Evidence fusion and joint risk assessment method based on GNSS abnormal phase

Abstract

The application discloses a kind of based on GNSS abnormal stage's evidence fusion and joint risk assessment method, it is related to satellite navigation abnormal monitoring and anti-interference technical field, including: at least including power degradation, observation consistency destruction and continuity failure three kinds of feature evidence intensity, abnormal evidence set is constructed;For each kind of abnormal evidence, four self-adaptive adjustment factors of observation reliability factor, abnormal stage matching factor, mutation sensitive factor and redundancy suppression factor are constructed;After four factors are integrated, dynamic fusion weight is generated after normalization processing;According to feature evidence intensity and self-adaptive weight, main risk aggregation item is constructed, and based on the stage evolution relationship between three kinds of abnormalities, cooperative enhancement item is constructed;After the combination of main risk aggregation item and cooperative enhancement item, time smoothing output joint early warning risk score is generated. The application can realize the adaptive fusion and joint risk assessment of GNSS abnormal evidence, with the advantages of good adaptability, high stability and strong interpretability.

Description

Technical Field

[0001] This invention relates to the field of satellite navigation anomaly monitoring and anti-interference technology, and in particular to an evidence fusion and joint risk assessment method based on GNSS anomaly phase. Background Technology

[0002] Global Navigation Satellite Systems (GNSS) are widely used in civil aviation, autonomous driving, power timing, intelligent transportation, unmanned systems, and safety-critical infrastructure. These applications typically require receivers to continuously output reliable positioning, velocity, and timing results in complex environments. However, GNSS signals reach ground receivers at relatively low power, and their open service signal structure makes them susceptible to external electromagnetic interference, changes in the propagation environment, obstruction, multipath effects, ionospheric disturbances, and receiver tracking anomalies. These factors can lead to degraded observation quality and increased navigation service risks. To promptly reflect changes in GNSS observation status, it is necessary to fuse and assess anomaly evidence from different observation layers and develop a unified risk measure that characterizes the current level of anomaly risk.

[0003] In actual anomalies, the physical meanings reflected by different observational evidence are not the same. Suppression interference, signal blockage, or receiver front-end disturbances may initially manifest as a synchronous decrease in the carrier-to-noise ratio of multiple satellites, i.e., enhanced evidence of power degradation. Deception traction, observational disturbances, or complex propagation anomalies may disrupt the kinematic consistency between pseudorange, carrier phase, and Doppler observations, i.e., enhanced evidence of observation consistency disruption. When the anomaly further affects the receiver tracking loop, cycle slips, loss of lock, or observation interruptions may also occur, i.e., enhanced evidence of continuity failure. In some complex anomalies, the above evidence is not completely independent, but may exhibit a staged evolutionary relationship from power degradation to observation consistency disruption, and then to continuity failure. Judging only a single piece of evidence, or simply treating multiple types of evidence side by side, is insufficient to accurately reflect the overall risk level of the anomaly.

[0004] Existing GNSS anomaly assessment methods have the following main shortcomings:

[0005] First, fixed-weight fusion lacks dynamic adaptability. Some methods employ multi-index weighted fusion, but the fusion weights are usually preset and fixed. Due to the real-time changes in GNSS observation conditions (number of visible satellites, satellite geometric distribution, etc.) and anomaly development stages, fixed weights cannot reflect the credibility and indicative role of different evidence at different times. This may lead to insufficiently supported evidence being overemphasized, or the contribution of key evidence not being fully reflected.

[0006] Secondly, the evolutionary information of the anomaly stages is not fully utilized. Most existing methods treat various observation indicators as independent sources of evidence, failing to adequately consider the temporal correlation and stage evolution relationship between power degradation, disruption of observational consistency, and continuity failure. In the early stages of anomalies, some evidence has not yet been fully stimulated, and simple weighting may underestimate the risk; in the later stages of anomalies, the synergistic enhancement effect of various pieces of evidence has not been effectively characterized.

[0007] Third, improper handling of redundancy among evidence. Different pieces of anomalous evidence may have complementary relationships or homogeneous redundancy. For example, pseudorange-Doppler consistency evidence and carrier phase-Doppler consistency evidence are both related to Doppler constraints and have high correlation in some scenarios. If such correlations are not considered during fusion, the repeated contribution of redundant evidence will lead to an inflated risk score.

[0008] Fourth, the balance between interpretability and stability is poor. While data-driven methods (such as neural networks) perform well in certain scenarios, their "black box" nature makes the evaluation results difficult to trace and interpret. At the same time, single-epoch fusion results are susceptible to observation noise, and without an effective smoothing mechanism, fluctuations in risk scores may lead to frequent false alarms.

[0009] A search revealed that some existing patent and non-patent literature solutions detect anomalies based on a single indicator (such as carrier-to-noise ratio or Doppler consistency) without involving the joint evaluation and dynamic fusion of multiple types of evidence; some solutions use multi-indicator weighted fusion, but the weights are fixed preset values ​​and do not consider the time-varying characteristics of evidence reliability; some solutions use data-driven methods for signal quality classification, but the interpretability is poor and the generalization ability is limited.

