A low-altitude navigation environment real-time monitoring method and system of a multi-source fusion perception terminal

Abstract

The application discloses a kind of low-altitude navigation environment real-time monitoring method and system of multi-source fusion perception terminal, belong to low-altitude safety monitoring technical field.System is divided into edge integration perception terminal and cloud grid monitoring platform;Terminal integrates Beidou synchronous time service, spectrum, acoustics, multi-mode navigation, RID identification five big perception units, each sensor built-in self-calibration circuit, edge side completes original data lightweight feature extraction and carries data quality label upload;Cloud platform is three-dimensional grid to low-altitude airspace, based on signal-to-noise ratio, acquisition completeness, calibration state, data integrity four types of indexes calculate data source credibility, dynamically self-adapting adjustment navigation, spectrum, acoustics, RID four-dimensional environmental factor fusion weight, calculate grid comprehensive navigation safety score;According to factor threshold value, four types of navigation anomalies, such as terrain shelter, radio frequency interference, acoustic noise and cluster overload, are automatically distinguished.

Description

Technical Field

[0001] This invention relates to the field of low-altitude safety monitoring technology, specifically to a method for real-time monitoring of the low-altitude navigation environment using a multi-source fusion sensing terminal. Background Technology

[0002] With the comprehensive development of China's low-altitude economy, low-altitude flight activities are showing a trend of high density, high frequency, and increasing complexity. However, the low-altitude airspace environment itself is characterized by complex building obstructions, chaotic electromagnetic interference, and variable environmental noise, making aircraft navigation signals highly susceptible to the negative impacts of radio frequency interference, acoustic disturbances, or dense flight activities.

[0003] Currently, existing technologies and systems have the following significant drawbacks: 1) Hardware Level: Devices are scattered, data cannot be synchronized from the same source, and self-calibration capabilities are lacking. Spectrum detection equipment, acoustic detection equipment, navigation receivers, and RID identification equipment are mostly deployed independently, resulting in a large number of devices, high costs, and difficult maintenance. More importantly, the acquisition clocks of each device are inconsistent, and the spatial locations are inconsistent, making it impossible to achieve simultaneous acquisition of multi-source data at the same location, making it difficult to trace navigation anomalies to the actual source of interference. Furthermore, after long-term outdoor operation, the sensor performance of existing equipment will deviate due to temperature drift and component aging, but the lack of a built-in self-calibration mechanism causes data confidence to decrease over time.

[0004] 2) Monitoring Methodology: Focusing only on the target, neglecting the environment; fixed weights, lacking adaptability. Traditional monitoring systems only monitor the aircraft's position and trajectory, lacking the ability to quantitatively perceive the navigation environment itself (such as signal quality, electromagnetic spectrum, and acoustic anomalies). Furthermore, existing methods assign fixed weights to each data source, failing to dynamically adjust the fusion weights based on real-time data quality (e.g., spectrum module saturation due to strong local interference at a certain time, acoustic module noise increase due to strong winds). This leads to distorted evaluation results and makes it impossible to scientifically distinguish the specific causes of navigation anomalies (such as terrain obstruction, radio frequency interference, acoustic disturbances, or cluster overload).

[0005] Therefore, the industry urgently needs a complete technical solution that integrates hardware for synchronous data acquisition and self-calibration, and software for dynamic weighted intelligent analysis, in order to build a truly safe foundation for low-altitude navigation. Summary of the Invention

[0006] The purpose of this invention is to provide a method for real-time monitoring of the low-altitude navigation environment using a multi-source fusion sensing terminal, so as to solve the problems mentioned in the background art.

[0007] This invention provides the following technical solution: a method for real-time monitoring of the low-altitude navigation environment using a multi-source fusion sensing terminal, comprising the following steps: S1: Divide the controlled low-altitude airspace into grids in the horizontal and vertical directions, assign a unique code to each grid, and build a gridded digital base for the entire airspace. S2: Within each grid, collect four types of core data, including: grid navigation signal data, grid spectrum interference data, grid acoustic anomaly data, and grid RID aircraft data; S3: Calculate data quality labels for the four types of core data collected, calculate the comprehensive credibility based on the data quality labels, and output the four types of core data and their corresponding comprehensive credibility for each grid. S4: Establish a four-dimensional weighted algorithm to calculate the navigation quality factor, spectrum interference factor, acoustic disturbance factor, and RID density factor based on the four types of core data collected. Construct a dynamic adaptive weighted fusion engine and calculate the fusion weights of the navigation quality factor, spectrum interference factor, acoustic disturbance factor, and RID density factor based on the comprehensive credibility of the four types of core data. Finally, calculate the spatiotemporal comprehensive navigation safety score of each grid. S5: Identify and label the root cause categories of navigation anomalies based on navigation quality factor, spectral interference factor, acoustic disturbance factor, and RID density factor; S6: Perform full-domain situational visualization and simulation of the controlled low-altitude airspace, including: generating a low-altitude navigation heat map and automatically marking navigation blind spots, high-incidence interference areas, and high-disturbance grid areas; storing the visualized operation data according to the time axis and recording the root cause categories of navigation anomalies; and finally, predicting the short-term trend simulation for the next 1-5 minutes based on the time-series change rate of the spatiotemporal integrated navigation safety score of each grid.

[0008] Furthermore, step S1 includes: S11: Using the WGS-84 geodetic coordinate system as the spatial reference, horizontal position is described by latitude and longitude coordinates, and vertical position is described by altitude. S12: The horizontal range of the controlled low-altitude airspace is progressively meshed along the longitude and latitude directions. The specific steps are as follows: First, set the longitude range [Lon_min, Lon_max] and latitude range [Lat_min, Lat_max] of the controlled airspace. Second, according to the application scenario and monitoring accuracy requirements, calculate the longitude step size and latitude step size of the planar mesh. The calculation formula is as follows: ΔLon=GridSize / (111320×cos(Lat)); ΔLat = GridSize / 111320; Where ΔLon is the longitude step, Lat is the latitude, ΔLat is the latitude step, GridSize is the side length of the planar grid, and 111320 is the number of meters corresponding to 1° latitude near the Earth's equator; S13: Divide the controlled area into M columns along the longitude direction with a step size ΔLon. The controlled area is divided into N rows along the latitudinal direction with a step size ΔLat. ,in, This represents the rounding operation. Each planar grid is identified by its row index i and column index j, where 1≤i≤N, 1≤j≤M, and its longitude range is [Lon_min+(j-1)×ΔLon, Lon_min+j×ΔLon], and its latitude range is [Lat_min+(i-1)×ΔLat, Lat_min+i×ΔLat]; S14: In the vertical dimension, set 0 meters as the lowest altitude H_min and 1000 meters as the highest altitude H_max. Then, divide the airspace into K altitude layers at fixed altitude intervals ΔH. The ΔH settings are as follows: coarse-grained ΔH is 100 meters per layer, medium-grained ΔH is 50 meters per layer, and fine-grained ΔH is 25 meters per layer. The height of each layer is k, and 1≤k≤K. S15: Combines the planar mesh with the height layer using Cartesian product to generate a three-dimensional mesh. A complete mesh is uniquely determined by a triple (i, j, k). Each mesh is assigned a globally unique mesh code, supporting fast indexing and spatial retrieval.

[0009] Furthermore, step S2 includes: S21: The grid navigation signal data includes: the average of the carrier-to-noise ratios of all visible satellite signals reported by all terminals as the grid navigation signal strength; the maximum value of the horizontal or vertical positioning error reported by all terminals as the grid positioning error; the average of the horizontal and vertical accuracy factors reported by all terminals as the grid accuracy factor; and the calculation result of 1 - (number of lock-outs / total number of observation epochs) as the signal continuity. S22: The spectral interference data within the grid includes: interference intensity, calculated using the following formula: ; Where A_p is the amplitude value of the p-th frequency point, N_floor is the estimated noise floor value, and P is the total number of frequency points; The number of sudden events in which the spectral interference intensity exceeds the preset threshold by 3dB within the current 60-second time window is counted as the number of disturbances. The criteria for judging a sudden event are: the interference intensity of more than 3 consecutive sampling points exceeds the preset threshold and the duration is not less than 100ms. The total duration of all spectrum interference events within the current time window is calculated as the interference duration. If the interference is still ongoing, it is accumulated to the current time. S23: The grid acoustic anomaly data includes: acoustic anomaly intensity = E_current / E_baseline, where E_current is the short-term energy average of the current time window, and E_baseline is the average noise floor energy of the historical normal period; The perturbation feature is defined as R1·σ_ZCR + R2·ΔE_sub, where σ_ZCR is the variance of the zero-crossing rate within the current time window, ΔE_sub is the consistency change of the subband energy distribution relative to the historical baseline, and R1 and R2 represent the weights of the zero-crossing rate variance and the consistency change of the subband energy distribution, respectively. The weights satisfy R1 + R2 = 1. S24: The grid RID aircraft data includes: the total number of deduplicated aircraft whose broadcast positions fall within the grid range within the current time window, which is taken as the number of RID aircraft in the grid; Cluster density = number of aircraft in the grid / grid volume, where grid volume = grid area × altitude layer thickness, and the unit is aircraft / cubic kilometer; Flight activity = (Number of moving aircraft / Total number of aircraft) × Normalized average speed.

