A multi-ship GNSS receiver interference information comprehensive processing method and system

Abstract

The present application relates to the field of radio navigation technology, and more particularly to a multi-ship satellite navigation receiver interference information comprehensive processing method and system, the method comprising: collecting real-time original interference data and auxiliary data of each ship satellite navigation receiver; preprocessing the collected original interference data and auxiliary data of each ship satellite navigation receiver to obtain preprocessed synchronous interference data and preprocessed synchronous auxiliary data; performing multi-ship interference information collaborative fusion and misjudgment correction to obtain fused interference feature data set; sequentially completing interference subtype subdivision, multi-interference source separation, single-source precise positioning, and multi-dimensional parameter evaluation to ultimately obtain the position of each interference source, multi-dimensional parameter evaluation, and comprehensive grading result; and outputting the position of each interference source, multi-dimensional parameter evaluation, and comprehensive grading result in a standardized format. The present application realizes multi-ship collaborative precise positioning of interference sources and multi-dimensional parameter evaluation.

Description

Technical Field

[0001] This invention relates to the field of radio navigation technology, and in particular to a method and system for comprehensively processing interference information from multiple ship satellite navigation receivers. Background Technology

[0002] Global Navigation Satellite System (GNSS), or satellite navigation system for short, is a core navigation device for ships during navigation, and its positioning accuracy and stability are directly related to the safety of ship navigation. With the rapid development of maritime shipping, GNSS receivers are being used more and more widely. However, GNSS signals are susceptible to various types of interference, including human interference (such as suppression interference, spoofing interference, and relay spoofing interference) and natural interference (such as multipath effects, hull obstruction, and atmospheric attenuation). These interferences can lead to increased positioning deviations and signal loss in GNSS receivers, and in severe cases, may cause safety accidents such as ship collisions and groundings.

[0003] Currently, interference processing for ship GNSS receivers mostly adopts a single-device independent detection mode. This means each ship's GNSS receiver can only collect interference signals at its own location and perform simple interference identification, unable to correlate interference data from surrounding ships. This mode has the following technical limitations: First, the observation range of a single ship receiver is limited, making it difficult to accurately locate the interference source and evaluate multi-dimensional parameters. It can only determine whether interference exists, but cannot determine the location, power, coverage area, or type of interference. Second, ships are susceptible to factors such as wave motion, hull obstruction, and near-shore multipath effects during navigation, making single-device detection prone to misjudgment of interference. Third, existing technologies lack effective collaborative processing solutions designed for dynamic ship scenarios, failing to fully utilize existing ship equipment resources. Summary of the Invention

[0004] The technical problem to be solved by the present invention is to provide a method and system for comprehensive processing of interference information from multiple ship satellite navigation receivers. This method and system solves the problems of limited observation range, difficulty in locating interference sources, high misjudgment rate, and reliance on dedicated equipment in the detection of single ship satellite navigation receivers. It enables accurate location of interference sources and multi-dimensional parameter evaluation through multi-ship collaboration.

[0005] This invention is achieved through the following technical solution: A method for comprehensively processing interference information from multiple ship satellite navigation receivers includes the following steps: S1: Select multiple ships sailing in the same sea area as distributed observation nodes, and collect the raw interference data of each ship's satellite navigation receiver in real time, as well as auxiliary data including the ship's real-time position coordinates, navigation status data, environmental data and receiver equipment parameters. S2: Preprocess the raw interference data and auxiliary data collected from each ship's satellite navigation receiver, and simultaneously complete the spatiotemporal synchronization of the raw interference data and auxiliary data to obtain preprocessed synchronization interference data and preprocessed synchronization auxiliary data. S3: Taking the preprocessed synchronization interference data as the core processing object, and combining it with the preprocessed synchronization auxiliary data, the collaborative fusion of multi-ship interference information and misjudgment correction are carried out to obtain the fused interference feature dataset. S4: Based on the fused interference feature dataset, the interference subtype subdivision, multi-interference source separation, single-source precise localization, and multi-dimensional parameter evaluation are completed in sequence to finally obtain the location, multi-dimensional parameter evaluation and comprehensive classification results of each interference source. S5: Output the location of each interference source, multi-dimensional parameter evaluation and comprehensive classification results in a standardized format, and implement them in practice.

[0006] The optimized raw interference data collected in step S1 for each ship's satellite navigation receiver covers two categories: suppression interference features and deception interference features.

[0007] In the optimized version, the navigation status data in step S1 includes speed, heading, roll and pitch angles; the environmental data includes sea state, atmospheric visibility, and near-shore distance; and the receiver equipment parameters include antenna gain, noise figure, and equipment accuracy level.

[0008] In the optimized version, the acquisition frequency of the original interference data in step S1 is ≥10Hz, and the acquisition frequency of the auxiliary data is ≥1Hz.

[0009] Furthermore, the method for preprocessing the raw interference data and auxiliary data collected from the satellite navigation receivers of each ship in step S2 is as follows: S211: During normal, interference-free periods, if the data within the window follows a normal distribution, the threshold method combined with the Laida criterion based on the 3σ principle is used for outlier detection. If the data within the window does not follow a normal distribution, the interquartile range method is used for outlier detection. If the standard deviation of the signal strength within the calculation window exceeds the preset threshold range, the original data of interference is judged as outlier data and removed. During periods of suspected interference, when the signal strength or carrier-to-noise ratio is determined to be a deceptive interference feature by the preliminary discrimination rules, outlier removal is suspended, and only extreme invalid values ​​caused by hard faults in the filtering equipment are removed. S212: Standardize the interference data from different models of satellite navigation receivers, and convert the signal strength, timestamp, and coordinate system into a standard format to obtain standardized interference data; S213: Perform initial screening based on the auxiliary data of a single ship. If any item in the auxiliary data exceeds the preset threshold, mark the data for that period as a suspected occlusion misjudgment and remove it, thus completing the first level of misjudgment correction.

[0010] Furthermore, in step S3, the fused interference feature dataset is obtained using the following method: S311: Each ship periodically broadcasts its computing power level, communication link signal-to-noise ratio, and online status, and calculates a comprehensive election score according to formula (1). The ship with the highest comprehensive election score is elected as the core node, responsible for aggregating the observation data of all ships as distributed observation nodes and uniformly executing all subsequent steps. The core node is dynamically determined according to the preset election cycle. If a ship of the core node loses contact for more than a set time, the remaining ships trigger an emergency re-election mechanism based on the comprehensive election score. (1); in: This represents the overall election score. This represents the computing power score. This represents the computing power score coefficient. Indicates the communication quality score. This represents the communication quality score coefficient; S312: Based on the three dimensions of ship position reliability, receiver equipment accuracy, and data integrity, the original weight coefficient of interference data for each ship is calculated according to equation (2), and then the original weight coefficient is normalized by linear mapping to... The final weight range; (2); in: Indicates the first The original weighting coefficients of the ships, This represents the location reliability weighting coefficient. Indicates the first The ship The reliability coefficient of the ship's position. This represents the weighting coefficient for equipment accuracy. Indicates the first The accuracy coefficient of the ship's equipment This represents the data integrity weighting coefficient. Indicates the first Ship data integrity coefficient; S313: Introduce spatial consistency of multi-ship positioning deviation to dynamically correct the position reliability coefficient and update it synchronously with the positioning results of satellite navigation receiver. According to formula (3), the data integrity coefficient is updated synchronously with the fusion calculation cycle, and the weight coefficient is updated synchronously. (3); in: Indicates the duration of valid data. Indicates the total observation duration; S314: The preprocessed synchronous interference data of multiple ships are correlated and fused according to the time and spatial dimensions. Based on the updated weight coefficients, the fused interference signal strength of the multi-ship observations is calculated according to equation (4): (4); in: This indicates the intensity of the fusion interference signal from multiple ship observations. Indicates the first The strength of the interference signal measured by the ship Indicates the total number of ships; S315: Construct a misjudgment correction model based on the supervised machine learning binary classification algorithm CART decision tree, and input the feature vector containing the intensity of fused interference signals from multiple ship observations into the misjudgment correction model for the second-level misjudgment correction, and output the binary classification result containing real interference or misjudged interference. S316: Perform consistency verification on the fused interference signal strength observed by multiple ships. If the deviation between the fused interference signal strength observed by multiple ships corresponding to the same interference source and the interference signal strength measured by a single ship is less than or equal to 3dB, it is determined that the consistency verification requirement is met, and there is no need to adjust the weight coefficient of the ship. If the deviation between the fused interference signal strength observed by multiple ships corresponding to the same interference source and the interference signal strength measured by a single ship is greater than 3dB, it is determined that the consistency verification requirement is not met. In this case, an adjustment coefficient is generated according to the inverse proportional rule that the larger the deviation, the lower the weight. The original weight coefficient of the ship is corrected and normalized to obtain the fused interference feature dataset.