[0010] As can be seen from the above, existing GNSS anomaly risk assessment methods still struggle to simultaneously meet the requirements of dynamic fusion of multiple types of evidence, anomaly stage evolution modeling, evidence redundancy suppression, and stable risk output. Considering the dynamic characteristics of GNSS anomaly evidence in terms of reliability, stage attributes, temporal variations, and correlation, it is necessary to propose an evidence fusion and joint risk assessment method that can adaptively adjust evidence contributions based on anomaly stage relationships and output a unified joint early warning risk score. Summary of the Invention

[0011] This invention provides an evidence fusion and joint risk assessment method based on GNSS anomaly stages. It achieves dynamic adjustment of evidence contribution by constructing multidimensional adjustment factors and constructs synergistic enhancement terms in combination with stage evolution relationships to improve the adaptability, stability and interpretability of GNSS anomaly risk assessment.

[0012] To solve the above-mentioned technical problems, the technical solution proposed by this invention is as follows:

[0013] A method for evidence fusion and joint risk assessment based on GNSS anomaly phase includes the following steps:

[0014] Step S1: Obtain and construct an anomalous evidence set: Obtain evidence strengths including at least three types of features: power degradation, observation consistency violation, and continuity failure, and construct an anomalous evidence set using observation epochs as time indexes and evidence types as category indexes; Simultaneously, record the number of valid satellites participating in the statistics for each type of evidence and the number of currently visible satellites;

[0015] Step S2, constructing adaptive adjustment factors: For each type of abnormal evidence in the abnormal evidence set, construct corresponding adaptive adjustment factors; the adaptive adjustment factors include observation reliability factor, abnormal stage matching factor, mutation sensitivity factor and redundancy suppression factor;

[0016] Step S3, Generate Adaptive Weights: Integrate the four adaptive adjustment factors obtained in Step S2, calculate the comprehensive support of each type of evidence, and obtain the final dynamic fusion weights through normalization.

[0017] Step S4, construct the main risk aggregation term and the synergistic enhancement term: construct the main risk aggregation term based on the strength of the various feature evidences and the corresponding adaptive weights, and construct the synergistic enhancement term based on the stage evolution relationship between power degradation, observation consistency violation and continuity failure;

[0018] Step S5, Combining and Time Smoothing, Outputting Joint Early Warning Risk Score: The main risk aggregation term and the collaborative enhancement term are combined to obtain the instantaneous joint anomaly score of the current epoch, and the instantaneous joint anomaly score is time smoothed to obtain the GNSS anomaly joint early warning risk score.

[0019] A further improvement to the above technical solution is as follows:

[0020] Preferably, the strength of the three types of characteristic evidence specifically includes:

[0021] (1) The strength of evidence for power degradation is obtained by mapping the common-mode degradation characteristics of the carrier-to-noise ratio of multiple satellites, and is used to characterize the overall degradation degree of GNSS signal quality;

[0022] (2) Strength of evidence related to breaches of observational consistency, including:

[0023] The strength of the first observational consistency evidence is obtained by pseudorange-Doppler consistency feature mapping;

[0024] The strength of the second observational consistency evidence is obtained by carrier phase-Doppler consistency feature mapping;

[0025] (3) The strength of evidence for continuous failure is obtained by mapping the cycle slip, loss of lock or observation interruption state, and is used to characterize the degree of failure of continuous tracking state;

[0026] To observe the epoch Using time as the index and evidence type as the category index, construct an abnormal evidence set:

[0027]

[0028] in, Represents the epoch A collection of unusual evidence, Indicates the number of types of evidence involved in the fusion. Indicates the first Abnormal evidence in the historical period The strength of characteristic evidence, Indicates epoch index, Indicates the evidence type index.

[0029] Preferably, in step S2, the observation reliability factor is constructed in the following manner:

[0030] According to the Abnormal evidence in the historical period Number of valid satellites included in the statistics Number of visible satellites at the current epoch The ratio is used to construct the observation reliability factor:

[0031]

[0032] in, Indicates the first Abnormal evidence in the historical period The observation reliability factor Indicates the first Abnormal evidence in the historical period The number of valid satellites included in the statistics Indicates the current epoch. The number of visible satellites, To prevent positive numbers with a denominator of zero.

[0033] Preferably, in step S2, the abnormal phase matching factor is constructed in the following manner:

[0034]

[0035] in, Indicates the first Abnormal evidence in the historical period The abnormal phase matching factor, To match the scale parameters for the stage, Represents the epoch Abnormal phase index, Indicates the first The depth of the abnormal stage corresponding to the abnormal evidence; when and When the difference is zero, Take the maximum value of 1; as the absolute value of this difference increases, The difference decreases monotonically and approaches 0; the smaller the absolute value of the difference, the higher the degree of matching between this type of evidence and the current stage of evolution, and the contribution of this type of evidence is amplified; the larger the absolute value of the difference, the lower the degree of matching, and the contribution of this type of evidence is suppressed.

[0036] Preferably, in step S2, the mutation-sensitive factor is constructed in the following manner:

[0037]

[0038] in, Indicates the first Abnormal evidence in the historical period Mutation-sensitive factors Indicates the first Abnormal evidence in the previous epoch The strength of characteristic evidence, Indicates the parameter for regulating the mutation response. Indicates the first Abnormal evidence in the historical period The strength of characteristic evidence.

[0039] Preferably, in step S2, the redundancy suppression factor is constructed in the following manner:

[0040] and

[0041] in, Redundancy suppression factor, representing the strength of evidence of consistency among first observations. Redundancy suppression factor representing the strength of evidence for second observational consistency. This represents the redundancy suppression strength coefficient. Indicates the relevance threshold. This indicates that the first and second observational consistency types of evidence are consistent across different epochs. The short-term window correlation coefficient, Redundancy suppression factor representing evidence of power degradation. This represents a redundancy suppression factor for evidence of successive failures.