[0010] Furthermore, step S3 includes: S31: The data quality labels corresponding to each type of core data include: signal-to-noise ratio, effective acquisition duration, sensor self-calibration status, and data integrity identifier; S32: The signal-to-noise ratio (SNR)_Index is calculated as follows: For navigation signal data, the average carrier-to-noise ratio (C / No_avg) of visible satellites reported by each terminal in the grid is normalized to the range of 0 to 1 to obtain SNR_Index. When C / No_avg is below 30dB-Hz, SNR_Index is 0; when C / No_avg is above 50dB-Hz, SNR_Index is 1; when C / No_avg is between 30dB-Hz and 50dB-Hz, SNR_Index is calculated by linear interpolation, i.e., SNR_Index = (C / No_avg - 30) / (50 - 30); For spectral interference data, the SNR_Index is obtained by normalizing the ratio of the average signal amplitude to the average noise floor within the current frequency band; for acoustic anomaly data, the SNR_Index is obtained by normalizing the ratio of the current signal energy to the electronic noise floor energy; for RID aircraft data, the SNR_Index is obtained by normalizing the average RID signal strength of all aircraft within the grid. S33: The effective acquisition time Valid_Duty is calculated as follows: Valid_Duty = Total effective acquisition time within the grid / Total expected acquisition time within the grid, with a value of 0 to 1; S34: The calculation method of the sensor self-calibration status Calib_Status is as follows: Each terminal periodically performs self-calibration and reports the calibration status code. Calib_Status 0x00 indicates that the calibration is normal and the deviation is within the tolerance range. Calib_Status 0x01 indicates that the calibration deviation exceeds the limit but has not yet affected the validity of the data. Calib_Status 0x02 indicates that the calibration deviation is serious and the data needs to be used with reduced weight. Calib_Status 0x03 indicates that the sensor is faulty and the data is unavailable. The calibration status code is mapped to a continuous value from 0 to 1, with 0x00 mapped to 1.0, 0x01 to 0.6, 0x02 to 0.2, and 0x03 to 0. If there are multiple terminals in the grid, the average value of the mapped values ​​of each terminal is taken as Calib_Status. S35: The calculation method of the data integrity flag Integrity_Flag is as follows: CRC check, sequence number continuity check, and timestamp rationality check are performed on each frame of reported data; four aspects of verification are performed and identified as a combination of binary status bits: Bit0 indicates CRC check passed, and the value is 1 for pass; Bit1 indicates sequence number continuity, and the value is 1 for continuity; Bit2 indicates timestamp is normal, and the value is 1 for normal; Bit3 indicates data logic consistency, and the value is 1 for consistency; if all verifications pass, the Integrity_Flag value is 1.0; if some verifications pass, the Integrity_Flag value is the number of passed items / 4; if all verifications fail, the Integrity_Flag value is 0. S36: Calculate the overall trustworthiness weight based on the data quality label, including: when there is a single sensing terminal in any grid, the formula for calculating the overall trustworthiness Trust_Source of each type of core data is as follows: Trust_Source=α·SNR_Index+β·Valid_Duty+γ·Calib_Status+δ·Integrity_Flag; Where α, β, γ, and δ are preset weight coefficients that satisfy α+β+γ+δ=1, and Source is any one of Nav, Spec, Acou, and RID, which represent the comprehensive credibility of the core data of the corresponding class. When multiple sensing terminals simultaneously report the same type of core data within any given grid, the overall trustworthiness of the same type of core data from the nth terminal is calculated. Terminal data with an overall trustworthiness below 0.3 is marked as unusable and removed. The overall trustworthiness Trust_Source for each type of core data is then calculated using the following formula: Trust_Source=Σ(Source_n×Trust_Source_n) / ΣTrust_Source_n; Wherein, Source_n represents any type of core data reported by the nth terminal, and Trust_Source_n=α·SNR_Index_n+β·Valid_Duty_n+γ·Calib_Status_n+δ·Integrity_Flag_n, where SNR_Index_n, Valid_Duty_n, Calib_Status_n, and Integrity_Flag_n are the data quality labels of any type of core data collected by the nth terminal, respectively. S37: Based on the four core data values ​​of each grid and their corresponding comprehensive trustworthiness, output the forming vector: [Navigation data value, Trust_Nav], [Spectrum data value, Trust_Spec], [Acoustic data value, Trust_Acou], [RID data value, Trust_RID], which are used to drive the dynamic weight adjustment of each factor in subsequent steps.

[0011] Furthermore, step S4 includes: S41: Calculate the navigation quality factor F_nav, using the following formula: F_nav=w1·S_strength+w2·S_accuracy+w3·S_DOP+w4·S_continuity, Among them, w1, w2, w3, and w4 are the preset weights of the sub-indicators, and the sum of the four weights is 1. The ratio of the measured value of the navigation signal strength within the grid to the nominal normal value is the signal strength sub-index S_strength; The formula for calculating the positioning accuracy sub-index S_accuracy is: S_accuracy=1 / (1+exp(-k1·(HPE_threshold-HPE))); Where HPE is the measured value of the horizontal positioning error within the grid, in meters, calculated and output by the navigation perception module, representing the deviation between the estimated horizontal position of the aircraft or terminal device within the grid and its true position; HPE_threshold is the positioning error tolerance threshold; k1 is the steepness coefficient; exp is the natural exponential function; The formula for calculating the precision factor sub-index S_DOP is: S_DOP=max(0,1-HDOP / HDOP_max); Where HDOP_max is the maximum allowed precision factor, and the index is reset to zero when HDOP is exceeded; HDOP is the measured value of the horizontal precision factor within the grid. The signal continuity sub-index S_continuity adopts the signal continuity value described in step S2; S42: Calculate the spectral interference factor F_spec, using the following formula: F_spec = 1 - ( v1·(1-I_intensity) + v2·(1-I_frequency) + v3·(1-I_duration) ); Where v1, v2, and v3 are the preset weights of the sub-indices, and the interference intensity sub-indice I_intensity = exp(-λ·Intensity), where Intensity is the measured value of the interference intensity and λ is the attenuation coefficient; The disturbance frequency sub-index I_frequency = max(0, 1-Count / Count_max), where Count is the number of disturbances in the current time window, Count_max is the maximum tolerable number of disturbances, and I_frequency returns to zero when Count exceeds Count_max. The interference duration sub-index is I_duration=max(0,1-Duration / Duration_max), where Duration is the total duration of the interference, Duration_max is the maximum tolerable duration, and I_duration returns to zero when Duration exceeds Duration_max. S43: Calculate the acoustic disturbance factor F_acou, using the following formula: F_acou= u1·A_intensity+u2·A_disturbance; Where u1 and u2 are the preset weights of the sub-indicators; The anomaly intensity sub-index A_intensity adopts a piecewise linear mapping: it takes 1.0 when the anomaly intensity is ≤1.0, linearly decreases to 0.3 when the anomaly intensity is between 1.0 and 3.0, and takes 0.3 when the anomaly intensity is >3.0, to avoid the factor from suddenly dropping to zero due to short-term extreme noise; The voiceprint disturbance sub-index A_disturbance=max(0,1-Disturbance_norm / Disturbance_max), where Disturbance_norm is the measured normalized value of the disturbance feature, and Disturbance_max is the maximum tolerable disturbance value; when Disturbance_norm exceeds Disturbance_max, A_disturbance returns to zero. S44: Calculate the RID density factor F_rid, using the following formula: F_rid=t1·D_density+t2·D_activity+t3·D_violation; Where t1, t2, and t3 are the preset weights of the sub-indicators; The density sub-index D_density = max(0, 1-Density / Density_max), where Density is the cluster density, Density_max is the maximum tolerable density, and D_density returns to zero when Density exceeds Density_max. The activity sub-index is D_activity = 1 - Activity, where Activity is the measured value of flight activity, ranging from 0 to 1. The sub-indicator for the proportion of violations is D_violation = 1 - Violation_Ratio, where Violation_Ratio is the proportion of aircraft that violated regulations, and its value ranges from 0 to 1. S45: The method for calculating the fusion weight is as follows: ; ; ; Where p is the adjustment factor; W_nav, W_spec, W_acou, and W_rid are the fusion weights of the navigation quality factor F_nav, the spectral interference factor F_spec, the acoustic disturbance factor F_acou, and the RID density factor F_rid, respectively; Z is the normalization constant, ensuring that the sum of the four weights is 1, and the calculation formula is as follows: ; When Trust_Source < 0.3 for a certain type of data, the dynamic adaptive weighted fusion engine determines that the data source is basically unusable, multiplies its fusion weight coefficient by a penalty factor of 0.1, and renormalizes the other three factors proportionally to ensure that the overall evaluation is not misled by severely poor data. S46: Calculate the spatiotemporal integrated navigation safety score for each grid.

[0012] Furthermore, step S46 also includes: The spatiotemporal integrated navigation safety score S_grid for each grid is calculated using the following formula: S_grid=λ1·S_instant+λ2·S_mean+λ3·(1-S_var); Where λ1, λ2, and λ3 are dimension weights, and the sum of the three weights is 1; Transient score S_instant = W_nav·F_nav + W_spec·F_spec + W_acou·F_acou + W_rid·F_rid; moving average t is the current time, and T is the number of sampling points; variance .

[0013] Furthermore, step S5 includes: identifying and labeling the root cause category of the navigation anomaly, specifically the following root cause categories: S51: Signal attenuation caused by terrain / building obstruction. The following conditions must be met to be classified as signal attenuation: navigation quality factor F_nav < 0.40, spectral interference factor F_spec ≥ 0.55, acoustic disturbance factor F_acou ≥ 0.50, and RID density factor F_rid ≥ 0.50. S52: Interference degradation caused by radio frequency disturbances is classified as interference degradation if the following conditions are met: spectral interference factor F_spec < 0.40; navigation quality factor F_nav < 0.50; acoustic disturbance factor F_acou ≥ 0.50. S53: Environmental disturbances caused by acoustic anomalies and equipment vibration interference are classified as such if the following conditions are met: acoustic disturbance factor F_acou < 0.40; spectral interference factor F_spec ≥ 0.55; navigation quality factor F_nav ≥ 0.50. S54: Situational overload caused by excessively high aircraft cluster density. The following conditions must be met to be classified as situational overload: RID density factor F_rid < 0.35; navigation quality factor F_nav, spectral interference factor F_spec, and acoustic disturbance factor F_acou are all ≥ 0.50.