[0011] Furthermore, the method for obtaining the location of each interference source, multi-dimensional parameter evaluation, and comprehensive classification results in step S4 is as follows: S411: Based on the binary classification results and the fused interference feature dataset, the confirmed real interference events are further subdivided into human interference subtypes and natural interference subtypes; S412: Based on the fused interference feature dataset, multiple interference sources existing simultaneously in the sea area are separated and identified using a two-stage mechanism of hard frequency differentiation before positioning and fine differentiation of coarse frequency clustering after coarse positioning. After separation and identification, the observation dataset is split according to clusters so that each cluster corresponds to an independent co-frequency interference source. S413: For each independent interference source after separation, if the human interference subtype is suppression interference, the time difference of arrival technique is adopted, and the ship that receives the interference signal first is selected as the reference ship. Based on the time difference of arrival between the reference ship and other ships, a nonlinear positioning equation set is established as Equation (5). The Chan algorithm is used to solve the nonlinear positioning equation set to obtain the final preliminary position coordinates of the suppression interference source. Then, the mean-shift clustering algorithm is used to cluster multiple sets of final preliminary position coordinates, remove outliers, and take the cluster center as the final positioning result of the suppression interference source. (5); in: This indicates the time difference between the arrival of the reference vessel and other vessels. Represents the speed of light in a vacuum. Indicates the interference source to the first The distance between the ships Indicates the distance from the interference source to the reference vessel; S414: If the human interference subtype is deception interference, then read the deception signal pseudo-code phase observation value output by the tracking loop of each ship's satellite navigation receiver, eliminate the integer ambiguity of the deception signal pseudo-code phase observation value by combining chip period constraints with coarse time determination, obtain the precise distance difference between the two ships after ambiguity elimination, and then use the Chan algorithm to solve the nonlinear positioning equation set to obtain the final preliminary position coordinates of the deception interference source. For ships equipped with dual-antenna direction finding, the final preliminary position coordinates are then corrected by dual-antenna direction finding cross-positioning to obtain the corrected deception interference source positioning result. S415: For each independent interference source, perform multi-dimensional parameter classification and comprehensive classification based on the final location result of the suppressed interference source or the corrected location result of the deceitful interference source and the split observation dataset; S416: Package the multi-dimensional parameter classification evaluation and comprehensive grading results into a structured dataset and push it synchronously to the shipborne terminal and shore-based platform.

[0012] Optimized auxiliary data will be provided if any of the following exceeds the preset threshold: roll angle > 15°, antenna elevation angle < 5°, or near-shore distance < 1km.

[0013] The optimized step S5 outputs the interference detection results, the final locations of each interference source, and the multi-dimensional evaluation and comprehensive classification results in a standardized format, including three output methods: Shipborne terminal output: Overlay the location of the interference source, multi-dimensional assessment and comprehensive classification results, and coverage area onto the ship's electronic chart, and issue warnings to the crew in the form of sound and light, and recommend avoidance routes; Shore-based platform output: Summarizes all interference events in the sea area, generates a regional interference situation map, and supports historical interference data backtracking and statistical analysis; Data interface output: Provides a standard REST API interface, supporting integration with Automatic Identification System (AIS), Vessel Traffic Management System (VCR), and Vessel Navigation System (GNSS).

[0014] A multi-ship satellite navigation receiver jamming information integrated processing system is used to execute a multi-ship satellite navigation receiver jamming information integrated processing method as described in any one of the above, which includes a data acquisition module, a preprocessing module, a multi-source information collaborative fusion module, a jamming source localization and evaluation module, and a result output module; The data acquisition module is used to select multiple ships sailing in the same sea area as distributed observation nodes, and to collect the raw interference data of each ship's satellite navigation receiver and auxiliary data including the ship's real-time position coordinates, navigation status data, environmental data and receiver equipment parameters in real time. The preprocessing module is used to preprocess the raw interference data and auxiliary data of each ship's satellite navigation receiver, and at the same time complete the spatiotemporal synchronization of the raw interference data and auxiliary data to obtain preprocessed synchronization interference data and preprocessed synchronization auxiliary data. The multi-source information collaborative fusion module is used to take the pre-processed synchronization interference data as the core processing object, and combine it with the pre-processed synchronization auxiliary data to perform collaborative fusion and misjudgment correction of multi-ship interference information, so as to obtain the fused interference feature dataset. The interference source localization and evaluation module, based on the fused interference feature dataset, sequentially completes interference subtype subdivision, multi-interference source separation, single-source precise localization, and multi-dimensional parameter evaluation, and finally obtains the location, multi-dimensional parameter evaluation, and comprehensive classification results of each interference source. The result output module is used to output the interference detection results, the location of each interference source, the multi-dimensional parameter classification evaluation and comprehensive classification results in a standardized format, and to implement them in practice.

[0015] Beneficial effects of the invention: This invention provides a method and system for comprehensive processing of interference information from multiple shipborne satellite navigation receivers. Using multiple ships in the same sea area as distributed observation nodes, and upgrading the software of existing shipborne satellite navigation receivers, a distributed interference monitoring network of "ship-end edge computing + multi-ship collaborative fusion" can be constructed. A two-level misjudgment correction mechanism improves detection accuracy. Adaptive detection and location algorithms are designed for both suppression and deception interference. Combined with a maritime propagation model, multi-dimensional parameter evaluation of interference sources is achieved. Without increasing hardware costs, this invention overcomes the limitations of single-device observation and comprehensively enhances the anti-interference capability of maritime satellite navigation systems. Attached Figure Description

[0016] Figure 1 This is a schematic diagram of the overall process of the present invention. Detailed Implementation

[0017] A method for comprehensively processing interference information from multiple ship satellite navigation receivers, the overall flowchart of which is shown in Figure 1, specifically includes the following steps: S1: Select multiple ships sailing in the same sea area as distributed observation nodes, and collect the raw interference data of each ship's satellite navigation receiver in real time, as well as auxiliary data including the ship's real-time position coordinates, navigation status data, environmental data and receiver equipment parameters. Specifically, the raw interference data collected from each ship's satellite navigation receiver covers two categories: suppression interference characteristics and deception interference characteristics. The raw interference data is collected at a frequency ≥10Hz.

[0018] Among them, the characteristics of suppression interference include received signal strength, interference center frequency, signal modulation method, interference start time, interference duration, and carrier-to-noise ratio drop slope.

[0019] The preliminary identification rule for suppressing interference on a single device is as follows: the carrier-to-noise ratio of the GNSS signal is below 40dB-Hz for 3 consecutive seconds, and the decrease exceeds 10dB-Hz. The suppression of interference is characterized by a uniform decrease in the carrier-to-noise ratio across the entire frequency band, with a decrease slope ≥5dB / s, and the signal is a continuous wave or narrowband modulation.

[0020] The characteristics of spoofing interference include real-time positioning deviation, pseudo-code phase offset, and the number of abnormal satellite signals. The preliminary identification rules for spoofing interference by a single device are as follows: the carrier-to-noise ratio does not decrease significantly, but the positioning result shows a jump, and there is a pseudo signal with the same code pattern as the real satellite signal; the signal delay characteristics of forwarding spoofing interference are obvious, and the positioning deviation increases slowly over time.

[0021] The auxiliary data acquisition frequency is ≥1Hz. The navigation status data includes speed, heading, roll and pitch angles, the environmental data includes sea state, atmospheric visibility, and near-shore distance, and the receiver equipment parameters include antenna gain, noise figure, and equipment accuracy level.