[0042] Preferably, in step S3, the adaptive weights are constructed in the following manner:

[0043]

[0044] in, Indicates the first Abnormal evidence in the historical period Adaptive weights, Indicates the first Abnormal evidence in the historical period The overall support volume Indicates the first Abnormal evidence in the historical period The overall support volume To prevent positive numbers with a denominator of zero, and Indicates the evidence type index.

[0045] Preferably, in step S4, the main risk aggregation item is constructed in the following manner:

[0046] Based on the Abnormal evidence in the historical period Characteristic Evidence Strength and corresponding adaptive weights Construct the main risk aggregation item:

[0047]

[0048] in, Represents the epoch The main risk aggregation item, Indicates the first Abnormal evidence in the historical period Adaptive weights, Indicates the first Abnormal evidence in the historical period The strength of characteristic evidence, Indicates the evidence type index.

[0049] Preferably, in step S4, the collaborative enhancement term is constructed in the following manner:

[0050] Based on the linkage between power degradation evidence, intermediate layer evidence of observation consistency disruption, and continuous failure evidence, a synergistic enhancement term is constructed:

[0051]

[0052] in, Represents the epoch Synergistic enhancement terms; , , The weighting coefficients of the synergistic enhancement terms are pre-defined non-negative constants that control the strength of the three synergistic effects respectively. This indicates evidence of power degradation in epochs. Evidence with time-delay compensation characteristics; This indicates that the evidence of consistency between the first and second observations is consistent at different epochs. The average value of evidence with time-delay compensation characteristics; Evidence indicating continuous failure in epochs Evidence with time-delay compensation characteristics, This indicates that the first observational consistency evidence is in epochs. The strength of characteristic evidence, This indicates that the second observational consistency evidence is in epochs. The strength of characteristic evidence.

[0053] Preferably, in step S5, the instantaneous joint anomaly score and the joint early warning risk score are constructed in the following manner:

[0054] Combining the main risk aggregation term with the synergistic enhancement term yields the instantaneous joint anomaly score:

[0055]

[0056] in, Represents the epoch Instantaneous joint anomaly score, Represents the epoch The main risk aggregation item, Indicates the weight of collaborative enhancement. Represents the epoch Synergistic enhancement terms;

[0057] The joint early warning risk score is obtained by performing time smoothing on the instantaneous joint anomaly score:

[0058]

[0059] in, Represents the epoch Joint early warning risk score, This is the time smoothing coefficient. Indicates the previous epoch Joint early warning risk score, Represents the epoch Instantaneous joint anomaly score.

[0060] The evidence fusion and joint risk assessment method based on the anomaly phase of GNSS provided by this invention has the following advantages compared with the prior art:

[0061] (1) The evidence fusion and joint risk assessment method based on GNSS anomaly phase of the present invention achieves dynamic adaptive adjustment of fusion weights according to observation conditions, anomaly evolution phase, and correlation between evidences by constructing four dimensions of adaptive adjustment factors: observation reliability factor, anomaly phase matching factor, mutation sensitivity factor, and redundancy suppression factor. The weights of different types of evidence show significant differences in different time periods. Power degradation evidence and pseudorange-Doppler consistency violation evidence maintain high weights in most time periods, while carrier phase consistency evidence and continuity failure evidence mainly play a dominant role in specific anomaly phases, effectively overcoming the lack of dynamic adaptability in existing fixed-weight fusion methods.

[0062] (2) The evidence fusion and joint risk assessment method based on the GNSS anomaly phase of the present invention effectively addresses the problem of homogeneity redundancy between pseudorange-Doppler consistency evidence and carrier phase-Doppler consistency evidence by constructing a redundancy suppression factor based on short-term window correlation. When the two types of consistency evidence are highly correlated, the redundancy suppression factor automatically reduces their repeated contribution in the fusion, avoiding the over-amplification of homogeneous evidence leading to an inflated score and improving the accuracy of risk assessment.

[0063] (3) The evidence fusion and joint risk assessment method based on GNSS anomaly phase of the present invention adopts a combination of main risk aggregation term and synergistic enhancement term, and outputs a joint early warning risk score after time smoothing, which has good interpretability and stability. The main risk aggregation term uses a product form to intuitively reflect the weighted contribution of multiple types of evidence, and the synergistic enhancement term explicitly describes the linkage enhancement effect between anomaly phases, making the assessment results traceable and interpretable; at the same time, the time smoothing process effectively suppresses the score jitter caused by single-epoch observation noise, and the joint early warning risk score R(k) can still accurately track the suddenness, persistence and recovery process of anomalies while maintaining an overall stable trend, achieving a good balance between interpretability and stability. Attached Figure Description

[0064] Figure 1 This is the overall flowchart of the evaluation method of the present invention.

[0065] Figure 2 This is the time-series result of the strength of the four types of abnormal evidence in this invention.

[0066] Figure 3 This is the time-series result of the adaptive weights corresponding to the four types of abnormal evidence in this invention.

[0067] Figure 4 It is the joint early warning risk score in experimental verification. And the results of the classification threshold determination. Detailed Implementation

[0068] The following provides a detailed description of specific embodiments of the present invention. It should be understood that the specific embodiments described herein are for illustrative and explanatory purposes only and are not intended to limit the scope of the invention.

