to obtain a standard flux sequence; Step 2: Perform dynamic parallel dual-mode triggering monitoring on the standard flux sequence simultaneously in both high-speed and normal modes; Step 3: When any mode meets the trigger criterion, a parallel dual termination criterion mechanism of conventional attenuation determination and nested interrupt truncation is adopted to obtain the event end time and determine the closed loop interval of the flare event. Step 4: For each closed-loop interval of a flare event, extract the maximum flux value and the corresponding peak time, backtrack in reverse based on the peak time, correct the start time, and output the flare event and its corresponding start and end times.

[0007] As an improvement to the above method, the preprocessing in step 1 includes: A time series with fixed intervals is constructed. A short-time de-pulse method is used to identify single-point anomalies using local medians. Abnormal data is replaced by neighbor interpolation, and the physical rising edge of the real flare is preserved to obtain a standard flux sequence.

[0008] As an improvement to the above method, the high-speed mode in step 2 is used to capture rapidly rising pulsed flares, and the triggering criterion is: the flux increase exceeds a set ratio within 2 minutes. And the flux sequence is to It maintains a strictly monotonically increasing trend within minutes, that is, it satisfies the following formula:

[0009] in, They are respectively The flux sequence.

[0010] As an improvement to the above method, the conventional mode of step 2 is used to capture long-duration flares with a gentle rising phase, and the triggering criterion is: the cumulative flux increase exceeds a set proportion within 4 minutes. And the flux sequence is to It maintains a strictly monotonically increasing trend within minutes, that is, it satisfies the following formula:

[0011] in, They are respectively The flux sequence.

[0012] As an improvement to the above method, the conventional attenuation determination in step 3 includes: Searching backward from the peak of the flare, when the flux first drops to the half-height threshold... The event is considered complete when the time is determined, and the corresponding time is the end time. If the flux has fallen back to the background level before dropping to the half-high threshold, then the event is considered complete. The end time is then taken as the time of fallback; where, the half-height threshold for:

[0013] In the formula, This represents the maximum throughput value recorded during the event tracing. The moment the event is triggered And the average flux over the preceding few minutes.

[0014] As an improvement to the above method, the nested interrupt truncation in step 3 includes: During the flux drop towards half-width at half-maximum, a sliding time window is used to continuously monitor the local minimum points on the decay curve, and the corresponding time is set as the valley time. If the flux changes from the half-width to the full-width limit before the end time is reached... If a rebound occurs again, and this rebound sequence strictly satisfies the trigger criteria for either the high-speed mode or the normal mode in step 2, then the interruption and truncation operation is executed directly: The time frame is forcibly determined to be the absolute end time of the currently tracked old event, and the old event parameters are saved; at the same time, This establishes the start time of the new secondary flare and initiates independent tracking of the new flare event.

[0015] As an improvement to the above method, step 4 includes: Extract the maximum flux value within the closed-loop interval of each flare event. and the time of occurrence Energy levels were determined according to the X-ray flare standard classification, and those that were excluded were discarded. Slight fluctuations; From the determined peak time The search proceeds backward along the timeline until the flux first falls to the baseline level. Or, if a clear local minimum stationary point is encountered, the backtracking point is formally anchored as the corrected precise start time. The output includes a normalized list of physical events, including flare events, energy levels, and corresponding start and end times.

[0016] On the other hand, the present invention provides an automatic X-ray solar flare detection system based on dual-mode triggering and nested interruption, comprising: The preprocessing module is used to acquire satellite X-ray flux data and perform preprocessing to obtain a standard flux sequence; The dynamic dual-mode trigger monitoring module is used to perform dynamic parallel dual-mode trigger monitoring of standard throughput sequences in both high-speed and normal modes. The event tracking and parallel dual-end determination module is used to obtain the event end time and determine the closed-loop interval of the flare event when either mode meets the trigger criterion, employing a parallel dual-end determination mechanism of conventional attenuation determination and nested interrupt truncation; and The event parameter extraction and post-processing module is used to extract the maximum flux value and the corresponding peak time for each flare event closed loop interval, perform reverse backtracking based on the peak time, correct the start time, and output the flare event and its corresponding start and end times.