[0022] S2: Preprocess the raw interference data and auxiliary data collected from each ship's satellite navigation receiver, and simultaneously complete the spatiotemporal synchronization of the raw interference data and auxiliary data to obtain preprocessed synchronization interference data and preprocessed synchronization auxiliary data. Specifically, the method for preprocessing the raw interference data and auxiliary data collected from each ship's satellite navigation receiver is as follows: For filtering invalid data, a time-segmented mechanism of "baseline anomaly detection + interference period data exemption" can be adopted to distinguish between normal interference periods and suspected interference periods.

[0023] S211: During normal, interference-free periods, if the data within the window follows a normal distribution, a threshold method combined with the Laida criterion based on the 3σ principle is used for outlier detection. If the data within the window does not follow a normal distribution, the interquartile range method is used for outlier detection. If the standard deviation of the signal strength within the calculated window exceeds a preset threshold, the original data causing interference is judged as outlier and removed. During periods of suspected interference, when the signal strength or carrier-to-noise ratio is determined to have deceptive interference characteristics using preliminary discrimination rules, outlier removal is suspended, and only extreme invalid values ​​caused by hard faults in the filtering equipment are removed. If the data within the window follows a normal distribution, the preset threshold is ±3. , This represents the standard deviation of the signal strength data within the window.

[0024] When using the interquartile range method for outlier detection: the upper quartiles are calculated after sorting the signal intensity samples within the window. and lower quartile Take the interquartile range The anomaly detection threshold is Data outside this range is considered outlier and removed.

[0025] The optimized option is to select a sliding window duration of 10 seconds. The basis for this setting is that a 10-second window can cover the typical fluctuation cycle of maritime signals while ensuring the real-time performance of anomaly detection.

[0026] S212: Standardize the interference data from different models of satellite navigation receivers, and convert the signal strength, timestamp, and coordinate system into a standard format to obtain standardized interference data; The optimized version converts signal strength to dBm units, with the conversion rule mapping the original AD sampling values ​​to standard received signal strength based on the satellite navigation receiver's factory calibration parameters; the timestamps are unified to UTC (Coordinated Universal Time); and the coordinate system is CGCS2000 geodetic coordinate system.

[0027] S213: Perform initial screening based on the auxiliary data of a single ship. If any item in the auxiliary data exceeds the preset threshold, mark the data for that period as a suspected occlusion misjudgment and remove it, thus completing the first level of misjudgment correction.

[0028] Specifically, when the roll angle in the auxiliary data is greater than 15°, the antenna elevation angle is less than 5°, or the near-shore distance is less than 1km, the data for that period is marked as a suspected misjudgment of occlusion and removed. Only high-confidence observations are retained to avoid invalid data from entering the subsequent fusion process and contaminating the results.

[0029] Achieve spatiotemporal synchronization of raw interference data and auxiliary data, that is, unify the time base of all ship data through GNSS satellite time synchronization, with an integer second time synchronization accuracy ≤1. The timestamp resolution for the arrival time of the interference signal is ≤0.1. This achieves time synchronization, providing a unified time reference for subsequent arrival time difference calculations; by unifying the coordinate systems and elevation references of all ships, spatial alignment is achieved, ensuring that the position data of multiple ships are under the same spatial reference system.

[0030] S3: Taking the preprocessed synchronization interference data as the core processing object, and combining it with the preprocessed synchronization auxiliary data, the collaborative fusion of multi-ship interference information and misjudgment correction are carried out to obtain the fused interference feature dataset. This step uses the edge nodes of the core nodes in the region to undertake the fusion computing task. With interference feature data as the core processing object, and combined with full auxiliary data such as ship position, equipment status, and navigation environment, it realizes the collaborative fusion of interference information of multiple ships and the correction of secondary misjudgments, and obtains the fused interference feature dataset.

[0031] Specifically, the fused interference feature dataset is obtained using the following method: S311: Each ship periodically broadcasts its computing power level, communication link signal-to-noise ratio, and online status, and calculates a comprehensive election score according to formula (1). The ship with the highest comprehensive election score is elected as the core node, responsible for aggregating the observation data of all ships as distributed observation nodes and uniformly executing all subsequent steps. The core node is dynamically determined according to the preset election cycle. If a ship of the core node loses contact for more than a set time, the remaining ships trigger an emergency re-election mechanism based on the comprehensive election score. (1); in: This represents the overall election score. This represents the computing power score. This represents the computing power score coefficient. Indicates the communication quality score. This represents the communication quality score coefficient; The preferred election period for core nodes is 30 minutes, and the preferred time for ships to lose contact is 5 seconds.

[0032] The computing power score is out of 100 points. It can be evaluated using a two-dimensional weighted approach of "general computing power + AI accelerated computing power". The weight of general computing power can be set to 60% and the weight of AI accelerated computing power to 40%, which is suitable for different levels of shipborne edge computing devices.

[0033] For general computing power: The FP32 floating-point performance or equivalent CPU computing power of the device is graded and quantified to adapt to mainstream general-purpose CPUs or GPUs. Specific scores are as follows: FP32 floating-point computing power ≥ 500 GFLOPS or 8-core or higher high-performance embedded CPU: General computing power score 100 points; 200 GFLOPS ≤ FP32 floating-point computing power < 500 GFLOPS or 4 to 8-core industrial-grade CPU: General computing power score 70 points; 50 GFLOPS ≤ FP32 floating-point computing power < 200 GFLOPS or 2 to 4-core ordinary embedded CPU: General computing power score 40 points; FP32 floating-point computing power < 50 GFLOPS or single-core or lower low-power CPU: General computing power score 10 points.

[0034] For AI acceleration computing power, the maximum score is 100 points. It is quantified by the peak computing power (TOPS) of INT8 based on the dedicated AI acceleration unit. The device without a dedicated AI chip gets 0 points for AI acceleration computing power; the peak computing power is ≥10 TOPS and the AI ​​acceleration computing power is 100 points; the peak computing power is ≤5 TOPS and the AI ​​acceleration computing power is 70 points; the peak computing power is ≤1 TOPS and the AI ​​acceleration computing power is 40 points; the peak computing power is <1 TOPS or there is no dedicated AI acceleration unit and the AI ​​acceleration computing power is 0 points.

[0035] If the device can only provide general computing power parameters and cannot provide AI acceleration computing power parameters, the AI ​​acceleration computing power item will be scored as 0 points. If the device can only provide AI acceleration computing power parameters and cannot obtain the nominal value of general computing power, the general computing power item will be recalculated according to the tier alignment principle: general computing power and AI acceleration computing power adopt a four-tier one-to-one correspondence hierarchical system. Based on the performance tier of AI computing power, it is directly mapped to the score of the same tier of general computing power, ensuring that all nodes can participate in the election. This recalculation rule is a fallback compatibility mechanism and is only activated when there are no explicit general computing power parameters.

[0036] Regarding the communication quality score: the full score is 100 points, which is composed of two sub-items: the signal-to-noise ratio of the communication link and the online status. The sub-weights can be preferably 70% and 30% respectively.

[0037] Communication link signal-to-noise ratio: Based on the average signal-to-noise ratio of the ad hoc network communication links between the ship and other surrounding ships, a linear quantization score is applied. Signal-to-noise ratio (SNR) ≥ 20dB: 100 points; 10dB ≤ SNR < 20dB: score calculated linearly as (SNR - 10) × 10; SNR < 10dB: 0 points.

[0038] Online status: Quantitatively scored based on the percentage of effective online time of the vessel within the last 30 minutes. Online rate = 100%, score 100 points; 95% ≤ online rate < 100%: score 60 points; 90% ≤ online rate < 95%: score 30 points; online rate < 90%: score 0 points.

[0039] When calculating the final comprehensive election score, the weights of the computing power score and the communication quality score can be preferably 0.4 and 0.6, respectively.

[0040] The remaining ships trigger an emergency reselection mechanism. During the reselection, the score is calculated based only on the real-time parameters of the currently online nodes to ensure that the fusion calculation is not interrupted.

[0041] S312: Based on the three dimensions of ship position reliability, receiver equipment accuracy, and data integrity, the original weight coefficient of interference data for each ship is calculated according to equation (2), and then the original weight coefficient is normalized by linear mapping to... The final weight range; (2); in: Indicates the first The original weighting coefficients of the ships, This represents the location reliability weighting coefficient. Indicates the first The ship The reliability coefficient of the ship's position. This represents the weighting coefficient for equipment accuracy. Indicates the first The accuracy coefficient of the ship's equipment This represents the data integrity weighting coefficient. Indicates the first Ship data integrity coefficient; , , The values ​​can be preferably 0.4, 0.3, and 0.3, respectively, and the original weight coefficients are normalized to 0.3 by linear mapping. The final weight range ensures that the original weight coefficients are no more than 0.9 and no less than 0.1, while preserving the relative proportion of each ship's weight.