[0069] like Figure 1 As shown, the evidence fusion and joint risk assessment method based on the GNSS anomaly phase of the present invention takes the strength of multiple feature evidence output during the GNSS anomaly monitoring process as input, generates dynamic fusion weights by constructing a multi-dimensional adaptive adjustment factor, and further constructs a main risk aggregation term and a synergistic enhancement term, and finally outputs a time-smoothed joint early warning risk score.

[0070] Specifically, the following steps are included:

[0071] Step S1: Obtain and construct a set of abnormal evidence.

[0072] This step collects various evidence strength information generated during GNSS anomaly monitoring and organizes it into a unified data structure as the basis for subsequent fusion processing. Specifically, it acquires evidence strength information with at least three characteristics: power degradation, observation consistency violation, and continuity failure. An anomaly evidence set is constructed using observation epochs as time indices and evidence types as category indices. Simultaneously, it records the number of valid satellites and currently visible satellites for each type of evidence in the statistics, used for quantifying the reliability of subsequent observations.

[0073] First, the strength of multiple feature evidence generated during GNSS anomaly monitoring is obtained. The strength of multiple feature evidence is a normalized metric obtained by robustly standardizing and evidence mapping the original observation domain anomaly features (such as carrier-to-noise ratio changes, pseudorange / Doppler residuals, cycle slip indicators, etc.). Its numerical range is usually mapped to the interval [0,1], and the larger the value, the more severe the anomaly.

[0074] In this embodiment, the strength of the multi-type feature evidence includes at least:

[0075] (1) Strength of evidence for power degradation, denoted as It is obtained by mapping the common-mode degradation characteristics of the carrier-to-noise ratio of multiple satellites and is used to characterize the overall degradation degree of GNSS signal quality.

[0076] (2) Evidence strength of observation consistency violation is used to characterize the degree to which the kinematic consistency between GNSS observations is violated, and is further subdivided into:

[0077] The strength of evidence for first-order observational consistency is denoted as... It is obtained by pseudorange-Doppler consistency feature mapping;

[0078] The strength of evidence for second observational consistency is denoted as It is obtained by carrier phase-Doppler consistency feature mapping.

[0079] (3) Strength of evidence for continuous failure, denoted as It is obtained by mapping cycle slip, loss of lock, or observation interruption states, and is used to characterize the degree of failure of continuous tracking state.

[0080] To observe the epoch Using time as the index and evidence type as the category index, construct an abnormal evidence set:

[0081] (1)

[0082] in, Describes the set of anomalous evidence for epoch k. Indicates the number of types of evidence involved in the fusion. Indicates the first Abnormal evidence in the historical period The strength of characteristic evidence, Indicates epoch index, In this embodiment, the evidence type index is represented. The evidence of breach of observational consistency includes two subclasses.

[0083] Furthermore, while constructing the set of anomalous evidence, the first... Abnormal evidence in the historical period Number of valid satellites included in the statistics Number of visible satellites at the current epoch The above parameters are used to characterize the observational support conditions for the corresponding anomalous evidence and serve as the basis for subsequent calculation of the observational reliability factor.

[0084] Step S2: Construct the adaptive adjustment factor.

[0085] For each type of anomalous evidence in the aforementioned set of anomalous evidence, a corresponding adaptive adjustment factor is constructed. This adaptive adjustment factor includes an observation reliability factor, anomaly stage matching factor, mutation sensitivity factor, and redundancy suppression factor. These factors reflect the credibility and importance of this type of evidence under current observation conditions and the background of anomalous evolution from different dimensions, providing multidimensional support for the subsequent generation of dynamic weights.

[0086] S2-1, Observational Reliability Factor: Used to characterize the degree of observational support for anomalous evidence.

[0087] According to the Abnormal evidence in the historical period Number of valid satellites included in the statistics Number of visible satellites at the current epoch The ratio is used to construct the observation reliability factor:

[0088] (2)

[0089] in, Indicates the first Abnormal evidence in the historical period The observation reliability factor Indicates the first Abnormal evidence in the historical period The number of valid satellites included in the statistics Indicates the current epoch. The number of visible satellites, To prevent positive numbers with a denominator of zero.

[0090] The observation reliability factor reflects whether the corresponding anomalous evidence has sufficient effective observational support. When the number of effective satellites participating in the statistical analysis for a certain type of anomalous evidence is small, its observation reliability factor decreases accordingly, thereby reducing the impact of this type of evidence in the fusion process. That is, when the number of effective satellites is close to the number of visible satellites, Approaching 1; if the number of effective satellites is small, then This reduces the impact of such evidence in the fusion process.

[0091] S2-2, Anomaly Stage Matching Factor: Used to reflect the degree of matching between a certain type of evidence and the current overall evolution stage of the anomaly.

[0092] To characterize the stage position of different anomalous evidence in the anomalous development process, targeting the first... Setting the depth of abnormal evidence stages Among them, evidence of power degradation corresponds to a smaller depth of anomaly stage, evidence of breached observation consistency corresponds to an intermediate depth of anomaly stage, and evidence of continuity failure corresponds to a larger depth of anomaly stage, i.e., satisfying the following:

[0093] (3)

[0094] in, This indicates the depth of the anomalous stage corresponding to power degradation evidence. This indicates the depth of the anomalous phase corresponding to the first evidence of observational consistency. This indicates the depth of the anomalous phase corresponding to the second evidence of observational consistency. This indicates the depth of the abnormal stage corresponding to evidence of continuous failure. Indicates the evidence type index. When evidence of observational inconsistency includes both pseudorange-Doppler consistency evidence and carrier phase-Doppler consistency evidence, the depth of the anomaly phase can be set to the same or similar levels.