[0014] Furthermore, step S6 includes: S61: Obtain the following data for all grids at a fixed refresh cycle of 1 second: spatiotemporal integrated navigation safety score S_grid for each grid, four factor values ​​F_nav, F_spec, F_acou, F_rid for each grid, and the root cause category of navigation anomalies for each grid; all grid data are organized in the form of a two-dimensional matrix, where the rows of the matrix correspond to the grid index i in the latitude direction, the columns correspond to the grid index j in the longitude direction, and the matrix element value is the S_grid value of that grid; Each grid's S_grid value is mapped to an RGB color value using a green-yellow-red gradient. For S_grid values ​​between two color gradient boundaries, linear interpolation is used to calculate the RGB components. The interpolation formula is as follows: Let the target value S be located within the interval [S_min, S_max], where S_min is the minimum target value S and S_max is the maximum target value S. The starting color of the interval is RGB_start and the ending color is RGB_end. Then, the interpolation ratio ratio = (S - S_min) / (S_max - S_min), with a value of 0 to 1. R = R_start + ratio × (R_end - R_start), where R_start and R_end are the red R component values ​​of the starting and ending colors of the color interval where the current S_grid value is located, respectively. The G and B components in the RGB components are also calculated in the same way as the R components. R, G, and B are the red, green, and blue components, respectively, with each component ranging from 0 to 255. For grids without terminal coverage, they are displayed in gray and marked "No Data" in the legend. Using a GIS rendering engine, the color-mapped grid data is overlaid and displayed on a low-altitude electronic map. Each grid is rendered as a rectangular block, the location of which is determined by the latitude and longitude range of the grid, and the color of the block is determined by the S_grid value mapping. The final result is a low-altitude navigation heat map, with a refresh cycle synchronized with the data update cycle of 1 second. S62: When the F_nav of three or more consecutive adjacent grids is less than 0.25, and the S_grid of these grids is less than 0.30, and this state lasts for more than 60 seconds, while the spectral interference factor F_spec is greater than or equal to 0.55, it is identified as a navigation blind zone. The blind zone grid area is covered by a blue diagonal stripe overlay on the low-altitude navigation heatmap, and the boundary of the area is outlined with a thick blue dashed line. The navigation blind zone text label and the size of the blind zone are marked. Two grids are extended outward to set a buffer zone label at the edge of the blind zone, indicating a warning message that the blind zone is about to be entered. The label text is "Entering the blind zone is about to begin." When a grid continuously meets the blind zone conditions for more than 60 seconds, it is included in the blind zone. When a grid no longer meets the blind zone conditions and continues to do so for more than 30 seconds, it is removed from the blind zone to avoid frequent fluctuations in the blind zone boundary. S63: Condition A: Persistent interference: F_spec < 0.30 for 3 or more consecutive adjacent grids, and this state lasts for more than 30 seconds; Condition B: Intermittent interference: In the past 10 minutes, the cumulative number of spectral disturbances in a single grid exceeds 10, or the cumulative total duration of interference exceeds 300 seconds; If either condition A or B is met, it is determined to be a high-incidence area of ​​interference. On the heatmap, the grid area of ​​the high-incidence area of ​​interference is covered by a red diagonal overlay, the boundary of the area is outlined with a red solid line, and the text label of high-incidence area of ​​interference is marked. The specific interference type and interference intensity level are superimposed within the area. S64: When the S_var of a single grid is greater than 0.08, S_grid is greater than or equal to 0.40, and the ratio of the average S_var of this grid to that of its neighboring grids is greater than 2.0, it is identified as a high-perturbation grid region. This grid is highlighted on the heatmap with an orange flashing outer glow, and the center of the grid is labeled with a text label and the current variance value. At the same time, the names of the factors that are the main sources of perturbation are also superimposed. S65: Store the visualized operational data according to the timeline and record the root cause categories of navigation anomalies; S66: Prediction of the temporal change rate of the spatiotemporal integrated navigation safety score of each grid, and short-term trend projection for the next 1-5 minutes.

[0015] Furthermore, step S66 also includes: S661: Single-grid short-term trend prediction, the specific steps are as follows: Calculate the historical rate of change, take the S_grid value sequence of the current time t and the previous historical time, use the least squares method to linearly fit, and obtain the rate of change r(t) of the sequence and the fitting residual ε. The fitting formula is: S_grid(t')=a+b·t', where a is the intercept of the fitted line, b is the rate of change r(t), and t' is the time variable; Calculate the prediction confidence score. Based on the fitting residual ε and the historical volatility S_var of S_grid, calculate the prediction confidence score C_pred: C_pred=1-[ε_norm+S_var_norm] / 2, where ε_norm is the normalized fitting residual and S_var_norm is the normalized variance, both of which are mapped to the interval 0 to 1. Perform multi-duration prediction, let the prediction duration be T1, the unit be minutes, and the value range be 1 to 5 minutes. Corresponding to the short-term trend extrapolation of the future 1-5 minutes described in step S6, the prediction value S_pred(t+T)=S_grid(t)+r(t)·T. At the same time, set the physical boundary constraint of the prediction value: S_pred is truncated to the interval [0,1]. S662: Risk grid diffusion direction prediction, the specific steps are as follows: identify risk boundary grids, mark all grids with S_grid<0.50 as risk grids, extract the outer boundary grids of the risk grid set, that is, risk grids with at least one adjacent non-risk grid, as the target grids for diffusion analysis; Calculate the S_grid gradient at the boundary. For each boundary grid, calculate the S_grid gradient value pointing to the adjacent non-risk grid. The S_grid value of boundary grid A is S_A, and the S_grid value of its adjacent non-risk grid B is S_B. Then the gradient from A to B is G=(S_B-S_A) / d, where d is the distance between the grid centers. To determine the diffusion direction, for each boundary grid, the direction with the largest gradient value is selected as the main diffusion direction of that grid. The diffusion directions of all boundary grids in the same risk area are vector-superimposed to obtain the overall main diffusion direction vector D_major of that risk area. Calculate the diffusion rate V_diffusion, which is the average movement speed of the risk area boundary over the most recent N time windows; Generate a diffusion prediction based on the diffusion direction D_major and diffusion velocity V_diffusion, predicting the movement trajectory of the risk boundary within the next 1-5 minutes.

[0016] This invention also discloses a real-time monitoring system for low-altitude navigation environment using a multi-source fusion sensing terminal, comprising: Synchronization Time Unit: Access to BeiDou satellite time synchronization provides a unified clock reference for all sensing modules, controls the timestamp error of all multi-source data acquisition to the microsecond level, and outputs high-precision timestamps for backend multi-source data association and alignment; Spectrum detection module: Covers low-altitude navigation frequency bands and general radio frequency bands, scans and monitors the airspace spectrum status in real time, collects frequency band occupancy, interference intensity, interference bandwidth, sudden spectrum disturbances, and interference azimuth information, identifies narrowband, broadband, and intermittent radio frequency interference, and has a built-in reference noise source component to support periodic self-calibration based on the environmental spectrum noise floor to compensate for receiver link gain drift. High-sensitivity acoustic detection module: Equipped with a noise-reducing pickup array, it accurately collects low-altitude rotor noise, aircraft cluster sound field, operational abnormal noise, and sudden environmental noise, outputs standardized voiceprint feature data, and has a built-in standard electrical signal calibration circuit that periodically injects calibration signals of known frequency and amplitude to correct microphone sensitivity and frequency response curve deviation. Navigation Sensing Module: Fusion calculation of BeiDou, GNSS and inertial navigation, real-time output of navigation quality parameters such as signal strength, number of visible satellites, positioning error, accuracy factor, signal continuity, and multi-mode switching stability, supports dual-antenna / multi-constellation redundant observation, and can self-check and report the confidence level of the terminal's position calculation; RID identification module: Real-time monitoring and parsing of aircraft broadcast RID data, extracting unique device ID, aircraft type, location, flight permission, take-off and landing information; and marking targets without RID, with tampered RID, or with abnormal RID. It has a built-in RID signal parsing confidence evaluation unit and outputs link quality indicators such as signal strength and bit error rate. Main control and data transmission unit: includes edge side lightweight processing module, which can perform on-chip feature extraction on the original spectrum and acoustic data, specifically including: (1) spectrum feature extraction submodule: performs fast Fourier transform on the spectrum sampling data, calculates the energy ratio, peak frequency and amplitude, mean background noise and time series change rate of each preset frequency band, and outputs spectrum feature vector; (2) acoustic feature extraction submodule: performs frame windowing processing on the acoustic sampling data, calculates short-time zero crossing rate, short-time energy, sub-band energy distribution and time domain envelope features, and optionally extracts Mel frequency cepstral coefficients as voiceprint recognition features; (3) the lightweight feature vector extracted on-chip is uploaded to the cloud through 4G / 5G / network port, the original data can be cached locally for traceability and evidence collection, thereby reducing the uplink data bandwidth requirement, uniformly scheduling the parallel acquisition of each module, uniform protocol encapsulation, and uploading multi-source synchronous data to the cloud in real time through 4G / 5G / network port, while reporting the self-calibration status of each sensor and data quality label; Power supply and protection unit: Supports mains power, DC power, and lithium battery backup power; features surge protection, electrostatic discharge protection, overvoltage and undervoltage protection; meets 24 / 7 outdoor uninterrupted operation requirements. Compared with the prior art, the beneficial effects of the present invention are: 1. The low-altitude navigation environment real-time monitoring method of this multi-source fusion sensing terminal, at the hardware terminal level: an integrated single device replaces multiple sets of separate monitoring equipment, reducing deployment, wiring, power supply and maintenance costs; unified BeiDou timing enables simultaneous acquisition of four types of data at the same location and at the microsecond level, solving the problem of difficult spatiotemporal tracing of navigation anomalies; the radio frequency and acoustic modules have built-in self-calibration circuits to automatically compensate for temperature drift and device aging deviations, maintaining high data reliability during long-term outdoor operation; on-chip feature extraction at the edge significantly compresses uplink data traffic, supporting large-scale, high-density network deployment across the entire city.

[0017] 2. The low-altitude navigation environment real-time monitoring method of this multi-source fusion sensing terminal, at the cloud algorithm level: enables the industry to upgrade from "only monitoring aircraft targets" to "full-domain grid-based quantitative monitoring of the airspace navigation environment"; abandons the traditional fixed-weight fusion scheme, dynamically adjusts the weight of each factor according to the reliability of real-time data, and automatically reduces the weight when a single sensor is degraded by environmental interference, avoiding global assessment distortion; it can automatically distinguish four types of navigation anomaly causes and subdivide the interference types, eliminating the need for manual item-by-item investigation, and greatly improving the efficiency of low-altitude risk handling.