[0042] Location reliability coefficient The value ranges from 0.2 to 1. The shipborne edge node reads three types of parameters in real time from the positioning solution message of the GNSS receiver: horizontal position error estimate (HPE, 95% confidence level), horizontal accuracy factor (HDOP), and number of effective tracked satellites. The value is determined according to the following priority rules: 1. Priority Decision Rule: The horizontal position error estimate (HPE) is used as the core criterion for decision-making. When HPE ≤ 10m, =1.0; when 10m < HPE ≤ 50m, =0.5; when HPE>50m, =0.2.

[0043] 2. Alternative Judgment Rule: If the receiver does not output HPE parameters, then HDOP and the number of effectively tracked satellites will be used as the alternative criteria. When HDOP ≤ 2.5 and the number of effectively tracked satellites ≥ 8, =1.0; when 2.5 < HDOP ≤ 4.0 and the number of effectively tracked satellites is ≥ 6, =0.5; when HDOP>4.0 or the number of effectively tracked satellites<6, =0.2.

[0044] 3. Conflict handling rules: If two types of parameters are available simultaneously and the judgment results are inconsistent, the higher value shall be taken. The value is used as the final value to avoid undue weighting of high-precision observation data.

[0045] 4. Weighting Correction for Spoofing and Interference Scenarios: When a suspected spoofing or interference event is identified, the above-mentioned weighting based on the receiver's own parameters is adjusted. The initial value is only used as a reference; dynamic correction is required by introducing spatial consistency of multi-ship positioning deviations: calculate the group mean of all ship positioning deviation vectors; the larger the magnitude of the difference between a single ship's deviation vector and the group mean, the better. The larger the correction magnitude, the more likely it is that when the difference modulus is greater than 50m, The value was adjusted downwards by 0.3-0.5 to prevent deceived ships from gaining excessive weight due to false high accuracy.

[0046] The parameters are updated synchronously with the GNSS positioning results, and the update frequency is consistent with the receiver positioning output frequency (≥1Hz).

[0047] Equipment accuracy coefficient These are static, inherent parameters that are calibrated once by the receiver hardware and are not dynamically adjusted during operation: 1 for surveying-grade receivers, 0.6 for industrial-grade receivers, and 0.3 for consumer-grade receivers.

[0048] S313: Introduce spatial consistency of multi-ship positioning deviation to dynamically correct the position reliability coefficient and update it synchronously with the positioning results of satellite navigation receiver. According to formula (3), the data integrity coefficient is updated synchronously with the fusion calculation cycle, and the weight coefficient is updated synchronously. (3); in: Indicates the duration of valid data. Indicates the total observation duration; Data integrity coefficient The value ranges from 0 to 1 and is used to characterize the continuity and reliability of the observation data. The effective data duration and total observation duration can be determined according to the following rules: 1. Statistical window: Taking the current fusion time as the endpoint, a sliding time window with a length of 30 seconds is taken as the statistical interval, and the total observation time is the theoretical duration of the window, 30 seconds; 2. Valid data duration: The cumulative duration of data within the window that simultaneously meets all of the following conditions: Invalid data filtering and 3σ outlier removal were achieved; the data was not marked as a suspected occlusion misjudgment after initial screening by single-ship level misjudgment; the data packet loss rate within the time period was ≤20%; the receiver was in normal tracking state with no restart or self-test error reports.

[0049] Update and adaptation rules: Under normal circumstances Updated synchronously with the fusion computing cycle (calculated every 1 second); for a single independent interference event, a dedicated window from the start time of the interference to the current time is used to calculate the event level. The value is used for weight allocation during the interference source localization phase.

[0050] S314: The preprocessed synchronous interference data of multiple ships are correlated and fused according to the time and spatial dimensions. Based on the updated weight coefficients, the fused interference signal strength of the multi-ship observations is calculated according to equation (4): (4); in: This indicates the intensity of the fusion interference signal from multiple ship observations. Indicates the first The strength of the interference signal measured by the ship Indicates the total number of ships; Specifically, the system first focuses on the ship that first detects strong interference (Received Signal Strength, RSS > -100dBm). Based on the logarithmic distance path loss model at sea, the effective coverage area of ​​the interference is estimated. All ships within the coverage area are included in the voting base, while ships far from the interference source and naturally free of interference signals are excluded to avoid distortion of the voting base. For a candidate interference event, if more than half of the ships in the voting base report interference signals of the same frequency (frequency deviation ≤ 1kHz, based on the GNSS L1 band signal bandwidth of 2.046MHz, a deviation of 1kHz can be considered as the same interference source) within a preset time window (e.g., when the ship speed is ≤ 30 knots, the displacement within 100ms is less than 0.5m, which can be considered as observations at the same spatial location), then the event is determined to be a real interference event. The minimum number of ships in the voting base is 3. When the number of ships within the coverage area is < 3, the multi-ship voting condition is not met, and the system is directly downgraded to a single-ship detection and early warning mode, without performing collaborative fusion judgment.

[0051] S315: Construct a misjudgment correction model based on the supervised machine learning binary classification algorithm CART decision tree, and input the feature vector containing the intensity of fused interference signals from multiple ship observations into the misjudgment correction model for the second-level misjudgment correction, and output the binary classification result containing real interference or misjudged interference. The input feature vector of the misjudgment correction model consists of 4 categories and 10 dimensions. All continuous features are Z-score standardized, and discrete features are one-hot encoded: Interference feature category (4 dimensions): fused interference signal strength, interference center frequency, interference duration, and signal modulation method (discrete encoding: continuous wave = 0, narrowband modulation = 1, wideband modulation = 2); Ship consistency category (1 dimension): the percentage of ships that detected the interference within the voting base; Navigation status category (2 dimensions): the average roll angle and speed change rate of ships within the observation window; Environmental feature category (3 dimensions): sea state level, nearshore distance, and atmospheric visibility.

[0052] The misclassification correction model is trained using labeled misclassification cases such as multipath effects and atmospheric attenuation, along with real interference cases, from historical navigation data. Model parameters include a maximum tree depth of 6, a Gini coefficient as the splitting criterion, a training set size of at least 1000 samples, and a 1:1 ratio of positive to negative samples. After training, cost complexity pruning (CCP) is used for post-pruning, and the optimal pruning parameters are selected through 10-fold cross-validation to improve the model's generalization ability. After training, the fused interference feature dataset is corrected to remove misclassification events caused by complex natural factors.

[0053] If the misjudged interference result output by the misjudged correction model corresponds to signal anomalies caused by natural factors such as multipath effects, ship hull obstruction, and atmospheric attenuation, it is determined to be a natural interference event. It will not enter the subsequent interference source location and assessment process, but will only be marked as a cause and archived to the local log and shore-based platform for environmental statistics and model optimization.

[0054] S316: Perform consistency verification on the fused interference signal strength observed by multiple ships. If the deviation between the fused interference signal strength observed by multiple ships corresponding to the same interference source and the interference signal strength measured by a single ship is less than or equal to 3dB, it is determined that the consistency verification requirement is met, and there is no need to adjust the weight coefficient of the ship. If the deviation between the fused interference signal strength observed by multiple ships corresponding to the same interference source and the interference signal strength measured by a single ship is greater than 3dB, it is determined that the consistency verification requirement is not met. In this case, an adjustment coefficient is generated according to the inverse proportional rule that the larger the deviation, the lower the weight. The original weight coefficient of the ship is corrected and normalized to obtain the fused interference feature dataset.