[0095] According to the Anomaly stage depth corresponding to anomaly evidence and the Abnormal evidence in the historical period Characteristic Evidence Strength Construct an abnormal phase index:

[0096] (4)

[0097] in, Represents the epoch The anomaly stage index is used to reflect the comprehensive stage position of various types of anomaly evidence in the evolution relationship of power degradation, observation consistency disruption and continuity failure at the current epoch. Indicates the number of types of evidence involved in the fusion. Indicates the first Abnormal evidence in the historical period The strength of characteristic evidence, Indicates the first The depth of the abnormal stage corresponding to the abnormal evidence. To prevent positive numbers with a denominator of zero.

[0098] Constructing anomaly phase matching factors:

[0099] (5)

[0100] in, Indicates the first Abnormal evidence in the historical period The abnormal phase matching factor, To match the scale parameters for the stage, Represents the epoch Abnormal phase index, Indicates the first The depth of the abnormal stage corresponding to the abnormal evidence. and The closer they get, The closer the value is to 1, the more closely the evidence matches the current stage of evolution, and its contribution should be amplified; conversely, if the two deviate significantly, then... As the value approaches zero, the contribution of this type of evidence is suppressed.

[0101] S2-3, Mutation Sensitive Factor: Used to enhance the response to the phase of rapid increase in the strength of evidence, enabling the risk score to sensitively capture sudden deterioration of anomalies.

[0102] Based on the positive increment of the strength of characteristic evidence between adjacent epochs, a mutation sensitivity factor is constructed:

[0103] (6)

[0104] in, Indicates the first Abnormal evidence in the historical period Mutation-sensitive factors Indicates the first Abnormal evidence in the previous epoch The strength of characteristic evidence, Indicates the parameter for regulating the mutation response. Indicates the first Abnormal evidence in the historical period The characteristic evidence strength. Mutation sensitivity factors are used to enhance the response during the rising phase of abnormal evidence strength, enabling risk scores to maintain high sensitivity in the early or escalating stages of abnormal development.

[0105] S2-4, Redundancy Suppression Factor: Used to reduce the redundant contribution of evidence caused by the homology constraint.

[0106] For the first and second observation consistency class evidences obtained from pseudorange-Doppler consistency features and carrier phase-Doppler consistency features, respectively, considering the existence of a common source constraint relationship between them, a redundancy suppression factor is constructed based on their correlation within a short-time window. The short-time window correlation is:

[0107] (7)

[0108] in, This indicates that the first and second observational consistency types of evidence are consistent across different epochs. The short-time window correlation coefficient, with a value range of [value range missing]. Used to quantify evidence of consistency in first observations. Evidence consistent with second observation The degree of linear correlation within a short time window, the closer the value is to 1, the more synchronous the two changes, indicating that the two types of consistent evidence reflect the same source of anomaly and there is redundancy. Indicates the current epoch. Before A short window consisting of epochs, Indicates the short-time window length. Indicates short window Calculate the Pearson correlation coefficient. This indicates that the second type of feature (pseudorange-Doppler consistency feature) is in the epoch. The strength of characteristic evidence, This indicates that the third type of feature (carrier phase-Doppler consistency feature) is in the epoch. The strength of characteristic evidence.

[0109] The corresponding redundancy suppression factor is:

[0110] and (8)

[0111] in, The redundancy suppression factor representing the second type of feature (pseudorange-Doppler consistency) The redundancy suppression factor representing the third type of feature (carrier phase-Doppler consistency) This represents the redundancy suppression strength coefficient. Indicates the relevance threshold. This indicates that the first and second observational consistency types of evidence are consistent across different epochs. The short-term window correlation coefficient, The redundancy suppression factor represents the type 1 characteristic (power degradation evidence). The redundancy suppression factor represents the fourth type of feature (evidence of continuous failure).

[0112] Through the above processing, when the correlation between two types of observational consistency evidence is high, their redundant contribution in the fusion process can be reduced. When the correlation does not exceed the correlation initiation threshold, the redundancy suppression effect is weak or not activated.

[0113] Step S3: Generate adaptive weights.

[0114] The four adaptive adjustment factors obtained in step S2 are integrated to calculate the overall support for each type of evidence, and the final dynamic fusion weight is obtained through normalization. This weight will be directly used in the construction of the subsequent main risk aggregation term to achieve adaptive adjustment of evidence contribution according to the state.

[0115] First, the first Multiplying the four factors of this type of evidence yields its overall support level:

[0116] (9)

[0117] in, Indicates the first Abnormal evidence in the historical period The overall support volume Indicates the first Abnormal evidence in the historical period The observation reliability factor Indicates the first Abnormal evidence in the historical period The abnormal phase matching factor, Indicates the first Abnormal evidence in the historical period Mutation-sensitive factors Indicates the first Abnormal evidence in the historical period Redundancy suppression factor, Indicates the evidence type index. .

[0118] Then, the combined support quantities of all evidence types involved in the fusion are normalized to obtain the first... Adaptive weights for class evidence :

[0119] (10)

[0120] in, Indicates the first Abnormal evidence in the historical period Adaptive weights, Indicates the first Abnormal evidence in the historical period The overall support volume Indicates the first Abnormal evidence in the historical period The overall support volume To prevent positive numbers with a denominator of zero, and Indicates the evidence type index.

[0121] The resulting adaptive weights can dynamically change with observation conditions, evidence strength, stage evolution relationships, and evidence redundancy.

[0122] Step S4: Construct the main risk aggregation item and the synergistic enhancement item.