[0018] 3. The real-time monitoring method for low-altitude navigation environment of this multi-source fusion sensing terminal, at the system level, forms a complete closed-loop system through hardware synchronous acquisition, edge preprocessing, cloud-based grid-based analysis, automatic anomaly tracing, full-domain thermal visualization, and short-term risk simulation; it outputs standardized grid safety scores, anomaly classifications, and risk diffusion prediction results, providing quantitative and robust decision support for low-altitude dynamic airspace management, electromagnetic interference enforcement, and UAV flow restriction scheduling. Attached Figure Description

[0019] Figure 1 Architecture diagram for multi-source synchronous data gridding import and data quality assessment; Figure 2 Here is a flowchart of the four-dimensional factor dynamic adaptive weighted fusion algorithm; Figure 3 A logic block diagram for automatically identifying and tracing the causes of navigation anomalies. Detailed Implementation

[0020] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0021] Example 1: A method for real-time monitoring of the low-altitude navigation environment using a multi-source fusion sensing terminal, comprising the following steps: S1: Divide the controlled low-altitude airspace into grids in both the horizontal and vertical directions, assign a unique code to each grid, and construct a gridded digital foundation for the entire airspace, including: S11: Using the WGS-84 geodetic coordinate system as the spatial reference, horizontal position is described by latitude and longitude coordinates, and vertical position is described by altitude. S12: The horizontal range of the controlled low-altitude airspace is progressively meshed along the longitude and latitude directions. The specific steps are as follows: First, set the longitude range [Lon_min, Lon_max] and latitude range [Lat_min, Lat_max] of the controlled airspace. Second, according to the application scenario and monitoring accuracy requirements, calculate the longitude step size and latitude step size of the planar mesh. The calculation formula is as follows: ΔLon=GridSize / (111320×cos(Lat)); ΔLat = GridSize / 111320; Where ΔLon is the longitude step, Lat is the latitude, ΔLat is the latitude step, GridSize is the side length of the planar grid, and 111320 is the number of meters corresponding to 1° latitude near the Earth's equator; S13: Divide the controlled area into M columns along the longitude direction with a step size ΔLon. The controlled area is divided into N rows along the latitudinal direction with a step size ΔLat. ,in, This represents the rounding operation. Each planar grid is identified by its row index i and column index j, where 1≤i≤N, 1≤j≤M, and its longitude range is [Lon_min+(j-1)×ΔLon, Lon_min+j×ΔLon], and its latitude range is [Lat_min+(i-1)×ΔLat, Lat_min+i×ΔLat]; S14: In the vertical dimension, set 0 meters as the lowest altitude H_min and 1000 meters as the highest altitude H_max. Then, divide the airspace into K altitude layers at fixed altitude intervals ΔH. ,in, This indicates a rounding operation. ΔH is set as follows: coarse-grained ΔH is 100 meters per layer, medium-grained ΔH is 50 meters per layer, and fine-grained ΔH is 25 meters per layer. Each height layer is k, and 1≤k≤K. S15: Combines the planar mesh with the height layer using Cartesian product to generate a three-dimensional mesh. A complete mesh is uniquely determined by a triple (i, j, k). Each mesh is assigned a globally unique mesh code, supporting fast indexing and spatial retrieval.

[0022] S2: Within each grid, four types of core data are collected, including: grid navigation signal data, grid spectrum interference data, grid acoustic anomaly data, and grid RID aircraft data, including: S21: The grid navigation signal data includes: the average of the carrier-to-noise ratios of all visible satellite signals reported by all terminals as the grid navigation signal strength; the maximum value of the horizontal or vertical positioning error reported by all terminals as the grid positioning error; the average of the horizontal and vertical accuracy factors reported by all terminals as the grid accuracy factor; and the calculation result of 1 - (number of lock-outs / total number of observation epochs) as the signal continuity. S22: The spectral interference data within the grid includes: interference intensity, calculated using the following formula: ; Where A_p is the amplitude value of the p-th frequency point, N_floor is the estimated noise floor value, and P is the total number of frequency points; The number of sudden events in which the spectral interference intensity exceeds the preset threshold by 3dB within the current 60-second time window is counted as the number of disturbances. The criteria for judging a sudden event are: the interference intensity of more than 3 consecutive sampling points exceeds the preset threshold and the duration is not less than 100ms. The total duration of all spectrum interference events within the current time window is calculated as the interference duration. If the interference is still ongoing, it is accumulated to the current time. S23: The grid acoustic anomaly data includes: acoustic anomaly intensity = E_current / E_baseline, where E_current is the short-term energy average of the current time window, and E_baseline is the average noise floor energy of the historical normal period; The perturbation feature is defined as R1·σ_ZCR + R2·ΔE_sub, where σ_ZCR is the variance of the zero-crossing rate within the current time window, ΔE_sub is the consistency change of the subband energy distribution relative to the historical baseline, and R1 and R2 represent the weights of the zero-crossing rate variance and the consistency change of the subband energy distribution, respectively. The weights satisfy R1 + R2 = 1. S24: The grid RID aircraft data includes: the total number of deduplicated aircraft whose broadcast positions fall within the grid range within the current time window, which is taken as the number of RID aircraft in the grid; Cluster density = number of aircraft in the grid / grid volume, where grid volume = grid area × altitude layer thickness, and the unit is aircraft / cubic kilometer; Flight activity = (Number of moving aircraft / Total number of aircraft) × Normalized average speed.

[0023] S3: Calculate data quality labels for the four types of core data collected, calculate the comprehensive credibility based on the data quality labels, and output the four types of core data and their corresponding comprehensive credibility for each grid, including: S31: The data quality labels corresponding to each type of core data include: signal-to-noise ratio, effective acquisition duration, sensor self-calibration status, and data integrity identifier; S32: The signal-to-noise ratio (SNR)_Index is calculated as follows: For navigation signal data, the average carrier-to-noise ratio (C / No_avg) of visible satellites reported by each terminal in the grid is normalized to the range of 0 to 1 to obtain SNR_Index. When C / No_avg is below 30dB-Hz, SNR_Index is 0; when C / No_avg is above 50dB-Hz, SNR_Index is 1; when C / No_avg is between 30dB-Hz and 50dB-Hz, SNR_Index is calculated by linear interpolation, i.e., SNR_Index = (C / No_avg - 30) / (50 - 30); For spectral interference data, the SNR_Index is obtained by normalizing the ratio of the average signal amplitude to the average noise floor within the current frequency band; for acoustic anomaly data, the SNR_Index is obtained by normalizing the ratio of the current signal energy to the electronic noise floor energy; for RID aircraft data, the SNR_Index is obtained by normalizing the average RID signal strength of all aircraft within the grid. S33: The effective acquisition time Valid_Duty is calculated as follows: Valid_Duty = Total effective acquisition time within the grid / Total expected acquisition time within the grid, with a value of 0 to 1; S34: The calculation method of the sensor self-calibration status Calib_Status is as follows: Each terminal periodically performs self-calibration and reports the calibration status code. Calib_Status 0x00 indicates that the calibration is normal and the deviation is within the tolerance range. Calib_Status 0x01 indicates that the calibration deviation exceeds the limit but has not yet affected the validity of the data. Calib_Status 0x02 indicates that the calibration deviation is serious and the data needs to be used with reduced weight. Calib_Status 0x03 indicates that the sensor is faulty and the data is unavailable. The calibration status code is mapped to a continuous value from 0 to 1, with 0x00 mapped to 1.0, 0x01 to 0.6, 0x02 to 0.2, and 0x03 to 0. If there are multiple terminals in the grid, the average value of the mapped values ​​of each terminal is taken as Calib_Status. S35: The calculation method of the data integrity flag Integrity_Flag is as follows: CRC check, sequence number continuity check, and timestamp rationality check are performed on each frame of reported data; four aspects of verification are performed and identified as a combination of binary status bits: Bit0 indicates CRC check passed, and the value is 1 for pass; Bit1 indicates sequence number continuity, and the value is 1 for continuity; Bit2 indicates timestamp is normal, and the value is 1 for normal; Bit3 indicates data logic consistency, and the value is 1 for consistency; if all verifications pass, the Integrity_Flag value is 1.0; if some verifications pass, the Integrity_Flag value is the number of passed items / 4; if all verifications fail, the Integrity_Flag value is 0. S36: Calculate the overall trustworthiness weight based on the data quality label, including: when there is a single sensing terminal in any grid, the formula for calculating the overall trustworthiness Trust_Source of each type of core data is as follows: Trust_Source=α·SNR_Index+β·Valid_Duty+γ·Calib_Status+δ·Integrity_Flag; Where α, β, γ, and δ are preset weight coefficients that satisfy α+β+γ+δ=1, and Source is any one of Nav, Spec, Acou, and RID, which represent the comprehensive credibility of the core data of the corresponding class. When multiple sensing terminals simultaneously report the same type of core data within any given grid, the overall trustworthiness of the same type of core data from the nth terminal is calculated. Terminal data with an overall trustworthiness below 0.3 is marked as unusable and removed. The overall trustworthiness Trust_Source for each type of core data is then calculated using the following formula: Trust_Source=Σ(Source_n×Trust_Source_n) / ΣTrust_Source_n; Wherein, Source_n represents any type of core data reported by the nth terminal, and Trust_Source_n=α·SNR_Index_n+β·Valid_Duty_n+γ·Calib_Status_n+δ·Integrity_Flag_n, where SNR_Index_n, Valid_Duty_n, Calib_Status_n, and Integrity_Flag_n are the data quality labels of any type of core data collected by the nth terminal, respectively. S37: Based on the four core data values ​​of each grid and their corresponding comprehensive trustworthiness, output the forming vector: [Navigation data value, Trust_Nav], [Spectrum data value, Trust_Spec], [Acoustic data value, Trust_Acou], [RID data value, Trust_RID], which are used to drive the dynamic weight adjustment of each factor in subsequent steps.