[0055] Specifically, if the consistency verification requirements are not met, the original weight coefficients are iteratively adjusted based on the observation residuals, and the fusion calculation is re-performed: using the current fusion result as a benchmark, the signal strength residuals (dB values) of each vessel are calculated, and adjustment coefficients are generated according to the inverse proportional rule of "the larger the deviation, the lower the weight," correcting the original weight coefficients and normalizing them. The single weight adjustment range does not exceed 60% of the original weight; the weight adjustment is performed a maximum of 3 times, and consistency is re-verified after each adjustment. If the verification requirements are still not met after 3 adjustments, the outlier vessel with the largest residual is removed, and the remaining nodes are fused again; if there are fewer than 3 remaining nodes, the system is downgraded to a single-vessel detection and early warning mode, the event is marked as pending confirmation, and pushed to the shore-based platform for manual review.

[0056] S4: Based on the fused interference feature dataset, the interference subtype subdivision, multi-interference source separation, single-source precise localization, and multi-dimensional parameter evaluation are completed in sequence to finally obtain the location, multi-dimensional parameter evaluation and comprehensive classification results of each interference source. Furthermore, the method for obtaining the location of each interference source, multi-dimensional parameter evaluation, and comprehensive classification results in step S4 is as follows: S411: Based on the binary classification results and the fused interference feature dataset, the confirmed real interference events are further subdivided into human interference subtypes and natural interference subtypes; First, the sources of interference are broadly categorized into human-caused interference and natural interference.

[0057] The core basis for identifying human interference is that the interference observed by multiple ships within the coverage area has a unified center frequency and modulation method, the signal strength changes with distance in accordance with the attenuation law of maritime propagation, and there is no significant correlation with environmental parameters such as sea state and ship attitude. The interference is sudden.

[0058] The core basis for natural interference is that the interference has no uniform frequency and modulation characteristics, is highly correlated with environmental factors such as near-shore distance, sea state level, and ship hull obstruction, and only local ships experience abnormalities, without large-scale spatial consistency.

[0059] Artificial interference subtypes include: Suppressing interference: The carrier-to-noise ratio decreases uniformly across the entire frequency band with a decrease slope ≥ 5dB / s; multiple ships simultaneously exhibit signal attenuation at the same frequency; and the intensity distribution matches the propagation model. Generative deception interference: The carrier-to-noise ratio does not decrease significantly but the positioning suddenly changes. The spatial correlation of positioning deviations of multiple ships is weak, the magnitude of the deviation difference is >50m and the spatial correlation coefficient is <0.3. Forwarding deception interference: The carrier-to-noise ratio does not decrease significantly, the positioning error increases slowly over time, and the error distribution has a gradient relationship with the distance of the interference source.

[0060] Natural disturbances are further divided into three categories: a) Multipath interference: The carrier-to-noise ratio fluctuates periodically, and the positioning oscillates slightly. It only occurs in near-shore areas and is strongly correlated with the distance to the near shore. b) Hull obstruction interference: The carrier-to-noise ratio fluctuates synchronously with the roll and pitch angles, and only occurs on a single ship or in some ships. It is strongly correlated with the antenna elevation angle and the ship's attitude. c) Atmospheric attenuation interference: The carrier-to-noise ratio decreases gradually across the entire frequency band, and a large number of ships experience uniform attenuation synchronously, which is strongly correlated with atmospheric visibility and meteorological conditions.

[0061] The following applies only to the location and assessment of events determined to be genuine human interference; natural interference events will be further subdivided into specific causes and archived for environmental statistics and model optimization, and will not be included in subsequent calculations. Subdividing human interference subtypes can support subsequent differentiated location and assessment.

[0062] S412: Based on the fused interference feature dataset, multiple interference sources existing simultaneously in the sea area are separated and identified using a two-stage mechanism of hard frequency differentiation before positioning and fine differentiation of coarse frequency clustering after coarse positioning. After separation and identification, the observation dataset is split according to clusters so that each cluster corresponds to an independent co-frequency interference source. Specifically, the first stage is pre-location frequency hard differentiation: based on the frequency domain features of the fused interference feature dataset, interference with a center frequency deviation of more than 5kHz or with different modulation methods is directly identified as an independent interference source.

[0063] The second stage is coarse localization followed by fine-grained clustering with the same frequency: For co-frequency aliasing interference that cannot be directly separated in the first stage, unconstrained coarse localization is first solved based on the nonlinear localization equations to obtain discrete candidate location point clouds; then, a clustering algorithm is used to spatially cluster the point clouds to identify multiple independent interference sources: when the number of observation nodes is <5, the DBSCAN density clustering algorithm is used, with a neighborhood radius of 500m and a minimum sample size of 2, without the need to pre-specify the number of clusters, which is suitable for small sample scenarios; when the number of observation nodes is ≥5, the K-means clustering algorithm is used, with the planar coordinates of the candidate locations as the core feature, combined with signal strength and the time of interference occurrence as auxiliary constraints, and the K value is automatically determined by the elbow rule, with a value range of 2-10.

[0064] Separating and identifying multiple interference sources that coexist in the sea area can provide accurate individual datasets for subsequent independent positioning and evaluation, avoiding the distortion of positioning and evaluation results caused by the superposition of multiple source signals.

[0065] S413: For each independent interference source after separation, if the human interference subtype is suppression interference, the Time Difference of Arrival (TDOA) technique is adopted. The ship that first receives the interference signal is selected as the reference ship. Based on the time difference of arrival between the reference ship and other ships, a set of nonlinear positioning equations is established as Equation (5). The Chan algorithm is used to solve the set of nonlinear positioning equations to obtain the final preliminary position coordinates of the suppression interference source. Then, the mean-shift clustering algorithm is used to cluster multiple sets of final preliminary position coordinates, remove outliers, and take the cluster center as the final positioning result of the suppression interference source. (5); in: This indicates the time difference between the arrival of the reference vessel and other vessels. Represents the speed of light in a vacuum. Indicates the interference source to the first The distance between the ships Indicates the distance from the interference source to the reference vessel; Chan's algorithm is a classic two-step weighted least squares localization algorithm. Its positioning accuracy in Gaussian noise environments can approach the lower bound of the Cramer-Rao algorithm, making it suitable for TDOA localization scenarios with suppressed maritime interference. The specific solution steps are as follows: 1. Variables and system of equations construction: Let the true location coordinates of the interference source be... The ship that first detected the interference signal was selected as the reference node. (Number 1, coordinates are) The remaining N-1 ships are the observation node numbers ( ,coordinate ), No. The time difference between the arrival of the interference signals between the node and the reference node is Corresponding distance difference for ,in The speed of light in a vacuum is 299,792,458 m / s. Based on the geometric relationship of the distance difference, N-1 hyperbolic equations can be constructed, forming a nonlinear positioning equation system.

[0066] 2. First step: Weighted Least Squares (WLS) to find the initial position: Each nonlinear distance difference equation in the nonlinear positioning equation system is algebraically transformed into a function of unknown quantities. ( Given a pseudo-linear system of equations (where T is the distance from the interference source to the reference node and T is the matrix transpose), construct a measurement error vector. The diagonal elements of the weighting matrix are the reciprocals of the variances of each TDOA measurement. By minimizing the sum of squared weighted residuals, the initial estimate of the interference source location and its estimated covariance matrix are obtained.

[0067] 3. Second step: Weighted least squares correction optimization By utilizing the error correlation of the first step estimation result, a constraint relationship of the position estimation error is introduced to construct a second pseudo-linear equation system; the weighting matrix is ​​updated to the inverse of the first step estimation covariance matrix, and the position correction is solved again by weighted least squares to obtain the preliminary positioning result in closed-form solution.

[0068] 4. Iterative Refinement and Scene Adaptation To further improve positioning accuracy in dynamic scenarios, an iterative weighted least squares approach is introduced to refine the solution based on the two-step solution of the Chan algorithm. The iteration termination condition is: the position correction magnitude value of this iteration is less than 1m, or the number of iterations reaches the maximum limit of 15. For ship dynamic scenarios, the positions of each ship are first interpolated to the same reference time before being substituted into the solution to eliminate motion errors. After the iteration converges, the final preliminary position coordinates of the interference source are output.

[0069] After obtaining the initial location coordinates of the interference source, the mean-shift clustering algorithm is used to cluster 10 consecutive sets of location results. Outliers are removed, and the cluster centers are taken as the final location results. The bandwidth parameter of the mean-shift algorithm is set to 30m. Finally, a heatmap of the interference source location can be generated, which visually displays the interference distribution.