[0123] Based on the strength of various feature evidences and their corresponding adaptive weights, a main risk aggregation term is constructed, and a synergistic enhancement term is constructed based on the stage evolution relationship between power degradation, observation consistency disruption, and continuity failure.

[0124] The construction process of the main risk aggregation item and the synergistic enhancement item is as follows:

[0125] Based on the Abnormal evidence in the historical period Characteristic Evidence Strength and corresponding adaptive weights Construct the main risk aggregation item:

[0126] (11)

[0127] in, Represents the epoch The main risk aggregation item, Indicates the first Abnormal evidence in the historical period Adaptive weights, Indicates the first Abnormal evidence in the historical period The strength of characteristic evidence, Indicates the evidence type index.

[0128] Based on the stage evolution relationship between power degradation, observation consistency violation, and continuity failure, time-delay compensation is applied to various types of anomalous evidence to obtain characteristic evidence with time-delay compensation:

[0129] (12)

[0130] in, Indicates the first Abnormal evidence in the historical period Evidence with time delay compensation features Indicates the first Class Anomalous Evidence in the The strength of characteristic evidence for each epoch, Indicates the epoch number that is lagging behind. Indicates the first The maximum time delay compensation window length for anomaly-related evidence. Time delay compensation is used to capture the temporal sequence of events during the evolution of anomaly phases. The maximum time delay compensation window length is set based on the physical time constant of the anomaly evolution; the time delay window for power degradation-related evidence is smaller than that for continuous failure-related evidence.

[0131] Construct intermediate layer evidence based on the first and second observational consistency evidence:

[0132] (13)

[0133] in, Represents the epoch The consistency of observations disrupts evidence from the intermediate layers. This indicates that the first observational consistency evidence is in epochs. Evidence with time-delay compensation characteristics, This indicates that the second observational consistency evidence is in epochs. Evidence with time-delay compensation features.

[0134] Based on the linkage between power degradation evidence, intermediate layer evidence of observation consistency disruption, and continuous failure evidence, a synergistic enhancement term is constructed:

[0135] (14)

[0136] in, Represents the epoch Synergistic enhancement terms; , , The weighting coefficients of the synergistic enhancement terms are pre-defined non-negative constants that control the strength of the three synergistic effects respectively. This indicates evidence of power degradation in epochs. Evidence with time-delay compensation characteristics; This indicates that the evidence of consistency between the first and second observations is consistent at different epochs. The average value of evidence with time-delay compensation characteristics; Evidence indicating continuous failure in epochs Evidence with time-delay compensation characteristics, This indicates that the first observational consistency evidence is in epochs. The strength of characteristic evidence, This indicates that the second observational consistency evidence is in epochs. The strength of characteristic evidence.

[0137] This synergistic enhancement term is used to characterize the combined enhancement between power degradation evidence and observational consistency violation evidence, between observational consistency violation evidence and continuity failure evidence, and between the two types of observational consistency evidence.

[0138] Step S5: Combine and smooth the time, and output the joint early warning risk score.

[0139] The main risk aggregation term is combined with the collaborative enhancement term to obtain the instantaneous joint anomaly score for the current epoch. The instantaneous joint anomaly score is then subjected to time smoothing to obtain the GNSS anomaly joint early warning risk score.

[0140] Combining the main risk aggregation term with the synergistic enhancement term yields the instantaneous joint anomaly score:

[0141] (15)

[0142] in, Represents the epoch Instantaneous joint anomaly score, Represents the epoch The main risk aggregation item, Indicates the weight of collaborative enhancement. Represents the epoch Synergistic enhancements.

[0143] Subsequently, the instantaneous joint anomaly score is smoothed over time to obtain the joint early warning risk score:

[0144] (16)

[0145] in, Represents the epoch Joint early warning risk score, This is the time smoothing coefficient. Indicates the previous epoch Joint early warning risk score, Represents the epoch The instantaneous joint anomaly score. Time smoothing can suppress instantaneous fluctuations and make the risk score changes more stable.

[0146] By using time smoothing, the impact of single-epoch peaks or short-term random fluctuations on risk scoring results can be reduced, making the output joint early warning risk score more stable.

[0147] Experimental verification:

[0148] Figure 2 and Figure 3 It demonstrates the intermediate process of various types of anomalous evidence from the original evidence strength to adaptive weight allocation. Figure 2 Four types of abnormal evidence strength Its value has been normalized to The interval. Wherein, This indicates evidence of power degradation or common-mode anomaly. This indicates evidence of pseudorange-Doppler consistency violation. This indicates evidence of carrier phase-Doppler consistency disruption. This indicates evidence of continuous failure. As shown in the graph, after 08:46, and The levels repeatedly rose above 0.4, and locally approached 0.8–1.0, indicating that evidence of power degradation and disruption of pseudorange-Doppler consistency were quite significant during this anomaly. The signal intensity increased significantly after approximately 08:56, with a local value approaching 1.0, indicating that carrier phase-Doppler consistency disruptions occurred more frequently in the middle and later stages of abnormal development. The high levels reached locally in the early stages, followed by a low level for most of the time, indicating that evidence of continuity failure was not consistently dominant, but rather provided supplementary anomalous information at specific stages.