[0017] Compared with the prior art, the advantages of the present invention are: 1. High precision in nested event segmentation significantly improves the detection rate of strong flares: By using nested interrupt logic, this method effectively avoids the defect of forcibly merging multiple independent flares into a single event. Experimental data shows that under the complex observation environment of solar maximum, this invention achieves 100% accuracy in identifying X-class flares and an F1 score as high as 0.969 for M-class flares, with an accuracy of 97.3%, eliminating the detection blind spots of traditional detection algorithms. 2. Effective control of time parameter measurement errors and elimination of subjective judgment uncertainties: Utilizing dual termination criteria and reverse backtracking correction technology, this invention avoids systematic biases of "over-truncation" or "over-stretching" when characterizing the flare lifecycle. The mean absolute error (MAE) of the start and end times and duration for the entire sample is controlled within five minutes (start and end time MAE are approximately 2.3 minutes, and duration MAE is approximately 4.3 minutes), meeting the needs of refined flare research and operational forecasting. 3. Lightweight algorithm with high physical interpretability, adaptable to low-power engineering deployment: This method completely abandons the "black box" architecture of deep learning and operates solely based on the physical morphological features in time series data. It consumes extremely low computing resources and does not require high-performance computing power such as GPUs. It can be directly deployed in spaceborne processors or station edge computing devices, perfectly meeting the engineering operation requirements of space weather early warning systems for low latency, high robustness, and lightweight operation. Attached Figure Description

[0018] Figure 1 This is a timing flowchart of an automatic X-ray solar flare detection method based on dual-mode triggering and nested interruption according to the present invention. Detailed Implementation

[0019] This invention provides an automatic detection method for X-ray solar flares based on dual-mode triggering and nested interruption, comprising the following steps, such as... Figure 1 As shown: Step 1: Acquisition and Preprocessing of X-ray Flux Data Acquire soft X-ray flux data observed by satellite X-ray sensors and construct a standard time series with a uniform resolution of 1 minute. .

[0020] To address the "spike" spurious signals generated by high-energy particle impact detectors, a short-time de-pulse method is employed. This method utilizes local medians for single-point anomaly detection and replaces anomalous data with neighbor-point interpolation, thus preserving the true physical rising edge of the flare.

[0021] Step 2: Execute dynamic dual-mode trigger monitoring For standard time series Simultaneous monitoring of "Extreme Speed ​​Mode" and "Normal Mode": (1) Extreme Speed ​​Mode: Used to capture rapidly rising pulse-type flares. The trigger criterion is: the flux increase must exceed a set ratio within 2 minutes. And the flux sequence is to It maintains a strictly monotonically increasing trend over a period of minutes. The mathematical expression is: .

[0022] (2) Normal mode: Used to capture long-duration flares with a gentle rising phase. The trigger criterion is: the cumulative flux increase must exceed a set percentage within 4 minutes. And the flux sequence is to It maintains a strictly monotonically increasing trend over a period of minutes. The mathematical expression is: .

[0023] When any of the triggering criteria of the above modes is met, the system enters the "event tracking" state and records the current time. Marked as suspected start time.

[0024] Step 3: Dual termination determination of event tracking and parallel processing During the flare decay phase tracking stage, a parallel dual termination determination mechanism is designed to accurately define the event end time: (1) Conventional attenuation determination (half-width at half maximum): Search backward from the peak of the flare, when the flux first drops to the half-width at half maximum threshold. The event is considered complete when the time is reached. The calculation formula is: .in, This represents the maximum throughput value recorded during the event tracing. The local background baseline before the event is triggered (defined as the event trigger time). (And the average flux over the preceding few minutes). If the flux has fallen back to the background level before dropping to the half-high value. The time of retracement is taken as the end time.

[0025] (2) Nested interrupt truncation (for consecutive flares): While waiting for the flux to drop to half-width at half-maximum, the system continuously monitors the local minimum point on the decay curve (set as the valley moment) with a sliding time window. If the flux from [a certain value] before the end time of the half-width limit is reached... If a rebound occurs again, and the rebound sequence strictly satisfies the trigger criterion of "high-speed mode" or "normal mode" in step 2, then the interruption and truncation operation is executed directly: The time frame is forcibly determined to be the absolute end time of the currently tracked old event, and the old event parameters are saved; at the same time, Establish the start time of the new secondary flare, reset the peak tracker, and begin independent tracking of the new event.

[0026] Step 4: Event Parameter Extraction and Post-processing (1) Peak energy level calibration and noise filtering: Extract the maximum flux value within each closed-loop event interval. and the time of occurrence Energy levels were determined according to the X-ray flare standard classification. (Removal) (i.e., slight fluctuations below level C1.0).

[0027] Precise start time correction (backtracking): starting from a defined peak moment The search proceeds backward along the timeline until the flux first falls to the baseline level. Or, it may encounter a significant local minimum stationary point. The backtracking point is then formally anchored as the corrected, precise start time to eliminate the inherent time delay of the triggering algorithm.