[0070] S414: If the human interference subtype is deception interference, then read the deception signal pseudo-code phase observation value output by the tracking loop of each ship's satellite navigation receiver, eliminate the integer ambiguity of the deception signal pseudo-code phase observation value by combining chip period constraints with coarse time determination, obtain the precise distance difference between the two ships after ambiguity elimination, and then use the Chan algorithm to solve the nonlinear positioning equation set to obtain the final preliminary position coordinates of the deception interference source. For ships equipped with dual-antenna direction finding, the final preliminary position coordinates are then corrected by dual-antenna direction finding cross-positioning to obtain the corrected deception interference source positioning result. The method for eliminating integer ambiguity in the phase observations of spoofing signals by combining chip period constraints with coarse time determination is as follows: The pseudocode of GNSS spoofing signals is periodic. Taking GPS L1 C / A code as an example, the chip rate is 1.023 Mcps, the code period is 1 ms, and the spatial propagation distance of a single code period is approximately 299.8 km. The pseudocode phase observation value output by the receiver code tracking loop is the relative phase after modulo-taking for a single code period, and does not contain complete cycle count information. If the distance difference between the two ships and the interference source exceeds the distance of a single code period, integer cycle ambiguity (i.e., distance difference) will exist. ),in It is an integer number of weeks. For the normalized code phase difference, (Code period). When eliminating integer ambiguity by combining code period constraints with coarse time determination, the distance difference is first roughly estimated based on the first detection time difference of the interference signal. The coarse estimate is divided by the single-code cycle distance and rounded down to obtain the nominal value of the integer cycle number. The fine code phase difference is then converted into a sub-chip level fine distance difference, which is superimposed on the integer cycle distance to obtain an unambiguous, precise distance difference. Since the distance between cooperating vessels in the same sea area is typically less than 300 km, the integer cycle number ranges from 0 to 1, requiring only one coarse time determination to eliminate ambiguity.

[0071] Based on the unambiguous and precise distance difference, the Chan algorithm, which is of the same origin as the suppression of interference, is used to solve the nonlinear positioning equations to obtain the final preliminary position coordinates.

[0072] The method of using the Chan algorithm to solve the nonlinear localization equations is the same as the method for suppressing interference.

[0073] The method for correcting the final preliminary position coordinates using dual-antenna direction finding cross-positioning is as follows: Ships equipped with dual-antenna direction finding systems supplement the azimuth angle of arrival for cross-positioning correction. The shipborne dual GNSS antennas are deployed laterally along the ship with a baseline length ≥1m. The azimuth angle of arrival of the deception signal is obtained through carrier phase interferometry direction finding technology, with a direction finding accuracy better than 1. The preliminary position result of TDOA is fused with the azimuth ray result using weighted least squares to obtain the corrected deception interference source positioning result. Ships without direction finding equipment complete positioning only through pseudocode phase difference TDOA.

[0074] S415: Based on the final location result of the suppressed interference source or the corrected location result of the deceiving interference source, as well as the observation vessel set and feature dataset of each independent interference source, conduct a multi-dimensional hierarchical evaluation and comprehensive classification of each independent interference source. First, suitable parameter evaluation systems can be constructed for both suppression and deception interference, and the parameters of suppression and deception interference can be evaluated. Then, using the final location result of the suppression interference source or the corrected location result of the deception interference source, as well as the observation vessel set and feature dataset of each independent interference source as input, an evaluation is performed on each independent interference source individually.

[0075] Calculation of specific parameters for suppressing interference: Transmission power inversion: The system of maritime signal propagation attenuation model is called up and the calculation is performed in different scenarios, namely far field and near field.

[0076] For open-line-of-sight far-field scenarios (distance between interference source and ship > 1km), a logarithmic distance path loss model considering seawater reflection and atmospheric attenuation is adopted. For the GNSS L1 band (1575.42MHz), the path loss is calculated using the following formula: ; in: For the distance of transmission Path loss at the location, in dB; For reference distance Path loss at reference distance The optimal range is 1km. The corresponding reference propagation loss can be calculated using the free space loss formula or determined through on-site calibration, making it suitable for open line-of-sight scenarios at sea. The path loss index is 2.2 for open-line-of-sight scenarios at sea. This value is based on the fact that multipath propagation caused by seawater reflection results in a path loss slightly greater than that in free space. =2; The shadow fading follows a normal distribution with a mean of 0 and a standard deviation of 3dB.

[0077] For near-field scenes (distance ≤ 1km), sea surface reflection correction is added, and a two-path propagation model is used instead of a single logarithmic distance model to improve the accuracy of near-field inversion.

[0078] Path loss is a core intermediate quantity in transmit power inversion: it is combined with the received signal strength measured by each ship. Corresponding path loss The interference source's transmit power is inversely calculated using the wireless link budget relationship; multiple sets of observations from multiple ships are used to eliminate random errors through least squares fitting, and finally the equivalent isotropic radiated power (EIRP) of the interference source is obtained. This parameter is the sole input for subsequent power intensity level assessment.

[0079] Effective radius calculation: Using the location coordinates of the interference source as the center and the interference sensitivity threshold (-130dBm) of the GNSS receiver as the boundary, the critical distance is solved in reverse based on the above propagation loss model to obtain the effective radius of action for suppressing the interference. This parameter is the input for subsequent impact range level assessment.

[0080] Deception and interference-specific parameter calculation: Positioning deviation magnitude assessment: This involves statistically analyzing the mean, peak, and rate of change of the two-dimensional positioning deviation of the affected vessel during the duration of the interference. Combined with the temporal evolution characteristics of the deviation, this helps distinguish between generative deception interference and forwarding deception interference. This parameter is the core input for subsequent deviation severity level assessment.

[0081] Effective radius estimation: Since the power of the deceptive jamming signal is close to that of the real satellite signal, the coverage area cannot be inverted from the received signal strength. Therefore, the radius is estimated based on fitting the spatial attenuation law of positioning deviation using multi-ship observation data. The specific process is as follows: calculate the planar distance from each observation ship to the jamming source. Match the average positioning deviation of the corresponding ship Multiple sets of distance-bias observation samples were obtained; a log-linear decay model was adopted. The attenuation characteristics of positioning error with distance are characterized, and the parameters of the logarithmic linear attenuation model are obtained by fitting using the least squares method. and Using "positioning deviation exceeding 10m" as the effective disturbance threshold, the effective radius of the deception interference is obtained by reverse calculation. : ,in: This indicates the effective location deviation threshold for deception interference.

[0082] If there are fewer than 3 observation samples, the estimation is downgraded to two-point linear interpolation; if the deviation of all samples is higher or lower than the threshold, the distance between the farthest or nearest observation point is used as the boundary estimate. This parameter is the input for subsequent impact range level assessment.

[0083] Source tracing accuracy characterization: Based on the fusion positioning results of pseudocode phase difference TDOA and dual-antenna direction finding, the 95% confidence interval of the interference source positioning results is given, characterizing the reliability of the source tracing results.

[0084] Calculation of general interference parameters: Trajectory and velocity fitting: Using the location results of the interference source at continuous time points, after Kalman filtering and smoothing, the trajectory of the interference source is fitted, and its real-time velocity and direction of motion are calculated. This parameter serves as an auxiliary input for comprehensive hazard assessment.

[0085] Based on the quantitative parameters calculated above, combined with maritime navigation safety thresholds, a graded assessment is completed from three dimensions, ultimately aggregated into a comprehensive hazard level, and corresponding response strategies are matched accordingly. a) Power intensity level (specific for suppressing interference) Mild threat: Equivalent omnidirectional radiated power <10W, affecting ships only at close range; Moderate threat: Equivalent isotropic radiated power 10-100W, which can affect a range of several kilometers; Severe threat: Equivalent omnidirectional radiated power >100W, which can affect a range of tens of kilometers.

[0086] b) Level of Impact Local impact: Effective radius of action < 5km, only a small area of ​​sea will be disturbed; Regional impact: Effective radius of action 5-20km, covering local waterway areas; Wide-area impact: Effective radius of action > 20km, covering a large area of ​​navigable waters.

[0087] c) Level of deviation Slight deviation: The average positioning deviation is less than 10m within 10 consecutive seconds, which does not affect the basic positioning function; Moderate deviation: The average positioning deviation is 10-50m within 10 consecutive seconds, and the positioning accuracy decreases significantly; Severe deviation: The average positioning deviation is greater than 50m within 10 consecutive seconds or the signal lock is lost, and the positioning function fails.