[0149] Figure 3 Adaptive weights for different types of evidence The changes in weights show that... and It remained at a high level for most of the time, among which It is mainly distributed in the range of approximately 0.55 to 0.70. The values ​​are mainly distributed between approximately 0.40 and 0.60, indicating that evidence of power degradation and pseudorange-Doppler consistency disruption have high fusion weights in this anomaly process. In contrast, For most of the time, the time interval is approximately 0.15 to 0.30. The value remained below 0.20 for most of the time, with increases only observed in certain periods, indicating that carrier phase consistency evidence and continuity failure evidence primarily function during specific anomalous phases. This result suggests that the weights are not fixed but dynamically change with observation reliability, the degree of anomalous phase matching, the state of abrupt response, and evidence redundancy, thus giving different pieces of evidence varying risk support roles at different times.

[0150] Figure 4 This demonstrates the joint early warning risk score obtained by further fusing multiple types of weighted evidence. This indicator is constructed by combining a main risk aggregation term and a synergistic enhancement term, and output after time smoothing. It is used to characterize the comprehensive anomaly risk of GNSS observations at the current epoch. The red solid line in the figure represents... The three dashed lines correspond to the three risk level thresholds: attention, warning, and severe warning.

[0151] Before 08:44, It has remained below 0.05, indicating a low-risk environment. After 08:44, The index rapidly rose to above 0.1 and remained within the range of approximately 0.15 to 0.35 for most of the abnormal periods, indicating that multiple types of abnormal evidence had transformed from localized fluctuations into a sustained, combined risk response. Between approximately 08:50 and 09:10, The repeated breaches of the intermediate threshold, with local peak values ​​approaching 0.4 and some periods approaching or exceeding the severe warning threshold, indicate that the cumulative effect of abnormal evidence is more pronounced during this stage, and the level of systemic risk is further enhanced.

[0152] Combination Figure 2 and Figure 3 The evolution of evidence in the process can be seen as follows: The increase is not directly determined by a single piece of evidence, but rather by the combined effect of multiple types of evidence after adaptive weighting. After 08:46, and Continuously enhancing, and through higher and promote The background level has entered the warning zone. Subsequently, and Participating in enhancement at specific stages enables The occurrence of higher amplitude fluctuations and local peaks reflects a phased synergistic effect between power degradation, disruption of observation consistency, and continuity failure.

[0153] After approximately 09:20, as the strength of anomalous evidence weakened and its weighted contribution decreased, The value gradually decreased to below 0.1 and approached the background level. This change indicates that time smoothing, while suppressing single-epoch peak fluctuations, did not weaken the risk score's ability to track the occurrence, persistence, and recovery processes of anomalies. Therefore, Figure 4 The method of this invention was verified to output the joint early warning risk score from multiple weighted contributions. It can uniformly express the temporal changes of multi-source GNSS anomaly evidence as a continuous risk curve, and provide quantitative basis for event-level early warning and anomaly boundary determination through hierarchical thresholds.

[0154] The above embodiments are merely preferred embodiments of the present invention and are not intended to limit the present invention in any way. Although the present invention has been disclosed above with reference to preferred embodiments, it is not intended to limit the present invention. Therefore, any simple modifications, equivalent changes, and alterations made to the above embodiments based on the technical essence of the present invention without departing from the scope of the present invention should fall within the protection scope of the present invention.

Claims

1. A method for evidence fusion and joint risk assessment based on GNSS anomaly phase, characterized in that, Includes the following steps: Step S1: Obtain and construct an anomalous evidence set: Obtain evidence strengths including at least three types of features: power degradation, observation consistency violation, and continuity failure, and construct an anomalous evidence set using observation epochs as time indexes and evidence types as category indexes; Simultaneously, record the number of valid satellites participating in the statistics for each type of evidence and the number of currently visible satellites; The strength of the three types of characteristic evidence specifically includes: (1) The strength of evidence for power degradation is obtained by mapping the common-mode degradation characteristics of the carrier-to-noise ratio of multiple satellites, and is used to characterize the overall degradation degree of GNSS signal quality; (2) Strength of evidence related to breaches of observational consistency, including: The strength of the first observational consistency evidence is obtained by pseudorange-Doppler consistency feature mapping; The strength of the second observational consistency evidence is obtained by carrier phase-Doppler consistency feature mapping; (3) The strength of evidence for continuous failure is obtained by mapping the cycle slip, loss of lock or observation interruption state, and is used to characterize the degree of failure of continuous tracking state; To observe the epoch Using time as the index and evidence type as the category index, construct an abnormal evidence set: ； in, Describes the set of anomalous evidence for epoch k. Indicates the number of types of evidence involved in the fusion. Indicates the first Abnormal evidence in the historical period The strength of characteristic evidence, Indicates epoch index, Indicates the evidence type index; Step S2, constructing adaptive adjustment factors: For each type of abnormal evidence in the abnormal evidence set, construct corresponding adaptive adjustment factors; the adaptive adjustment factors include observation reliability factor, abnormal stage matching factor, mutation sensitivity factor and redundancy suppression factor; The observation reliability factor is based on the first Abnormal evidence in the historical period Number of valid satellites included in the statistics Number of visible satellites at the current epoch The ratio is used to construct the anomaly stage matching factor, which reflects the degree of matching between a certain type of evidence and the current overall evolution stage of the anomaly; the mutation sensitivity factor is constructed based on the positive increment of the strength of characteristic evidence between adjacent epochs; the redundancy suppression factor is used to reduce the redundant and repetitive contribution of evidence caused by homology constraints. Step S3, Generate Adaptive Weights: Integrate the four adaptive adjustment factors obtained in Step S2, calculate the comprehensive support of each type of evidence, and obtain the final dynamic fusion weights through normalization. Step S4, construct the main risk aggregation term and the synergistic enhancement term: construct the main risk aggregation term based on the strength of the various feature evidences and the corresponding adaptive weights, and construct the synergistic enhancement term based on the stage evolution relationship between power degradation, observation consistency violation and continuity failure; The collaborative enhancements are constructed as follows: Based on the linkage between power degradation evidence, intermediate layer evidence of observation consistency disruption, and continuous failure evidence, a synergistic enhancement term is constructed: ； in, Represents the epoch Synergistic enhancement terms; , , The weighting coefficients of the synergistic enhancement terms are pre-defined non-negative constants that control the strength of the three synergistic effects respectively. This indicates evidence of power degradation in epochs. Evidence with time-delay compensation characteristics; This indicates that the evidence of consistency between the first and second observations is consistent at different epochs. The average value of evidence with time-delay compensation characteristics; Evidence indicating continuous failure in epochs Evidence with time-delay compensation characteristics, This indicates that the first observational consistency evidence is in epochs. The strength of characteristic evidence, This indicates that the second observational consistency evidence is in epochs. The strength of characteristic evidence; Step S5, Combining and Time Smoothing, Outputting Joint Early Warning Risk Score: The main risk aggregation term and the collaborative enhancement term are combined to obtain the instantaneous joint anomaly score of the current epoch, and the instantaneous joint anomaly score is time smoothed to obtain the GNSS anomaly joint early warning risk score.

