Method for measuring three-dimensional coordinates in tunnel based on multiple sensors

Abstract

The invention discloses a method for measuring three-dimensional coordinates in a tunnel based on multiple sensors, and the method comprises the steps: S1, collecting tunnel data, completing the hardware deployment of a positioning system, carrying out the data collection and calibration, and constructing a basic environment of a digital twin system; s2, introducing a measurement quality factor Q to comprehensively score the reliability of each observation value; s3, calculating a comprehensive cost function of the candidate triangular pyramid group, and selecting the candidate triangular pyramid group with the minimum cost function as calculation input; and S4, writing a signal module worn by a worker or an airborne signal module into a digital twin system, issuing the signal module or the airborne signal module, executing whole-network time synchronization, starting a tunnel three-dimensional positioning system, selecting four base stations in two positioning base station groups to form a composite triangular pyramid with the signal module according to the covariance matrix in the S1 and the comprehensive cost function in the S3, and transmitting the composite triangular pyramid. Calculating a three-dimensional coordinate of the signal module; and S5, edge data processing and early warning.

Description

TECHNICAL FIELD

[0001] The present application belongs to the technical field of railway system engineering design, and particularly relates to a three-dimensional coordinate measurement method in a tunnel based on multiple sensors. BACKGROUND

[0002] Nowadays, the railway system is developing rapidly, and as an important part of railway engineering, the construction and operation of tunnels have very high requirements for safety and efficiency. The traditional construction site management method relies on manual inspection and clock-in, which has strong hysteresis, relies on human subjective judgment, and has limited coverage. In actual operation, workers need to frequently operate in narrow, complex, and high-risk tunnel environments, and also need to work with large machinery, so any negligence may cause collisions, entering dangerous areas, or other safety accidents.

[0003] Most existing positioning devices can only work effectively within a short distance range, and are easily affected by multipath effects, cumulative errors, and other factors in long tunnels, making it difficult to achieve high-precision, long-distance, and continuous three-dimensional positioning of personnel and machinery. This not only cannot guarantee the safety of the construction process, but also limits real-time monitoring and risk assessment of worker status and machinery operation dynamics. Therefore, there is an urgent need for a system that can measure the three-dimensional position of workers and engineering machinery in long tunnels, combined with trajectory monitoring and danger warning, to effectively improve construction safety and management level. SUMMARY

[0004] The purpose of the present application is to provide a three-dimensional coordinate measurement method in a tunnel based on multiple sensors to achieve high-precision, continuous, and stable three-dimensional measurement of people and machines in a tunnel environment, and to improve construction and operation safety.

[0005] The technical solution adopted by the present application is a three-dimensional coordinate measurement method in a tunnel based on multiple sensors, comprising the following steps:

[0006] Step S1, collect tunnel data, complete hardware deployment of the positioning system, and perform data collection and calibration to build a digital twin system basic environment;

[0007] Step S2, introduce a measurement quality factor Q to comprehensively score the reliability of each observation value;

[0008] Step S3, calculate the comprehensive cost function of the candidate three-prism group, and select the candidate three-prism group with the smallest cost function as the calculation input;

[0009] Step S4, the signal module worn by the worker or the airborne signal module is written into the digital twin system and issued, the network time synchronization is performed, the tunnel three-dimensional positioning system is started, and according to the covariance matrix in S1 and the comprehensive cost function in S3, four base stations in the two groups of positioning base stations are selected to form a composite tri-prism with the signal module, and the three-dimensional coordinates of the signal module are calculated;

[0010] Step S5, edge data processing and early warning.

[0011] Further, the specific steps of S1 are as follows:

[0012] S11, according to the tunnel data, the positioning base station network is deployed, and the positioning base station group is arranged along the tunnel axis;

[0013] S12, after the deployment is completed, the unique label, installation height, orientation and surrounding environment data of each base station are recorded one by one, and all are input into the digital twin system to complete the initial state modeling;

[0014] S13, the installed positioning system is tested, the relative coordinates between the base stations and the clock bias between each other are estimated, and the base station coordinate table and the geometric precision factor distribution map are output.

[0015] Further, in S13, the geometric precision factor calculation method is as follows:

[0016] In the three-dimensional space of the tunnel, first, the target position to be positioned is determined as , and the positions of the first and the second base stations are , the ranging observation equation is listed and linearized at the approximate position to obtain the observation matrix of the positioning target in the tunnel ;

[0017] Then, according to the variance of single ranging error , the covariance matrix of position estimation and the geometric precision factor GDOP are calculated, and the formula is as follows:

[0018] ;

[0019] ;

[0020] In the formula, represents transposition, is the trace of the matrix.