[0024] S4: Establish a four-dimensional weighted algorithm, such as Figure 2 As shown, navigation quality factor, spectral interference factor, acoustic disturbance factor, and RID density factor are calculated based on the four types of core data collected. A dynamic adaptive weighted fusion engine is constructed, and the fusion weights of navigation quality factor, spectral interference factor, acoustic disturbance factor, and RID density factor are calculated based on the comprehensive reliability corresponding to the four types of core data. Finally, the spatiotemporal integrated navigation safety score of each grid is calculated, including: S41: Calculate the navigation quality factor F_nav, the calculation formula is: F_nav=w1·S_strength+w2·S_accuracy+w3·S_DOP+w4·S_continuity, Among them, w1, w2, w3, and w4 are the preset weights of the sub-indicators. We take w1=0.25, w2=0.30, w3=0.20, and w4=0.25 to satisfy the condition that the sum of the four weights is 1. The ratio of the measured value of the navigation signal strength within the grid to the nominal normal value is the signal strength sub-index S_strength, and the nominal normal value is 45dB-Hz; The formula for calculating the positioning accuracy sub-index S_accuracy is: S_accuracy=1 / (1+exp(-k1·(HPE_threshold-HPE))); Wherein, HPE is the measured value of the horizontal position error within the grid, in meters, calculated and output by the navigation perception module, representing the deviation between the estimated horizontal position of the aircraft or terminal equipment within the grid and the actual position; HPE_threshold is the positioning error tolerance threshold, which is 5m; k1 is the steepness coefficient, which is 1.0; exp is the natural exponential function; The formula for calculating the precision factor sub-index S_DOP is: S_DOP=max(0,1-HDOP / HDOP_max); Where HDOP_max is the maximum allowed precision factor, which is set to 10. When HDOP is exceeded, the index is reset to zero; HDOP is the measured value of the horizontal precision factor within the grid. The signal continuity sub-index S_continuity adopts the signal continuity value described in step S2; S42: Calculate the spectral interference factor F_spec, using the following formula: F_spec = 1 - ( v1·(1-I_intensity) + v2·(1-I_frequency) + v3·(1-I_duration) ); Where v1, v2, and v3 are the preset weights of the sub-indices, with v1=0.40, v2=0.30, and v3=0.30 respectively. The interference intensity sub-indice I_intensity=exp(-λ·Intensity), where Intensity is the measured value of the interference intensity, and λ is the attenuation coefficient, which is 0.2. The disturbance frequency sub-index I_frequency = max(0, 1-Count / Count_max), where Count is the number of disturbances in the current time window, and Count_max is the maximum tolerable number of disturbances, which is 20 times / hour. When Count exceeds Count_max, I_frequency returns to zero. The interference duration sub-index is I_duration=max(0,1-Duration / Duration_max), where Duration is the total duration of the interference, Duration_max is the maximum tolerable duration, which is 600 seconds / hour. When Duration exceeds Duration_max, I_duration is reset to zero. S43: Calculate the acoustic disturbance factor F_acou, using the following formula: F_acou= u1·A_intensity+u2·A_disturbance; Where u1 and u2 are the preset weights of the sub-indicators, with u1=0.50 and u2=0.50; The anomaly intensity sub-index A_intensity adopts a piecewise linear mapping: it takes 1.0 when the anomaly intensity is ≤1.0, linearly decreases to 0.3 when the anomaly intensity is between 1.0 and 3.0, and takes 0.3 when the anomaly intensity is >3.0, to avoid the factor from suddenly dropping to zero due to short-term extreme noise; The voiceprint disturbance sub-index A_disturbance = max(0, 1 - Disturbance_norm / Disturbance_max), where Disturbance_norm is the measured normalized value of the disturbance characteristic, and Disturbance_max is the maximum tolerable disturbance value, which is 0.8; when Disturbance_norm exceeds Disturbance_max, A_disturbance returns to zero. S44: Calculate the RID density factor F_rid, using the following formula: F_rid=t1·D_density+t2·D_activity+t3·D_violation; Where t1, t2, and t3 are the preset weights of the sub-indicators, with t1=0.35, t2=0.30, and t3=0.35; The density sub-index D_density = max(0, 1-Density / Density_max), where Density is the cluster density and Density_max is the maximum tolerable density, which is taken as 100 aircraft / cubic kilometer. When Density exceeds Density_max, D_density returns to zero. The activity sub-index is D_activity = 1 - Activity, where Activity is the measured value of flight activity, ranging from 0 to 1. The sub-indicator for the proportion of violations is D_violation = 1 - Violation_Ratio, where Violation_Ratio is the proportion of aircraft that violated regulations, and its value ranges from 0 to 1. S45: The method for calculating the fusion weight is as follows: ; ; ; Where p is the adjustment factor, set to 1.5; W_nav, W_spec, W_acou, and W_rid are the fusion weights of the navigation quality factor F_nav, the spectral interference factor F_spec, the acoustic disturbance factor F_acou, and the RID density factor F_rid, respectively; Z is the normalization constant, ensuring that the sum of the four weights is 1, and the calculation formula is as follows: ; When Trust_Source < 0.3 for a certain type of data, the dynamic adaptive weighted fusion engine determines that the data source is basically unusable, multiplies its fusion weight coefficient by a penalty factor of 0.1, and renormalizes the other three factors proportionally to ensure that the overall evaluation is not misled by severely poor data. S46: Calculate the spatiotemporal integrated navigation safety score for each grid.

[0025] Step S46 also includes: The spatiotemporal integrated navigation safety score S_grid for each grid is calculated using the following formula: S_grid=λ1·S_instant+λ2·S_mean+λ3·(1-S_var); Where λ1, λ2, and λ3 are dimension weights, with λ1=0.50, λ2=0.30, and λ3=0.20, satisfying that the sum of the three weights is 1; Transient score S_instant = W_nav·F_nav + W_spec·F_spec + W_acou·F_acou + W_rid·F_rid; moving average t is the current time, T is the number of sampling points, which is 600; variance .

[0026] S5: Identify and label the root cause categories of navigation anomalies based on navigation quality factor, spectral interference factor, acoustic disturbance factor, and RID density factor, such as... Figure 3 As shown, it includes: S51: Signal attenuation caused by terrain / building obstruction. The following conditions must be met to be classified as signal attenuation: navigation quality factor F_nav < 0.40, spectral interference factor F_spec ≥ 0.55, acoustic disturbance factor F_acou ≥ 0.50, and RID density factor F_rid ≥ 0.50. S52: Interference degradation caused by radio frequency disturbances is classified as interference degradation if the following conditions are met: spectral interference factor F_spec < 0.40; navigation quality factor F_nav < 0.50; acoustic disturbance factor F_acou ≥ 0.50. S53: Environmental disturbances caused by acoustic anomalies and equipment vibration interference are classified as such if the following conditions are met: acoustic disturbance factor F_acou < 0.40; spectral interference factor F_spec ≥ 0.55; navigation quality factor F_nav ≥ 0.50. S54: Situational overload caused by excessively high aircraft cluster density. The following conditions must be met to be classified as situational overload: RID density factor F_rid < 0.35; navigation quality factor F_nav, spectral interference factor F_spec, and acoustic disturbance factor F_acou are all ≥ 0.50.

[0027] S6: Perform full-domain situational visualization and simulation for the controlled low-altitude airspace, including: generating a low-altitude navigation heatmap and automatically marking navigation blind spots, high-incidence interference areas, and high-disturbance grid areas; storing the visualized operational data according to the time axis and recording the root cause categories of navigation anomalies; and finally, predicting the short-term trend simulation for the next 1-5 minutes based on the time-series change rate of the spatiotemporal integrated navigation safety score of each grid, including: S61: Obtain the following data for all grids at a fixed refresh cycle of 1 second: spatiotemporal integrated navigation safety score S_grid for each grid, four factor values ​​F_nav, F_spec, F_acou, F_rid for each grid, and the root cause category of navigation anomalies for each grid; all grid data are organized in the form of a two-dimensional matrix, where the rows of the matrix correspond to the grid index i in the latitude direction, the columns correspond to the grid index j in the longitude direction, and the matrix element value is the S_grid value of that grid; Each grid's S_grid value is mapped to an RGB color value using a green-yellow-red gradient. For S_grid values ​​between two color gradient boundaries, linear interpolation is used to calculate the RGB components. The interpolation formula is as follows: Let the target value S be located within the interval [S_min, S_max], where S_min is the minimum target value S and S_max is the maximum target value S. The starting color of the interval is RGB_start, and the ending color is RGB_end. Then, the interpolation ratio ratio = (S - S_min) / (S_max - S_min), taking values ​​from 0 to 1; R = R_start + ratio × (R_end - R_start), where R_start and R_end are the red (R) component values ​​of the starting and ending colors of the color interval where the current S_grid value is located, respectively. The G and B components in the RGB components are also calculated using the same method as the R components. R, G, and B are the red, green, and blue components, respectively, with each component taking values ​​from 0 to 255. For grids without terminal coverage, they are displayed in gray (RGB = (200, 200, ...). 200), and mark "No data" in the legend; Using a GIS rendering engine, the color-mapped grid data is overlaid and displayed on a low-altitude electronic map. Each grid is rendered as a rectangular block, the location of which is determined by the latitude and longitude range of the grid, and the color of the block is determined by the S_grid value mapping. The final result is a low-altitude navigation heat map, with a refresh cycle synchronized with the data update cycle of 1 second. S62: When the F_nav of three or more consecutive adjacent grids is less than 0.25, and the S_grid of these grids is less than 0.30, and this state lasts for more than 60 seconds, while the spectral interference factor F_spec is greater than or equal to 0.55, it is identified as a navigation blind zone. The blind zone grid area is covered by a blue diagonal stripe overlay on the low-altitude navigation heatmap, and the boundary of the area is outlined with a thick blue dashed line. The navigation blind zone text label and the size of the blind zone are marked. Two grids are extended outward to set a buffer zone label at the edge of the blind zone, indicating a warning message that the blind zone is about to be entered. The label text is "Entering the blind zone is about to begin." When a grid continuously meets the blind zone conditions for more than 60 seconds, it is included in the blind zone. When a grid no longer meets the blind zone conditions and continues to do so for more than 30 seconds, it is removed from the blind zone to avoid frequent fluctuations in the blind zone boundary. S63: Condition A: Persistent interference: F_spec < 0.30 for 3 or more consecutive adjacent grids, and this state lasts for more than 30 seconds; Condition B: Intermittent interference: In the past 10 minutes, the cumulative number of spectral disturbances in a single grid exceeds 10, or the cumulative total duration of interference exceeds 300 seconds; If either condition A or B is met, it is determined to be a high-incidence area of ​​interference. On the heatmap, the grid area of ​​the high-incidence area of ​​interference is covered by a red diagonal overlay, the boundary of the area is outlined with a red solid line, and the text label of high-incidence area of ​​interference is marked. The specific interference type and interference intensity level are superimposed within the area. S64: When the S_var of a single grid is greater than 0.08, S_grid is greater than or equal to 0.40, and the ratio of the average S_var of this grid to that of its neighboring grids is greater than 2.0, it is identified as a high-perturbation grid region. This grid is highlighted on the heatmap with an orange flashing outer glow, and the center of the grid is labeled with a text label and the current variance value. At the same time, the names of the factors that are the main sources of perturbation are also superimposed. S65: Store the visualized operational data according to the timeline and record the root cause categories of navigation anomalies; S66: Prediction of the temporal change rate of the spatiotemporal integrated navigation safety score of each grid, and short-term trend projection for the next 1-5 minutes.