[0028] Invention point: 1. Dynamic dual-mode triggering design: Compared with the traditional single threshold or fixed slope triggering, this invention is the first to create a detection architecture that combines high-speed and conventional parallel triggering, which can adapt to the differences in morphology between pulse-type and slowly varying flares.

[0029] 2. Attenuation Phase Nested Interruption Truncation Mechanism: To solve the problem of frequent missed detections of "consecutive flares" during the maxima phase, this invention breaks away from the rigid setting of traditional algorithms that "must wait for the flux to fall back to the background or half-width at half-maximum." This is achieved by dynamically locking the valley point during the attenuation phase. It also performs a secondary trigger test, enabling the precise "physical stripping" of secondary burst events from the decaying tail of a large flare without waiting for the flux to completely fall back.

[0030] 3. Refined reverse correction mechanism for time parameters: In response to the inherent time lag defect of derivative / slope triggering, this invention proposes a reverse backtracking correction logic with the peak value as the anchor point, which greatly improves the positioning accuracy of flare start and end times.

[0031] The technical solution of the present invention will be described in detail below with reference to the accompanying drawings and embodiments.

[0032] Example 1 Embodiment 1 of this application proposes an automatic X-ray solar flare detection method based on dual-mode triggering and nested interruption. The operation process is as follows: 1. Obtain soft X-ray scientific data in the 1-8 Å band from the NOAA GOES series satellites and construct a standard sequence with a 1-minute time interval. Local median filtering is used to remove steep pulse singularities caused by high-energy particles.

[0033] 2. In the dynamic dual-mode triggering phase, this embodiment sets a lower limit for the triggering ratio of "Extreme Speed ​​Mode". (i.e., a price increase of over 40% within 2 minutes), set the lower limit of the trigger ratio for "normal mode". (i.e., a price increase of over 20% within 4 minutes). Real-time streaming input sequence data; once either of these conditions is met, a suspected starting point is marked. And then proceed to event tracking.

[0034] 3. During the decay phase of event tracing, at the system cache trigger time The average flux over the first 4 minutes was used as the local background baseline. As flux declines from its peak, the system calculates the full width at half maximum (FWHM) threshold in real time. If flux falls to Previously, the system detected a local minimum. If subsequent sequences again satisfy any of the triggering criteria in step 2, then a nested interrupt is immediately triggered: Register the end time of the current event, and The calculation is restarted as the starting point for a new event. If no rebound occurs, it will naturally fall to... Record the end time when the background is horizontal.

[0035] 4. In the post-processing stage, the system traverses all closed-loop independent event segments. Peak throughput is discarded according to international standards. Background thermal noise events (below C1.0 level). Then, for each retained event, from the recorded peak time... Compare step by step forward until the flux falls below The point is updated to reflect the true initiation time of the flare event. Finally, a standardized list of physical events, including accurate occurrence and end times and energy levels (e.g., M2.0, X1.5), is output to the space weather warning terminal.

[0036] Example 2 Embodiment 2 of the present invention provides an automatic X-ray solar flare detection system based on dual-mode triggering and nested interruption, implemented based on the method of Embodiment 1, including: The preprocessing module is used to acquire satellite X-ray flux data and perform preprocessing to obtain a standard flux sequence; The dynamic dual-mode trigger monitoring module is used to perform dynamic parallel dual-mode trigger monitoring of standard throughput sequences in both high-speed and normal modes. The event tracking and parallel dual-end determination module is used to obtain the event end time and determine the closed-loop interval of the flare event when either mode meets the trigger criterion, using a parallel dual-end determination mechanism of conventional attenuation determination and nested interrupt truncation. The event parameter extraction and post-processing module is used to extract the maximum flux value and the corresponding peak time for each flare event closed loop interval, perform reverse backtracking based on the peak time, correct the start time, and output the flare event and its corresponding start and end times.

[0037] It is worth noting that in the embodiments of the above system, the modules included are divided according to functional logic, but are not limited to the above division, as long as the corresponding functions can be achieved; in addition, the specific names of each functional module are only for easy differentiation and are not used to limit the scope of protection of the present invention.

[0038] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and are not intended to limit it. Although the present invention has been described in detail with reference to the embodiments, those skilled in the art should understand that modifications or equivalent substitutions to the technical solutions of the present invention do not depart from the spirit and scope of the technical solutions of the present invention, and all such modifications or substitutions should be covered within the scope of the claims of the present invention.