[0021] Further, in S2, the edge node converts the CIR into a two-dimensional time-delay-energy spectrum graph, classifies via a lightweight convolutional neural network, and outputs the probability of the existence of a direct diameter If the LOS is determined, the SAGE algorithm is used to iteratively estimate the first path arrival time delay; if the NLOS is determined, the extracted time delay is modified by the Bayesian shrinkage compensation method, and the angle of arrival (AOA) and its confidence are obtained, and finally the measurement quality factor Q is calculated, as follows:

[0022] ;

[0023] wherein, is the signal-to-noise ratio, is the standard deviation of the angle of arrival, is the historical deviation, , , , is the weighting coefficient, is the natural logarithm, is the signal-to-noise ratio smoothing coefficient, represents the reference variance value measured under ideal line-of-sight conditions, , represents the probability of the existence of the direct path.

[0024] Further, in the S3, the calculation formula of the cost function is as follows:

[0025] ;

[0026] wherein, , is the weighting coefficient, is the candidate base station set, is the number of base stations in the candidate base station set, is the geometric dilution of precision, is the measurement quality factor of the i-th base station.

[0027] Further, the specific steps of the S5 are as follows:

[0028] S51, the edge node pre-processes the original data to generate a compressed observation package with quality annotation and reports to the central processing layer;

[0029] S52, based on the covariance matrix obtained in step S1 and the positioning terminal coordinates output by the center in step S4, the original positioning result is trajectory smoothed, and the behavior of collision or entering the danger zone is predicted, and then graded early warning is performed according to the set threshold.

[0030] Further, the specific steps of the S52 are as follows:

[0031] S52a, based on the coordinate change of the smoothed three-dimensional coordinates of the signal module , the velocity vector ​And according to the velocity vector, the short-time prediction coordinates of the target are calculated, and when the target has a significant acceleration change, an acceleration compensation term is automatically introduced, and the calculation formula is as follows:

[0032] ;

[0033] Wherein, is the predicted coordinates of the target, is the target acceleration, is the prediction step;

[0034] S52b, based on the predicted point , the minimum distance between the target prediction point and the boundary of the dangerous area and other moving targets is calculated, if the minimum distance is less than the safety distance threshold, it is determined as a potential risk event, and a warning is made;

[0035] When the predicted coordinates enter the threshold boundary, a first-level warning is carried out; when the actual coordinates enter the threshold boundary, a second-level warning is carried out; when the predicted coordinates enter the dangerous area, a third-level warning is carried out.

[0036] The beneficial effects of the present application are:

[0037] 1、The base station layout and coordinate calculation method of the present application upgrades the two-dimensional positioning capability of the traditional UWB to three-dimensional positioning capability.

[0038] 2、The present application avoids positioning errors caused by base station geometric degradation and multipath interference in the tunnel environment through the joint optimization mechanism of GDOP and quality factor Q based on the three-pyramid group, and can output high-precision three-dimensional coordinates with covariance.

[0039] 3、The present application introduces a multi-positioning base station group and an adaptive base station switching strategy, which realizes full coverage of the positioning area in the tunnel according to the long and narrow characteristics of the tunnel, and can still maintain continuous and stable three-dimensional positioning even if there is shielding, signal anomaly or multipath interference.

[0040] 4、The present application not only realizes real-time three-dimensional position measurement of workers and engineering machinery, but also combines trajectory anomaly detection and dangerous area warning, which can detect potential risks in advance and significantly improve the safety guarantee level of the tunnel construction and operation process. BRIEF DESCRIPTION OF DRAWINGS

[0041] In order to more clearly illustrate the technical solutions in the embodiments of the present application or the prior art, the following will briefly introduce the drawings needed to be used in the embodiment or prior art description. Obviously, the drawings in the following description are only some embodiments of the present application, and other drawings can also be obtained by those skilled in the art without creative labor.

[0042] Figure 1is a flow chart of the present application.

[0043] Figure 2 is a system arrangement flow chart of the present application.

[0044] Figure 3 is a three-dimensional positioning core solution flow chart of the present application.

[0045] Figure 4 is a three-prism ranging method principle diagram of the present application.

[0046] Figure 5 is a three-dimensional positioning method schematic diagram of the composite three-prism group in the present application.

[0047] Figure 6 is a safety monitoring and rescue flow chart of the present application. DETAILED DESCRIPTION

[0048] The technical solutions in the embodiments of the present application will be clearly and completely described below with reference to the drawings in the embodiments of the present application. Obviously, the described embodiments are only part of the embodiments of the present application, rather than all the embodiments. Based on the embodiments in the present application, all other embodiments obtained by those skilled in the art without creative labor fall within the scope of protection of the present application.