[0028] Step S66 further includes: S661: Single-grid short-term trend prediction, the specific steps are as follows: Calculate the historical rate of change, take the S_grid value sequence of the current time t and the previous historical time, use the least squares method for linear fitting, and obtain the rate of change (slope) r(t) of the sequence, and the fitting residual ε. The fitting formula is: S_grid(t')=a+b·t', where a is the intercept of the fitted line (dimensionless), b is the rate of change r(t) (unit is change / minute), and t' is the time variable (unit is minutes). Calculate the prediction confidence score. Based on the fitting residual ε and the historical volatility S_var of S_grid, calculate the prediction confidence score C_pred: C_pred=1-[ε_norm+S_var_norm] / 2, where ε_norm is the normalized fitting residual and S_var_norm is the normalized variance, both of which are mapped to the interval 0 to 1. Perform multi-duration prediction, let the prediction duration be T1, the unit be minutes, and the value range be 1 to 5 minutes. Corresponding to the short-term trend extrapolation of the future 1-5 minutes described in step S6, the prediction value S_pred(t+T)=S_grid(t)+r(t)·T. At the same time, set the physical boundary constraint of the prediction value: S_pred is truncated to the interval [0,1]. S662: Risk grid diffusion direction prediction, the specific steps are as follows: identify risk boundary grids, mark all grids with S_grid<0.50 as risk grids, extract the outer boundary grids of the risk grid set, that is, risk grids with at least one adjacent non-risk grid, as the target grids for diffusion analysis; Calculate the S_grid gradient at the boundary. For each boundary grid, calculate the S_grid gradient value pointing to the adjacent non-risk grid. The S_grid value of boundary grid A is S_A, and the S_grid value of its adjacent non-risk grid B is S_B. Then the gradient from A to B is G=(S_B-S_A) / d, where d is the distance between the grid centers. To determine the diffusion direction, for each boundary grid, the direction with the largest gradient value is selected as the main diffusion direction of that grid. The diffusion directions of all boundary grids in the same risk area are vector-superimposed to obtain the overall main diffusion direction vector D_major of that risk area. Calculate the diffusion rate V_diffusion, which is the average movement speed of the risk area boundary over the most recent N time windows; Generate a diffusion prediction based on the diffusion direction D_major and diffusion velocity V_diffusion, predicting the movement trajectory of the risk boundary within the next 1-5 minutes.

[0029] Example 2: A real-time monitoring system for low-altitude navigation environment using a multi-source fusion sensing terminal, such as... Figure 1 As shown, it includes: Synchronization Time Unit: Access to BeiDou satellite time synchronization provides a unified clock reference for all sensing modules, controls the timestamp error of all multi-source data acquisition to the microsecond level, and outputs high-precision timestamps for backend multi-source data association and alignment; Spectrum detection module: Covers low-altitude navigation frequency bands and general radio frequency bands, scans and monitors the airspace spectrum status in real time, collects frequency band occupancy, interference intensity, interference bandwidth, sudden spectrum disturbances, and interference azimuth information, identifies narrowband, broadband, and intermittent radio frequency interference, and has a built-in reference noise source component to support periodic self-calibration based on the environmental spectrum noise floor to compensate for receiver link gain drift. High-sensitivity acoustic detection module: Equipped with a noise-reducing pickup array, it accurately collects low-altitude rotor noise, aircraft cluster sound field, operational abnormal noise, and sudden environmental noise, outputs standardized voiceprint feature data, and has a built-in standard electrical signal calibration circuit that periodically injects calibration signals of known frequency and amplitude to correct microphone sensitivity and frequency response curve deviation. Navigation Sensing Module: Fusion calculation of BeiDou, GNSS and inertial navigation, real-time output of navigation quality parameters such as signal strength, number of visible satellites, positioning error, accuracy factor, signal continuity, and multi-mode switching stability, supports dual-antenna / multi-constellation redundant observation, and can self-check and report the confidence level of the terminal's position calculation; RID identification module: Real-time monitoring and parsing of aircraft broadcast RID data, extracting unique device ID, aircraft type, location, flight permission, take-off and landing information; and marking targets without RID, with tampered RID, or with abnormal RID. It has a built-in RID signal parsing confidence evaluation unit and outputs link quality indicators such as signal strength and bit error rate. Main control and data transmission unit: includes edge side lightweight processing module, which can perform on-chip feature extraction on the original spectrum and acoustic data, specifically including: (1) spectrum feature extraction submodule: performs fast Fourier transform on the spectrum sampling data, calculates the energy ratio, peak frequency and amplitude, mean background noise and time series change rate of each preset frequency band, and outputs spectrum feature vector; (2) acoustic feature extraction submodule: performs frame windowing processing on the acoustic sampling data, calculates short-time zero crossing rate, short-time energy, sub-band energy distribution and time domain envelope features, and optionally extracts Mel frequency cepstral coefficients as voiceprint recognition features; (3) the lightweight feature vector extracted on-chip is uploaded to the cloud through 4G / 5G / network port, the original data can be cached locally for traceability and evidence collection, thereby reducing the uplink data bandwidth requirement, uniformly scheduling the parallel acquisition of each module, uniform protocol encapsulation, and uploading multi-source synchronous data to the cloud in real time through 4G / 5G / network port, while reporting the self-calibration status of each sensor and data quality label; Power supply and protection unit: Supports mains power, DC power, and lithium battery backup power, and has lightning protection, static electricity protection, overvoltage and undervoltage protection, meeting the requirements of 24 / 7 outdoor uninterrupted operation.

[0030] The above are merely preferred embodiments of the present invention, but the scope of protection of the present invention is not limited thereto. Any equivalent substitutions or modifications made by those skilled in the art within the scope of the technology disclosed in the present invention, based on the technical solution and inventive concept of the present invention, should be covered within the scope of protection of the present invention.

Claims

1. A method for real-time monitoring of the low-altitude navigation environment using a multi-source fusion sensing terminal, characterized in that, Includes the following steps: S1: Divide the controlled low-altitude airspace into grids in the horizontal and vertical directions, assign a unique code to each grid, and build a gridded digital base for the entire airspace. S2: Within each grid, collect four types of core data, including: grid navigation signal data, grid spectrum interference data, grid acoustic anomaly data, and grid RID aircraft data; S3: Calculate data quality labels for the four types of core data collected, calculate the comprehensive credibility based on the data quality labels, and output the four types of core data and their corresponding comprehensive credibility for each grid. S4: Establish a four-dimensional weighted algorithm to calculate the navigation quality factor, spectrum interference factor, acoustic disturbance factor, and RID density factor based on the four types of core data collected. Construct a dynamic adaptive weighted fusion engine and calculate the fusion weights of the navigation quality factor, spectrum interference factor, acoustic disturbance factor, and RID density factor based on the comprehensive credibility of the four types of core data. Finally, calculate the spatiotemporal comprehensive navigation safety score of each grid. S5: Identify and label the root cause categories of navigation anomalies based on navigation quality factor, spectral interference factor, acoustic disturbance factor, and RID density factor; S6: Perform full-domain situational visualization and simulation of the controlled low-altitude airspace, including: generating a low-altitude navigation heat map and automatically marking navigation blind spots, high-incidence interference areas, and high-disturbance grid areas; storing the visualized operation data according to the time axis and recording the root cause categories of navigation anomalies; and finally, predicting the short-term trend simulation for the next 1-5 minutes based on the time-series change rate of the spatiotemporal integrated navigation safety score of each grid.

2. The method for real-time monitoring of low-altitude navigation environment using a multi-source fusion sensing terminal according to claim 1, characterized in that: Step S1 includes: S11: Using the WGS-84 geodetic coordinate system as the spatial reference, horizontal position is described by latitude and longitude coordinates, and vertical position is described by altitude. S12: The horizontal range of the controlled low-altitude airspace is progressively meshed along the longitude and latitude directions. The specific steps are as follows: First, set the longitude range [Lon_min, Lon_max] and latitude range [Lat_min, Lat_max] of the controlled airspace. Second, according to the application scenario and monitoring accuracy requirements, calculate the longitude step size and latitude step size of the planar mesh. The calculation formula is as follows: ΔLon=GridSize / (111320×cos(Lat)); ΔLat = GridSize / 111320; Where ΔLon is the longitude step, Lat is the latitude, ΔLat is the latitude step, GridSize is the side length of the planar grid, and 111320 is the number of meters corresponding to 1° latitude near the Earth's equator; S13: Divide the controlled area into M columns along the longitude direction with a step size ΔLon. The controlled area is divided into N rows along the latitudinal direction with a step size ΔLat. ,in, This represents the rounding operation. Each planar grid is identified by its row index i and column index j, where 1≤i≤N, 1≤j≤M, and its longitude range is [Lon_min+(j-1)×ΔLon, Lon_min+j×ΔLon], and its latitude range is [Lat_min+(i-1)×ΔLat, Lat_min+i×ΔLat]; S14: In the vertical dimension, set 0 meters as the lowest altitude H_min and 1000 meters as the highest altitude H_max. Then, divide the airspace into K altitude layers at fixed altitude intervals ΔH. The ΔH settings are as follows: coarse-grained ΔH is 100 meters per layer, medium-grained ΔH is 50 meters per layer, and fine-grained ΔH is 25 meters per layer. The height of each layer is k, and 1≤k≤K. S15: Combines the planar mesh with the height layer using Cartesian product to generate a three-dimensional mesh. A complete mesh is uniquely determined by a triple (i, j, k). Each mesh is assigned a globally unique mesh code, supporting fast indexing and spatial retrieval.

3. The method for real-time monitoring of the low-altitude navigation environment using a multi-source fusion sensing terminal according to claim 1, characterized in that, Step S2 includes: S21: The grid navigation signal data includes: the average of the carrier-to-noise ratios of all visible satellite signals reported by all terminals as the grid navigation signal strength; the maximum value of the horizontal or vertical positioning error reported by all terminals as the grid positioning error; the average of the horizontal and vertical accuracy factors reported by all terminals as the grid accuracy factor; and the calculation result of 1 - (number of lock-outs / total number of observation epochs) as the signal continuity. S22: The spectral interference data within the grid includes: interference intensity, calculated using the following formula: ； Where A_p is the amplitude value of the p-th frequency point, N_floor is the estimated noise floor value, and P is the total number of frequency points; The number of sudden events in which the spectral interference intensity exceeds the preset threshold by 3dB within the current 60-second time window is counted as the number of disturbances. The criteria for judging a sudden event are: the interference intensity of more than 3 consecutive sampling points exceeds the preset threshold and the duration is not less than 100ms. The total duration of all spectrum interference events within the current time window is calculated as the interference duration. If the interference is still ongoing, it is accumulated to the current time. S23: The grid acoustic anomaly data includes: acoustic anomaly intensity = E_current / E_baseline, where E_current is the short-term energy average of the current time window, and E_baseline is the average noise floor energy of the historical normal period; The perturbation feature is defined as R1·σ_ZCR + R2·ΔE_sub, where σ_ZCR is the variance of the zero-crossing rate within the current time window, ΔE_sub is the consistency change of the subband energy distribution relative to the historical baseline, and R1 and R2 represent the weights of the zero-crossing rate variance and the consistency change of the subband energy distribution, respectively. The weights satisfy R1 + R2 = 1. S24: The grid RID aircraft data includes: the total number of deduplicated aircraft whose broadcast positions fall within the grid range within the current time window, which is taken as the number of RID aircraft in the grid; Cluster density = number of aircraft in the grid / grid volume, where grid volume = grid area × altitude layer thickness, and the unit is aircraft / cubic kilometer; Flight activity = (Number of moving aircraft / Total number of aircraft) × Normalized average speed.