[0088] d) Comprehensive Hazard Level Classification Rules The classification results of the three dimensions are aggregated using the principle of "choosing the higher standard": if any severe condition is met, it is judged as severe hazard; if it does not meet severe condition but meets any moderate condition, it is judged as moderate hazard; if all dimensions are mild, it is judged as mild hazard.

[0089] Minor hazard: Does not affect normal navigation; only requires event logging and local alerts. Moderate hazard: Drivers need to increase their vigilance, switch to alternative navigation methods, and continuously monitor changes in interference. Severe hazard: Immediate evasive action is required, including adjusting flight paths to move away from the interference area and simultaneously reporting to the shore-based platform.

[0090] This invention uses multiple ships in the same sea area as distributed observation nodes, and constructs a distributed interference monitoring network of "ship-end edge computing + multi-ship collaborative fusion". It improves detection accuracy through a two-level misjudgment correction mechanism, designs appropriate detection and location algorithms for suppressing interference and deceptive interference respectively, and conducts multi-dimensional evaluation of interference sources in combination with the maritime propagation model. Without increasing hardware costs, it breaks through the limitations of single-device observation and comprehensively improves the anti-interference capability of the maritime satellite navigation system.

[0091] The assessment results of a single interference source are packaged into a structured dataset and simultaneously pushed to shipboard terminals and shore-based platforms as the basis for decision-making in early warning classification, route planning, and maritime supervision.

[0092] S416: Package the multi-dimensional hierarchical assessment and comprehensive grading results into a unified structured dataset and push it synchronously to the shipborne terminal and shore-based platform.

[0093] S5: Output the location, multi-dimensional parameter classification and comprehensive rating results of each interference source in a standardized format and implement them in practice.

[0094] Specifically, interference detection results, the location of each interference source, multi-dimensional hierarchical evaluation and comprehensive classification results can be output in a standardized JSON format.

[0095] The optimized step S5 outputs the interference detection results, the locations of each interference source, and the multi-dimensional hierarchical evaluation and comprehensive classification results in a standardized format, including three output methods: Shipborne terminal output: Overlay the location of the interference source, multi-dimensional hierarchical assessment and comprehensive classification results, and coverage area onto the ship's electronic nautical chart, and issue warnings to the crew in the form of sound and light, and recommend avoidance routes; Shore-based platform output: Summarizes all interference events in the sea area, generates a regional interference situation map, and supports historical interference data backtracking and statistical analysis; Data interface output: Provides a standard REST API interface, supporting integration with Automatic Identification System (AIS), Vessel Traffic Management System (VCR), and Vessel Navigation System (GNSS).

[0096] A multi-ship satellite navigation receiver jamming information integrated processing system is used to execute a multi-ship satellite navigation receiver jamming information integrated processing method as described in any one of the above, which includes a data acquisition module, a preprocessing module, a multi-source information collaborative fusion module, a jamming source localization and evaluation module, and a result output module; The data acquisition module is used to select multiple ships sailing in the same sea area as distributed observation nodes, and to collect the raw interference data of each ship's satellite navigation receiver and auxiliary data including the ship's real-time position coordinates, navigation status data, environmental data and receiver equipment parameters in real time. The preprocessing module is used to preprocess the raw interference data and auxiliary data of each ship's satellite navigation receiver, and at the same time complete the spatiotemporal synchronization of the raw interference data and auxiliary data to obtain preprocessed synchronization interference data and preprocessed synchronization auxiliary data. The multi-source information collaborative fusion module is used to take the pre-processed synchronization interference data as the core processing object, and combine it with the pre-processed synchronization auxiliary data to perform collaborative fusion and misjudgment correction of multi-ship interference information, so as to obtain the fused interference feature dataset. The interference source localization and assessment module is based on the fused interference feature dataset, with auxiliary data such as ship position, navigation status, and equipment parameters as the calculation support. Combined with a unified marine signal propagation attenuation model system, it performs interference detection and multi-dimensional assessment of interference sources by type, and finally obtains the location, multi-dimensional hierarchical assessment and comprehensive classification results of each interference source. The result output module is used to output the interference detection results, the location of each interference source, the multi-dimensional hierarchical evaluation and comprehensive classification results in a standardized format, and to implement them in practice.

[0097] In summary, the present invention provides a method and system for comprehensive processing of interference information from multiple ship satellite navigation receivers. This method and system can automatically transform users' personalized needs into manufacturable patterns and process solutions. Within a micro-factory, it enables autonomous collaborative execution and closed-loop optimization of multiple processes and equipment, including cutting, AGV handling, multi-station sewing, ironing and finishing, and quality inspection. This solves problems in existing customized production such as "fragmented links from demand to design to manufacturing, reliance on manual scheduling for cross-process collaboration, high adaptation costs due to differences in equipment capabilities, and lack of closed-loop linkage between quality inspection and process parameters." As a result, it significantly improves the delivery efficiency, process stability, and finished product consistency of single or small-batch customized jeans, while reducing reliance on manual labor and rework rates.

[0098] The above description is merely a preferred embodiment of the present invention and is not intended to limit the invention. Various modifications and variations can be made to the present invention by those skilled in the art. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the scope of protection of the present invention.

Claims

1. A method for comprehensively processing interference information from multiple shipborne satellite navigation receivers, characterized in that: Includes the following steps: S1: Select multiple ships sailing in the same sea area as distributed observation nodes, and collect the raw interference data of each ship's satellite navigation receiver in real time, as well as auxiliary data including the ship's real-time position coordinates, navigation status data, environmental data and receiver equipment parameters. S2: Preprocess the raw interference data and auxiliary data collected from each ship's satellite navigation receiver, and simultaneously complete the spatiotemporal synchronization of the raw interference data and auxiliary data to obtain preprocessed synchronization interference data and preprocessed synchronization auxiliary data. S3: Taking the preprocessed synchronization interference data as the core processing object, and combining it with the preprocessed synchronization auxiliary data, the collaborative fusion of multi-ship interference information and misjudgment correction are carried out to obtain the fused interference feature dataset. S4: Based on the fused interference feature dataset, the interference subtype subdivision, multi-interference source separation, single-source precise localization, and multi-dimensional parameter evaluation are completed in sequence to finally obtain the location, multi-dimensional parameter evaluation and comprehensive classification results of each interference source. The methods for obtaining the location, multi-dimensional parameter evaluation, and comprehensive classification results of each interference source are as follows: S411: Based on the binary classification results and the fused interference feature dataset, the confirmed real interference events are further subdivided into human interference subtypes and natural interference subtypes; S412: Based on the fused interference feature dataset, multiple interference sources existing simultaneously in the sea area are separated and identified using a two-stage mechanism of hard frequency differentiation before positioning and fine differentiation of coarse frequency clustering after coarse positioning. After separation and identification, the observation dataset is split according to clusters so that each cluster corresponds to an independent co-frequency interference source. S413: For each independent interference source after separation, if the human interference subtype is suppression interference, the time difference of arrival technique is adopted, and the ship that receives the interference signal first is selected as the reference ship. Based on the time difference of arrival between the reference ship and other ships, a nonlinear positioning equation set is established as Equation (5). The Chan algorithm is used to solve the nonlinear positioning equation set to obtain the final preliminary position coordinates of the suppression interference source. Then, the mean-shift clustering algorithm is used to cluster multiple sets of final preliminary position coordinates, remove outliers, and take the cluster center as the final positioning result of the suppression interference source. （5）； in: This indicates the time difference between the arrival of the reference vessel and other vessels. Represents the speed of light in a vacuum. Indicates the interference source to the first The distance between the ships Indicates the distance from the interference source to the reference vessel; S414: If the human interference subtype is deception interference, then read the deception signal pseudo-code phase observation value output by the tracking loop of each ship's satellite navigation receiver, eliminate the integer ambiguity of the deception signal pseudo-code phase observation value by combining chip period constraints with coarse time determination, obtain the precise distance difference between the two ships after ambiguity elimination, and then use the Chan algorithm to solve the nonlinear positioning equation set to obtain the final preliminary position coordinates of the deception interference source. For ships equipped with dual-antenna direction finding, the final preliminary position coordinates are then corrected by dual-antenna direction finding cross-positioning to obtain the corrected deception interference source positioning result. S415: For each independent interference source, perform multi-dimensional parameter classification and comprehensive classification based on the final location result of the suppressed interference source or the corrected location result of the deceitful interference source and the split observation dataset; S416: Package the multi-dimensional parameter classification evaluation and comprehensive classification results into a structured dataset and push it synchronously to the shipborne terminal and shore-based platform. S5: Output the location of each interference source, multi-dimensional parameter evaluation and comprehensive classification results in a standardized format, and implement them in practice.