[0049] EMBODIMENT

[0050] The embodiments of the present application provide a three-dimensional coordinate measurement method in a tunnel based on multiple sensors, and a flow chart is as shown in Figure 1 The specific steps are as follows:

[0051] Step S1, tunnel data (exit and entrance positions, tunnel center line, cross section size, internal facilities, dangerous areas, construction machinery operation area, etc.) is collected, hardware deployment of the positioning system is completed, and data collection and calibration are performed to construct a digital twin system basic environment, and a system arrangement flow chart is as shown in Figure 2 The specific steps are as follows:

[0052] S11, positioning base station network deployment is performed according to the tunnel data, UWB positioning base station groups are arranged along the tunnel axis at an interval of 10-30 m, each positioning base station group is composed of three positioning base stations, forming a triangle to ensure that any point has at least four base stations visible. As shown in Figure 5 The three positioning base stations of the previous positioning base station group and one positioning base station of the next base station group form a composite three-prism group, with positioning base station 1-1 as the coordinate origin, positioning base station 1-1 and positioning base station 2-1 as the positive direction of the x-axis, positioning base station 1-1 and positioning base station 1-2 as the positive direction of the y-axis, and positioning base station 1-1 and positioning base station 1-3 as the positive direction of the z-axis. If necessary, the positioning base station position can be changed or the positioning base station group can be increased to meet the coordinate calculation principle.

[0053] S12, after deployment, record the unique tag, installation height, orientation and surrounding environment data of each base station one by one, and all enter the digital twin system to complete the initial state modeling.

[0054] S13, test the installed positioning system, use a mobile calibration terminal equipped with an inertial measurement unit (IMU), collect two-way ranging (TWR), channel impulse response (CIR) and time series data synchronized with the IMU along the preset trajectory. Based on the above data, run the base station joint calibration algorithm to estimate the relative coordinates between each base station and the clock bias between each other, and finally output the base station coordinate table with covariance and the geometric dilution of precision (GDOP) distribution map reflecting the positioning geometric precision. The specific process is as follows:

[0055] Let the target position of positioning in the three-dimensional space of the tunnel be , the position of the th base station be , and the ranging observation be: The principle diagram of the three-prism ranging method is shown in Figure 4 .

[0056] Linearize the above ranging observation equation at the approximate position to obtain the observation matrix of the positioning target in the tunnel:

[0057] ;

[0058] ;

[0059] According to the variance of single ranging error , calculate the covariance matrix of the position estimate, and the formula is as follows:

[0060] ;

[0061] Calculate GDOP, and the specific formula is as follows:

[0062] ;

[0063] In the formula, is the ranging noise, is the observed distance of the base station, is the geometric distance between the target position and the base station position, is the three-dimensional coordinate of the th base station, is the three-dimensional coordinate of the positioning target, is the total number of base stations, denotes distance, denotes transpose, Let be the trace of the matrix.

[0064] GDOP reflects the amplification effect of base station geometry on positioning accuracy. The smaller the value, the better the geometry distribution and the better the stability of positioning solution.

[0065] Step S2: In complex tunnel environments, the multipath effect of signals is significant. Therefore, a measurement quality factor Q is introduced to comprehensively score the reliability of each observation. The edge nodes convert the CIR (energy distribution of signal decay over time) into a two-dimensional time-delay-energy spectrum, which is then classified by a lightweight convolutional neural network to output the probability of the existence of a direct path. If the signal is determined to be LOS (Line-of-Sight, meaning there is a direct line of sight between the target and the base station, and the signal is not reflected or blocked), the SAGE algorithm is used to iteratively estimate the first path arrival delay. If the signal is determined to be NLOS (Non-Line-of-Sight, meaning the signal reaches the receiver after reflection or diffraction, and the path is longer than a straight line, leading to an overestimation of the distance), the extracted delay is corrected using a Bayesian shrinkage compensation method to reduce the error. The MUSIC algorithm is used to obtain the angle of arrival (AOA) and its confidence level. Finally, the measurement quality factor Q is calculated to reflect the reliability of each observation. The formula for calculating the measurement quality factor is as follows:

[0066] ;

[0067] in, For signal-to-noise ratio, The standard deviation of the angle of arrival. Due to historical bias, , , , These are weighting coefficients. It is the natural logarithm; This is the signal-to-noise ratio (SNR) smoothing coefficient, used to adjust the growth rate of the SNR scoring function to achieve a smooth transition in SNR scoring under complex multipath environments. Its value ranges from 8 to 12, with a preferred value of 10. At that time, the signal-to-noise ratio score reached 63% of its maximum value; This represents the reference variance value measured under ideal line-of-sight (LOS) conditions. , This represents the probability that a direct path exists.

[0068] In multiple sets of measured and simulated data from long tunnels (base station spacing 25–40 m, tunnel radius 5–6 m, humidity 80%), the optimal values ​​of the weighting coefficients were obtained by minimizing the overall positioning error, satisfying the following conditions: + + + =1, the parameter range and preferred values ​​are shown in Table 1:

[0069] Table 1 , , , The range of values

[0070]

[0071] This parameter set reduces positioning error by approximately 25% and GDOP fluctuation by approximately 18% in typical tunnel scenarios.