4. The method for real-time monitoring of low-altitude navigation environment using a multi-source fusion sensing terminal according to claim 3, characterized in that, Step S3 includes: S31: The data quality labels corresponding to each type of core data include: signal-to-noise ratio, effective acquisition duration, sensor self-calibration status, and data integrity identifier; S32: The signal-to-noise ratio (SNR)_Index is calculated as follows: For navigation signal data, the average carrier-to-noise ratio (C / No_avg) of visible satellites reported by each terminal in the grid is normalized to the range of 0 to 1 to obtain SNR_Index. When C / No_avg is below 30dB-Hz, SNR_Index is 0; when C / No_avg is above 50dB-Hz, SNR_Index is 1; when C / No_avg is between 30dB-Hz and 50dB-Hz, SNR_Index is calculated by linear interpolation, i.e., SNR_Index = (C / No_avg - 30) / (50 - 30); For spectral interference data, the SNR_Index is obtained by normalizing the ratio of the average signal amplitude to the average noise floor within the current frequency band; for acoustic anomaly data, the SNR_Index is obtained by normalizing the ratio of the current signal energy to the electronic noise floor energy; for RID aircraft data, the SNR_Index is obtained by normalizing the average RID signal strength of all aircraft within the grid. S33: The effective acquisition time Valid_Duty is calculated as follows: Valid_Duty = Total effective acquisition time within the grid / Total expected acquisition time within the grid, with a value of 0 to 1; S34: The calculation method of the sensor self-calibration status Calib_Status is as follows: Each terminal performs self-calibration periodically and reports the calibration status code. The calibration status code 0x00 indicates that the calibration is normal and the deviation is within the normal range. The calibration status code 0x01 indicates that the calibration deviation exceeds the limit but has not yet affected the validity of the data. The calibration status code 0x02 indicates that the calibration deviation is serious and the data needs to be used with reduced weight. The calibration status code 0x03 indicates that the sensor is faulty and the data is unavailable. The calibration status code is mapped to a continuous value from 0 to 1. 0x00 is mapped to 1.0, 0x01 is mapped to 0.6, 0x02 is mapped to 0.2, and 0x03 is mapped to 0. When there are multiple terminals in the grid, the average value of the mapped values ​​of each terminal is taken as Calib_Status. S35: The calculation method of the data integrity flag Integrity_Flag is as follows: CRC check, sequence number continuity check, and timestamp rationality check are performed on each frame of reported data; four aspects of verification are performed and identified as a combination of binary status bits: Bit0 indicates CRC check passed, and the value is 1 for pass; Bit1 indicates sequence number continuity, and the value is 1 for continuity; Bit2 indicates timestamp is normal, and the value is 1 for normal; Bit3 indicates data logic consistency, and the value is 1 for consistency; if all verifications pass, the Integrity_Flag value is 1.0; if some verifications pass, the Integrity_Flag value is the number of passed items / 4; if all verifications fail, the Integrity_Flag value is 0. S36: Calculate the overall trustworthiness weight based on the data quality label, including: when there is a single sensing terminal in any grid, the formula for calculating the overall trustworthiness Trust_Source of each type of core data is as follows: Trust_Source=α·SNR_Index+β·Valid_Duty+γ·Calib_Status+δ·Integrity_Flag; Where α, β, γ, and δ are preset weight coefficients that satisfy α+β+γ+δ=1, and Source is any one of Nav, Spec, Acou, and RID, which represent the comprehensive credibility of the core data of the corresponding class. When multiple sensing terminals simultaneously report the same type of core data within any given grid, the overall trustworthiness of the same type of core data from the nth terminal is calculated. Terminal data with an overall trustworthiness below 0.3 is marked as unusable and removed. The overall trustworthiness Trust_Source for each type of core data is then calculated using the following formula: Trust_Source=Σ(Source_n×Trust_Source_n) / ΣTrust_Source_n; Wherein, Source_n represents any type of core data reported by the nth terminal, and Trust_Source_n=α·SNR_Index_n+β·Valid_Duty_n+γ·Calib_Status_n+δ·Integrity_Flag_n, where SNR_Index_n, Valid_Duty_n, Calib_Status_n, and Integrity_Flag_n are the data quality labels of any type of core data collected by the nth terminal, respectively. S37: Based on the four types of core data values ​​of each grid and their corresponding comprehensive trustworthiness, output the following vectors: [Navigation data value, Trust_Nav], [Spectrum data value, Trust_Spec], [Acoustic data value, Trust_Acou], [RID data value, Trust_RID].

5. The method for real-time monitoring of the low-altitude navigation environment using a multi-source fusion sensing terminal according to claim 4, characterized in that, Step S4 includes: S41: Calculate the navigation quality factor F_nav, using the following formula: F_nav=w1·S_strength+w2·S_accuracy+w3·S_DOP+w4·S_continuity, Among them, w1, w2, w3, and w4 are the preset weights of the sub-indicators, and the sum of the four weights is 1. The ratio of the measured value of the navigation signal strength within the grid to the nominal normal value is the signal strength sub-index S_strength; The formula for calculating the positioning accuracy sub-index S_accuracy is: S_accuracy=1 / (1+exp(-k1·(HPE_threshold-HPE))); Where HPE is the measured value of the horizontal positioning error within the grid, in meters, calculated and output by the navigation perception module, representing the deviation between the estimated horizontal position of the aircraft or terminal device within the grid and its true position; HPE_threshold is the positioning error tolerance threshold; k1 is the steepness coefficient; exp is the natural exponential function; The formula for calculating the precision factor sub-index S_DOP is: S_DOP=max(0,1-HDOP / HDOP_max); Where HDOP_max is the maximum allowed precision factor, and S_DOP returns to zero when HDOP exceeds HDOP_max; HDOP is the measured value of the horizontal precision factor within the grid. The signal continuity sub-index S_continuity adopts the signal continuity value described in step S2; S42: Calculate the spectral interference factor F_spec, using the following formula: F_spec = 1 - ( v1·(1-I_intensity) + v2·(1-I_frequency) + v3·(1-I_duration) ); Where v1, v2, and v3 are the preset weights of the sub-indices, and the interference intensity sub-indice I_intensity = exp(-λ·Intensity), where Intensity is the measured value of the interference intensity and λ is the attenuation coefficient. The disturbance frequency sub-index I_frequency = max(0, 1-Count / Count_max), where Count is the number of disturbances in the current time window, Count_max is the maximum tolerable number of disturbances, and I_frequency returns to zero when Count exceeds Count_max; The interference duration sub-index is I_duration=max(0,1-Duration / Duration_max), where Duration is the total duration of the interference, Duration_max is the maximum tolerable duration, and I_duration returns to zero when Duration exceeds Duration_max. S43: Calculate the acoustic disturbance factor F_acou, using the following formula: F_acou= u1·A_intensity+u2·A_disturbance; Where u1 and u2 are the preset weights of the sub-indicators; The anomaly intensity sub-index A_intensity adopts a piecewise linear mapping: it takes 1.0 when the anomaly intensity is ≤1.0, linearly decreases to 0.3 when the anomaly intensity is between 1.0 and 3.0, and takes 0.3 when the anomaly intensity is >3.0, to avoid the factor from suddenly dropping to zero due to short-term extreme noise; The voiceprint disturbance sub-index A_disturbance=max(0,1-Disturbance_norm / Disturbance_max), where Disturbance_norm is the measured normalized value of the disturbance feature, and Disturbance_max is the maximum tolerable disturbance value; when Disturbance_norm exceeds Disturbance_max, A_disturbance returns to zero. S44: Calculate the RID density factor F_rid, using the following formula: F_rid=t1·D_density+t2·D_activity+t3·D_violation; Where t1, t2, and t3 are the preset weights of the sub-indicators; The density sub-index D_density = max(0, 1-Density / Density_max), where Density is the cluster density, Density_max is the maximum tolerable density, and D_density returns to zero when Density exceeds Density_max. The activity sub-index is D_activity = 1 - Activity, where Activity is the measured value of flight activity, ranging from 0 to 1. The sub-indicator for the proportion of violations is D_violation = 1 - Violation_Ratio, where Violation_Ratio is the proportion of aircraft that violated regulations, and its value ranges from 0 to 1. S45: The method for calculating the fusion weight is as follows: ； ； ； Where p is the adjustment factor; W_nav, W_spec, W_acou, and W_rid are the fusion weights of the navigation quality factor F_nav, the spectral interference factor F_spec, the acoustic disturbance factor F_acou, and the RID density factor F_rid, respectively; Z is the normalization constant, ensuring that the sum of the four weights is 1, and the calculation formula is as follows: ； When Trust_Source < 0.3 for a certain type of data, the dynamic adaptive weighted fusion engine determines that the data source is basically unusable, multiplies its fusion weight coefficient by a penalty factor of 0.1, and renormalizes the other three factors proportionally to ensure that the overall evaluation is not misled by severely poor data. S46: Calculate the spatiotemporal integrated navigation safety score for each grid.

6. The method for real-time monitoring of the low-altitude navigation environment using a multi-source fusion sensing terminal according to claim 5, characterized in that, Step S46 further includes: The spatiotemporal integrated navigation safety score S_grid for each grid is calculated using the following formula: S_grid=λ1·S_instant+λ2·S_mean+λ3·(1-S_var); Where λ1, λ2, and λ3 are dimension weights, and the sum of the three weights is 1; Transient score S_instant = W_nav·F_nav + W_spec·F_spec + W_acou·F_acou + W_rid·F_rid; moving average t is the current time, and T is the number of sampling points; variance .

7. The method for real-time monitoring of the low-altitude navigation environment using a multi-source fusion sensing terminal according to claim 6, characterized in that, Step S5 includes: identifying and labeling the root cause category of the navigation anomaly, specifically the following root cause categories: S51: Signal attenuation caused by terrain / building obstruction. The following conditions must be met to be classified as signal attenuation: navigation quality factor F_nav < 0.40, spectral interference factor F_spec ≥ 0.55, acoustic disturbance factor F_acou ≥ 0.50, and RID density factor F_rid ≥ 0.