2. The method for comprehensively processing interference information from multiple ship satellite navigation receivers according to claim 1, characterized in that: The raw interference data collected in step S1 for each ship's satellite navigation receiver covers two categories: suppression interference features and deception interference features.

3. The method for comprehensively processing interference information from multiple shipborne satellite navigation receivers according to claim 1, characterized in that: The navigation status data mentioned in step S1 includes speed, heading, roll and pitch angles; environmental data includes sea state, atmospheric visibility, and near-shore distance; and receiver equipment parameters include antenna gain, noise figure, and equipment accuracy level.

4. The method for comprehensively processing interference information from multiple ship satellite navigation receivers according to claim 1, characterized in that: In step S1, the acquisition frequency of the original interference data is ≥10Hz, and the acquisition frequency of the auxiliary data is ≥1Hz.

5. The method for comprehensively processing interference information from multiple ship satellite navigation receivers according to claim 1, characterized in that: The method for preprocessing the raw interference data and auxiliary data collected from the satellite navigation receivers of each ship in step S2 is as follows: S211: During normal, interference-free periods, if the data within the window follows a normal distribution, the threshold method combined with the Laida criterion based on the 3σ principle is used for outlier detection. If the data within the window does not follow a normal distribution, the interquartile range method is used for outlier detection. If the standard deviation of the signal strength within the calculation window exceeds the preset threshold range, the original data of interference is judged as outlier data and removed. During periods of suspected interference, when the signal strength or carrier-to-noise ratio is determined to be a deceptive interference feature by the preliminary discrimination rules, outlier removal is suspended, and only extreme invalid values ​​caused by hard faults in the filtering equipment are removed. S212: Standardize the interference data from different models of satellite navigation receivers, and convert the signal strength, timestamp, and coordinate system into a standard format to obtain standardized interference data; S213: Perform initial screening based on the auxiliary data of a single ship. If any item in the auxiliary data exceeds the preset threshold, mark the data for that period as a suspected occlusion misjudgment and remove it, thus completing the first level of misjudgment correction.

6. The method for comprehensively processing interference information from multiple ship satellite navigation receivers according to claim 1, characterized in that: In step S3, the fused interference feature dataset is obtained using the following method: S311: Each ship periodically broadcasts its computing power level, communication link signal-to-noise ratio, and online status, and calculates a comprehensive election score according to formula (1). The ship with the highest comprehensive election score is elected as the core node, responsible for aggregating the observation data of all ships as distributed observation nodes and uniformly executing all subsequent steps. The core node is dynamically determined according to the preset election cycle. If a ship of the core node loses contact for more than a set time, the remaining ships trigger an emergency re-election mechanism based on the comprehensive election score. （1）； in: This represents the overall election score. This represents the computing power score. This represents the computing power score coefficient. Indicates the communication quality score. This represents the communication quality score coefficient; S312: Based on the three dimensions of ship position reliability, receiver equipment accuracy, and data integrity, the original weight coefficient of interference data for each ship is calculated according to equation (2), and then the original weight coefficient is normalized by linear mapping to... The final weight range; （2）； in: Indicates the first The original weighting coefficients of the ships, This represents the location reliability weighting coefficient. Indicates the first The reliability coefficient of the ship's position. This represents the weighting coefficient for equipment accuracy. Indicates the first The accuracy coefficient of the ship's equipment This represents the data integrity weighting coefficient. Indicates the first Ship data integrity coefficient; S313: Introduce spatial consistency of multi-ship positioning deviation to dynamically correct the position reliability coefficient and update it synchronously with the positioning results of satellite navigation receiver. According to formula (3), the data integrity coefficient is updated synchronously with the fusion calculation cycle, and the weight coefficient is updated synchronously. （3）； in: Indicates the duration of valid data. Indicates the total observation duration; S314: The preprocessed synchronous interference data of multiple ships are correlated and fused according to the time and spatial dimensions. Based on the updated weight coefficients, the fused interference signal strength of the multi-ship observations is calculated according to equation (4): （4）； in: This indicates the intensity of the fusion interference signal from multiple ship observations. Indicates the first The strength of the interference signal measured by the ship Indicates the total number of ships; S315: Construct a misjudgment correction model based on the supervised machine learning binary classification algorithm CART decision tree, and input the feature vector containing the intensity of fused interference signals from multiple ship observations into the misjudgment correction model for the second-level misjudgment correction, and output the binary classification result containing real interference or misjudged interference. S316: Perform consistency verification on the fused interference signal strength observed by multiple ships. If the deviation between the fused interference signal strength observed by multiple ships corresponding to the same interference source and the interference signal strength measured by a single ship is less than or equal to 3dB, it is determined that the consistency verification requirement is met, and there is no need to adjust the weight coefficient of the ship. If the deviation between the fused interference signal strength observed by multiple ships corresponding to the same interference source and the interference signal strength measured by a single ship is greater than 3dB, it is determined that the consistency verification requirement is not met. In this case, an adjustment coefficient is generated according to the inverse proportional rule that the larger the deviation, the lower the weight. The original weight coefficient of the ship is corrected and normalized to obtain the fused interference feature dataset.

7. The method for comprehensively processing interference information from multiple ship satellite navigation receivers according to claim 5, characterized in that: Any of the auxiliary data exceeding the preset threshold includes roll angle > 15°, antenna elevation angle < 5°, and near-shore distance < 1km.

8. The method for comprehensively processing interference information from multiple ship satellite navigation receivers according to claim 1, characterized in that: Step S5 outputs the interference detection results, the locations of each interference source, and the multi-dimensional hierarchical evaluation and comprehensive classification results in a standardized format, including three output methods: Shipborne terminal output: Overlay the location of the interference source, multi-dimensional hierarchical assessment and comprehensive classification results, and coverage area onto the ship's electronic nautical chart, and issue warnings to the crew in the form of sound and light, and recommend avoidance routes; Shore-based platform output: Summarizes all interference events in the sea area, generates a regional interference situation map, and supports historical interference data backtracking and statistical analysis; Data interface output: Provides a standard REST API interface, supporting integration with Automatic Identification System (AIS), Vessel Traffic Management System (VCR), and Vessel Navigation System (GNSS).

9. A multi-ship satellite navigation receiver jamming information integrated processing system, characterized in that: The method for comprehensively processing interference information of multiple ship satellite navigation receivers as described in any one of claims 1 to 8 includes a data acquisition module, a preprocessing module, a multi-source information collaborative fusion module, an interference source localization and evaluation module, and a result output module. The data acquisition module is used to select multiple ships sailing in the same sea area as distributed observation nodes, and to collect the raw interference data of each ship's satellite navigation receiver and auxiliary data including the ship's real-time position coordinates, navigation status data, environmental data and receiver equipment parameters in real time. The preprocessing module is used to preprocess the raw interference data and auxiliary data of each ship's satellite navigation receiver, and at the same time complete the spatiotemporal synchronization of the raw interference data and auxiliary data to obtain preprocessed synchronization interference data and preprocessed synchronization auxiliary data. The multi-source information collaborative fusion module is used to take the pre-processed synchronization interference data as the core processing object, and combine it with the pre-processed synchronization auxiliary data to perform collaborative fusion and misjudgment correction of multi-ship interference information, so as to obtain the fused interference feature dataset. The interference source localization and evaluation module, based on the fused interference feature dataset, sequentially completes interference subtype subdivision, multi-interference source separation, single-source precise localization, and multi-dimensional parameter evaluation, and finally obtains the location, multi-dimensional parameter evaluation, and comprehensive classification results of each interference source. The result output module is used to output the location, multi-dimensional parameter evaluation and comprehensive classification results of each interference source in a standardized format and to implement them in practice.

Images