2. The evidence fusion and joint risk assessment method based on the GNSS anomaly phase according to claim 1, characterized in that, In step S2, the observation reliability factor is constructed in the following manner: The formula for constructing the observation reliability factor is as follows: ； in, Indicates the first Abnormal evidence in the historical period The observation reliability factor Indicates the first Abnormal evidence in the historical period The number of valid satellites included in the statistics Indicates the current epoch. The number of visible satellites, To prevent positive numbers with a denominator of zero.

3. The evidence fusion and joint risk assessment method based on the GNSS anomaly phase according to claim 2, characterized in that, In step S2, the abnormal phase matching factor is constructed in the following manner: ； in, Indicates the first Abnormal evidence in the historical period The abnormal phase matching factor, To match the scale parameters for the stage, Represents the epoch Abnormal phase index, Indicates the first The depth of the abnormal stage corresponding to the abnormal evidence; when and When the difference is zero, Take the maximum value of 1; as this difference increases, The difference decreases monotonically and approaches 0; the smaller the difference, the higher the match between this type of evidence and the current stage of evolution, and the contribution of this type of evidence is amplified; the larger the difference, the lower the match, and the contribution of this type of evidence is suppressed.

4. The evidence fusion and joint risk assessment method based on the GNSS anomaly phase according to claim 3, characterized in that, In step S2, the mutation-sensitive factor is constructed in the following manner: ； in, Indicates the first Abnormal evidence in the historical period Mutation-sensitive factors Indicates the first Abnormal evidence in the previous epoch The strength of characteristic evidence, Indicates the parameter for regulating the mutation response. Indicates the first Abnormal evidence in the historical period The strength of characteristic evidence.

5. The evidence fusion and joint risk assessment method based on the GNSS anomaly phase according to claim 4, characterized in that, In step S2, the redundancy suppression factor is constructed in the following manner: and ; in, Redundancy suppression factor representing the strength of evidence of consistency among first observations. Redundancy suppression factor representing the strength of evidence for second observational consistency. This represents the redundancy suppression strength coefficient. Indicates the relevance threshold. This indicates that the first and second observational consistency types of evidence are consistent across different epochs. The short-term window correlation coefficient, Redundancy suppression factor representing evidence of power degradation. This represents a redundancy suppression factor for evidence of successive failures.

6. The evidence fusion and joint risk assessment method based on the GNSS anomaly phase according to claim 5, characterized in that, In step S3, the adaptive weights are constructed as follows: ； in, Indicates the first Abnormal evidence in the historical period Adaptive weights, Indicates the first Abnormal evidence in the historical period The overall support volume Indicates the first Abnormal evidence in the historical period The overall support volume To prevent positive numbers with a denominator of zero, and Indicates the evidence type index.

7. The evidence fusion and joint risk assessment method based on the GNSS anomaly phase according to claim 6, characterized in that, In step S4, the main risk aggregation item is constructed in the following manner: Based on the Abnormal evidence in the historical period Characteristic Evidence Strength and corresponding adaptive weights Construct the main risk aggregation item: ； in, Represents the epoch The main risk aggregation item, Indicates the first Abnormal evidence in the historical period Adaptive weights, Indicates the first Abnormal evidence in the historical period The strength of characteristic evidence, Indicates the evidence type index.

8. The evidence fusion and joint risk assessment method based on the GNSS anomaly phase according to claim 7, characterized in that, In step S5, the instantaneous joint anomaly score and the joint early warning risk score are constructed in the following manner: Combining the main risk aggregation term with the synergistic enhancement term yields the instantaneous joint anomaly score: ； in, Represents the epoch Instantaneous joint anomaly score, Represents the epoch The main risk aggregation item, Indicates the weight of collaborative enhancement. Represents the epoch Synergistic enhancement terms; The joint early warning risk score is obtained by performing time smoothing on the instantaneous joint anomaly score: ； in, Represents the epoch Joint early warning risk score, This is the time smoothing coefficient. Indicates the previous epoch Joint early warning risk score, Represents the epoch Instantaneous joint anomaly score.

Images