50. S52: Interference degradation caused by radio frequency disturbances is classified as interference degradation if the following conditions are met: spectral interference factor F_spec < 0.40; navigation quality factor F_nav < 0.50; acoustic disturbance factor F_acou ≥ 0.

50. S53: Environmental disturbances caused by acoustic anomalies and equipment vibration interference are classified as such if the following conditions are met: acoustic disturbance factor F_acou < 0.40; spectral interference factor F_spec ≥ 0.55; navigation quality factor F_nav ≥ 0.

50. S54: Situational overload caused by excessively high aircraft cluster density. The following conditions must be met to be classified as situational overload: RID density factor F_rid < 0.35; navigation quality factor F_nav, spectral interference factor F_spec, and acoustic disturbance factor F_acou are all ≥ 0.

50.

8. The method for real-time monitoring of low-altitude navigation environment using a multi-source fusion sensing terminal according to claim 7, characterized in that, Step S6 includes: S61: Obtain the following data for all grids at a fixed refresh cycle of 1 second: the spatiotemporal integrated navigation safety score S_grid for each grid, the four factor values ​​F_nav, F_spec, F_acou, and F_rid for each grid, and the root cause category of navigation anomalies for each grid; all grid data are organized in the form of a two-dimensional matrix, where the rows of the matrix correspond to the grid index i in the latitude direction, the columns correspond to the grid index j in the longitude direction, and the matrix element value is the S_grid value of that grid; Each grid's S_grid value is mapped to an RGB color value using a green-yellow-red gradient. For S_grid values ​​between two color gradient boundaries, linear interpolation is used to calculate the RGB components. The interpolation formula is as follows: Let the target value S be located within the interval [S_min, S_max], where S_min is the minimum target value S and S_max is the maximum target value S. The starting color of the interval is RGB_start and the ending color is RGB_end. Then, the interpolation ratio ratio = (S - S_min) / (S_max - S_min), with a value of 0 to 1; R = R_start + ratio × (R_end - R_start); R_start and R_end are the red R component values ​​of the starting and ending colors of the color interval where the current S_grid value is located, respectively. The G and B components in the RGB components are also calculated in the same way as the R components. R, G, and B are the red, green, and blue components, respectively, with each component ranging from 0 to 255. For grids without terminal coverage, they are displayed in gray and marked "No Data" in the legend. Using a GIS rendering engine, the color-mapped grid data is overlaid and displayed on a low-altitude electronic map. Each grid is rendered as a rectangular block, the location of which is determined by the latitude and longitude range of the grid, and the color of the block is determined by the S_grid value mapping. The final result is a low-altitude navigation heat map, with a refresh cycle synchronized with the data update cycle of 1 second. S62: When three or more consecutive adjacent grids have F_nav < 0.25 and S_grid < 0.30, and this state lasts for more than 60 seconds, while the spectral interference factor F_spec ≥ 0.55, it can be identified as a navigation blind zone. On the low-altitude navigation heatmap, the blind zone grid area is covered with a blue diagonal stripe overlay, and the boundary of the area is outlined with a thick blue dashed line. The navigation blind zone text label and the size of the blind zone are marked. Two grids are extended outward to set a buffer zone label at the edge of the blind zone, indicating a warning message that the blind zone is about to be entered. The label text is "Entering the blind zone is about to begin." When a grid continuously meets the blind zone conditions for more than 60 seconds, it is included in the blind zone. When a grid no longer meets the blind zone conditions and continues to do so for more than 30 seconds, it is removed from the blind zone to avoid frequent fluctuations in the blind zone boundary. S63: Condition A: Persistent interference: F_spec < 0.30 for 3 or more consecutive adjacent grids, and this state lasts for more than 30 seconds; Condition B: Intermittent interference: In the past 10 minutes, the cumulative number of spectral disturbances in a single grid exceeds 10, or the cumulative total duration of interference exceeds 300 seconds; If either condition A or B is met, it is determined to be a high-incidence area of ​​interference. On the heat map, the grid area of ​​the high-incidence area of ​​interference is covered by a red diagonal overlay, the boundary of the area is outlined with a red solid line, and the text label of high-incidence area of ​​interference is marked. The specific interference type and interference intensity level are superimposed within the area. S64: When the S_var of a single grid is greater than 0.08 and S_grid is greater than 0.40, and the ratio of the average S_var of the grid to that of its neighboring grids is greater than 2.0, it is identified as a high-disturbance grid region. The grid is highlighted on the heatmap with an orange flashing outer glow. The center of the grid is labeled with a text label and the current variance value. At the same time, the names of the factors that are the main sources of disturbance are also superimposed. S65: Store the visualized operational data according to the timeline and record the root cause categories of navigation anomalies; S66: Prediction of the temporal change rate of the spatiotemporal integrated navigation safety score of each grid, and short-term trend projection for the next 1-5 minutes.

9. The method for real-time monitoring of the low-altitude navigation environment using a multi-source fusion sensing terminal according to claim 8, characterized in that, Step S66 further includes: S661: Single-grid short-term trend prediction, the specific steps are as follows: Calculate the historical rate of change, take the S_grid value sequence of the current time t and the previous historical time, use the least squares method to linearly fit, and obtain the rate of change r(t) of the sequence and the fitting residual ε. The fitting formula is: S_grid(t')=a+b·t', where a is the intercept of the fitted line, b is the rate of change r(t), and t' is the time variable; Calculate the prediction confidence score. Based on the fitting residual ε and the historical volatility S_var of S_grid, calculate the prediction confidence score C_pred: C_pred=1-[ε_norm+S_var_norm] / 2, where ε_norm is the normalized fitting residual and S_var_norm is the normalized variance, both of which are mapped to the interval 0 to 1. Perform multi-duration prediction, let the prediction duration be T1, the unit be minutes, and the value range be 1 to 5 minutes. Corresponding to the short-term trend extrapolation of the future 1-5 minutes described in step S6, the prediction value S_pred(t+T)=S_grid(t)+r(t)·T. At the same time, set the physical boundary constraint of the prediction value: S_pred is truncated to the interval [0,1]. S662: Risk grid diffusion direction prediction, the specific steps are as follows: identify risk boundary grids, mark all grids with S_grid<0.50 as risk grids, extract the outer boundary grids of the risk grid set, that is, risk grids with at least one adjacent non-risk grid, as the target grids for diffusion analysis; Calculate the S_grid gradient at the boundary. For each boundary grid, calculate the S_grid gradient value pointing to the adjacent non-risk grid. The S_grid value of boundary grid A is S_A, and the S_grid value of its adjacent non-risk grid B is S_B. Then the gradient from A to B is G=(S_B-S_A) / d, where d is the distance between the grid centers. To determine the diffusion direction, for each boundary grid, the direction with the largest gradient value is selected as the main diffusion direction of that grid. The diffusion directions of all boundary grids in the same risk area are vector-superimposed to obtain the overall main diffusion direction vector D_major of that risk area. Calculate the diffusion rate V_diffusion, which is the average movement speed of the risk area boundary over the most recent N time windows; Generate a diffusion prediction based on the diffusion direction D_major and diffusion velocity V_diffusion, predicting the movement trajectory of the risk boundary within the next 1-5 minutes.

10. A real-time monitoring system for low-altitude navigation environment using a multi-source fusion sensing terminal, characterized in that, Synchronization Time Unit: Access to BeiDou satellite time synchronization provides a unified clock reference for all sensing modules, controls the timestamp error of all multi-source data acquisition to the microsecond level, and outputs high-precision timestamps for backend multi-source data association and alignment; Spectrum detection module: Covers low-altitude navigation frequency bands and general radio frequency bands, scans and monitors the airspace spectrum status in real time, collects frequency band occupancy, interference intensity, interference bandwidth, sudden spectrum disturbances, and interference azimuth information, identifies narrowband, broadband, and intermittent radio frequency interference, and has a built-in reference noise source component to support periodic self-calibration based on the environmental spectrum noise floor to compensate for receiver link gain drift. High-sensitivity acoustic detection module: Equipped with a noise-reducing pickup array, it accurately collects low-altitude rotor noise, aircraft cluster sound field, operational abnormal noise, and sudden environmental noise, outputs standardized voiceprint feature data, and has a built-in standard electrical signal calibration circuit that periodically injects calibration signals of known frequency and amplitude to correct microphone sensitivity and frequency response curve deviation. Navigation Sensing Module: Fusion calculation of BeiDou, GNSS and inertial navigation, real-time output of navigation quality parameters such as signal strength, number of visible satellites, positioning error, accuracy factor, signal continuity, and multi-mode switching stability, supports dual-antenna / multi-constellation redundant observation, and can self-check and report the confidence level of the terminal's position calculation; RID identification module: Real-time monitoring and parsing of aircraft broadcast RID data, extracting unique device ID, aircraft type, location, flight permission, take-off and landing information; and marking targets without RID, with tampered RID, or with abnormal RID. It has a built-in RID signal parsing confidence evaluation unit and outputs link quality indicators such as signal strength and bit error rate. Main control and data transmission unit: includes edge-side lightweight processing module, which can perform on-chip feature extraction on the original spectrum and acoustic data, specifically including: (1) spectrum feature extraction submodule: performs fast Fourier transform on the spectrum sampling data, calculates the energy ratio, peak frequency and amplitude, mean background noise and time series change rate of each preset frequency band, and outputs spectrum feature vector; (2) acoustic feature extraction submodule: performs frame windowing processing on the acoustic sampling data, calculates short-time zero crossing rate, short-time energy, sub-band energy distribution and time domain envelope features, and optionally extracts Mel frequency cepstral coefficients as voiceprint recognition features; (3) The lightweight feature vector extracted on the chip is uploaded to the cloud via 4G / 5G / network port. The original data can be cached locally for traceability and evidence collection, thereby reducing the uplink data bandwidth requirement. The parallel collection of each module is uniformly scheduled and the protocol is uniformly encapsulated. The multi-source synchronous data is uploaded to the cloud in real time via 4G / 5G / network port, and the self-calibration status of each sensor and data quality label are reported at the same time. Power supply and protection unit: Supports mains power, DC power, and lithium battery backup power, and has lightning protection, anti-static protection, overvoltage and undervoltage protection, meeting the requirements of 24 / 7 outdoor uninterrupted operation; To realize the method for real-time monitoring of low-altitude navigation environment of multi-source fusion sensing terminal as described in any one of claims 1-9.

Images