Claims

1. An automatic detection method for X-ray solar flares based on dual-mode triggering and nested interruption, comprising: Step 1: Acquire satellite X-ray flux data and preprocess it to obtain a standard flux sequence; Step 2: Perform dynamic parallel dual-mode triggering monitoring on the standard flux sequence simultaneously in both high-speed and normal modes; Step 3: When any mode meets the trigger criterion, a parallel dual termination criterion mechanism of conventional attenuation determination and nested interrupt truncation is adopted to obtain the event end time and determine the closed loop interval of the flare event. Step 4: For each closed-loop interval of a flare event, extract the maximum flux value and the corresponding peak time, backtrack in reverse based on the peak time, correct the start time, and output the flare event and its corresponding start and end times.

2. The automatic detection method for X-ray solar flares based on dual-mode triggering and nested interruption as described in claim 1, characterized in that, The preprocessing in step 1 includes: A time series with fixed intervals is constructed. A short-time de-pulse method is used to identify single-point anomalies using local medians. Abnormal data is replaced by neighbor interpolation, and the physical rising edge of the real flare is preserved to obtain a standard flux sequence.

3. The automatic X-ray solar flare detection method based on dual-mode triggering and nested interruption according to claim 2, wherein the rapid mode in step 2 is used to capture rapidly rising pulsed flares, and the triggering criterion is: the flux increases by more than a set proportion within 2 minutes. And the flux sequence is to It maintains a strictly monotonically increasing trend within minutes, that is, it satisfies the following formula: in, They are respectively The flux sequence.

4. The automatic X-ray solar flare detection method based on dual-mode triggering and nested interruption as described in claim 1, wherein the conventional mode in step 2 is used to capture long-duration flares with a gentle rising phase, and the triggering criterion is: the cumulative flux increase exceeds a set proportion within 4 minutes. And the flux sequence is to It maintains a strictly monotonically increasing trend within minutes, that is, it satisfies the following formula: in, They are respectively The flux sequence.

5. The automatic detection method for X-ray solar flares based on dual-mode triggering and nested interruption as described in claim 1, characterized in that, The conventional attenuation determination in step 3 includes: Searching backward from the peak of the flare, when the flux first drops to the half-height threshold... The event is considered complete when the time is determined, and the corresponding time is the end time. If the flux has fallen back to the background level before dropping to the half-high threshold, then the event is considered complete. The end time is then taken as the time of fallback; where, the half-height threshold for: In the formula, This represents the maximum throughput value recorded during the event tracing. The moment the event is triggered And the average flux over the preceding few minutes.

6. The automatic detection method for X-ray solar flares based on dual-mode triggering and nested interruption as described in claim 1, characterized in that, The nested interrupt truncation in step 3 includes: During the flux drop towards half-width at half-maximum, a sliding time window is used to continuously monitor the local minimum points on the decay curve, and the corresponding time is set as the valley time. If the flux changes from the half-width to the full-width limit before the end time is reached... If a rebound occurs again, and this rebound sequence strictly satisfies the trigger criteria for either the high-speed mode or the normal mode in step 2, then the interruption and truncation operation is executed directly: The time frame is forcibly determined to be the absolute end time of the currently tracked old event, and the old event parameters are saved; at the same time, This establishes the start time of the new secondary flare and initiates independent tracking of the new flare event.

7. The automatic detection method for X-ray solar flares based on dual-mode triggering and nested interruption as described in claim 5, characterized in that, Step 4 includes: Extract the maximum flux value within the closed-loop interval of each flare event. and the time of occurrence Energy levels were determined according to the X-ray flare standard classification, and those that were excluded were discarded. Slight fluctuations; From the determined peak time The search proceeds backward along the timeline until the flux first falls to the baseline level. Or, if a clear local minimum stationary point is encountered, the backtracking point is formally anchored as the corrected precise start time. The output includes a normalized list of physical events, including flare events, energy levels, and corresponding start and end times.

8. An automatic X-ray solar flare detection system based on dual-mode triggering and nested interruption, characterized in that, include: The preprocessing module is used to acquire satellite X-ray flux data and perform preprocessing to obtain a standard flux sequence; The dynamic dual-mode trigger monitoring module is used to perform dynamic parallel dual-mode trigger monitoring of standard throughput sequences in both high-speed and normal modes. The event tracking and parallel dual-end determination module is used to obtain the event end time and determine the closed-loop interval of the flare event when either mode meets the trigger criterion, using a parallel dual-end determination mechanism of conventional attenuation determination and nested interrupt truncation. and The event parameter extraction and post-processing module is used to extract the maximum flux value and the corresponding peak time for each flare event closed loop interval, perform reverse backtracking based on the peak time, correct the start time, and output the flare event and its corresponding start and end times.

Images