[0072] Step S3: Calculate the comprehensive cost function of the candidate triangular pyramid group, and select the candidate triangular pyramid group with the smallest cost function as the solution input to ensure the robustness and accuracy of the 3D solution in the complex environment of the tunnel. The solution flowchart is as follows: Figure 3 As shown, the formula for calculating the cost function is as follows:

[0073] ;

[0074] in, , These are weighting coefficients. For the candidate base station set, The number of base stations in the candidate base station set. Geometric precision factor, For the first Measurement quality factor of the base station group.

[0075] Verification through simulation and experimental results , The preferred values ​​are shown in Table 2:

[0076] Table 2 , Preferred value

[0077]

[0078] Step S4: Write the signal module worn by the worker or the airborne signal module into the digital twin system and distribute it, perform network-wide time synchronization, and then start the tunnel three-dimensional positioning system, such as... Figure 5 As shown, based on the covariance matrix in step S1 and the comprehensive cost function in step S3, four base stations from two groups of positioning base stations with relatively optimal signal quality (each group consists of three positioning base stations) are selected and combined with the signal module to form a composite triangular pyramid (base stations are selected based on signal quality). The three-dimensional coordinates of the signal module are then calculated. The formula for calculating the three-dimensional coordinates of the signal module is as follows:

[0079] ;

[0080] ;

[0081] ;

[0082] ;

[0083] wherein, is the output result of positioning, is the base station spatial relationship matrix, is the positioning observation vector, is the coordinate of the positioning base station 1-1, is the coordinate of the positioning base station 2-1, is the coordinate of the positioning base station 1-2, is the coordinate of the positioning base station 1-3, is the coordinate of the signal module, is the distance from the signal module to the positioning base station 1-1, 2-1, 1-2, 1-3, respectively.

[0084] Step S5, edge data processing and early warning, the specific steps are as follows:

[0085] S51, the edge node pre-processes the original data, including clock correction, CIR denoising, peak value and energy ratio extraction, LOS / NLOS rough judgment, preliminary TOA / AOA candidate identification, and generates a compressed observation package with quality annotation to report the central processing layer (data statistics and calculation center). In the case of communication anomaly or emergency, the edge node has local rapid judgment and alarm capability, and the observation package and quality annotation at this stage will be used as the input of the center fusion to ensure that the subsequent solution has reliable measurement support. The central processing layer and the digital twin system belong to the same level, and the central processing layer here is the data statistics and calculation center, and the edge processing layer is the data preprocessing center.

[0086] S52, based on the covariance matrix obtained in step S1 and the positioning terminal coordinates output by the center in step S4, the original positioning result is smoothed by using the covariance weighted Kalman filter model (the position covariance matrix Σp is introduced into the filter process as the observation noise matrix R). The smoothed trajectory is used to dynamically adjust the filter gain, combined with the geometry of the danger area, the motion constraint of the device and the relationship between multiple targets, to predict the behavior of collision or entering the danger area, and then perform hierarchical early warning (prompt, action restriction, emergency stop) according to the set threshold. The specific steps are as follows:

[0087] S52a, the smoothed three-dimensional coordinates of the signal module worn by the worker or the on-board signal module based on S4 the coordinate change to calculate the velocity vector And the short-time prediction coordinates of the target are calculated according to the velocity vector, and when the target has a significant acceleration change, an acceleration compensation term is automatically introduced, and the calculation formula is as follows:

[0088]

[0089] Wherein, is the prediction coordinate of the target, is the acceleration of the target, is the prediction step, preferably 1-3s.

[0090] S52b, based on the prediction point , the minimum distance between the target prediction point and the boundary of the dangerous area and other moving targets (mainly mechanical equipment) is calculated, if the minimum distance is less than the safety distance threshold (according to the specific setting of the construction site), it is determined as a potential risk event, and a warning is made, the calculation formula of the minimum distance is as follows:

[0091]

[0092]

[0093] Wherein, is the minimum distance from the target prediction point to the boundary of the dangerous area, is the minimum distance from the target prediction point to other targets. min represents the minimum value, is the dangerous area geometric boundary data set, is a data in the dangerous area geometric boundary data set, represents the distance, and are two different signal modules, in this embodiment, the signal module worn by the artificial and the signal module of the mechanical equipment, is the coordinate of the signal module worn by the artificial, is the coordinate of the signal module of the mechanical equipment.

[0094] Specifically, when the prediction coordinate enters the threshold boundary, that is, the minimum distance is less than or equal to the safety distance threshold, a first-level warning (prompt) is performed; when the actual coordinate enters the threshold boundary, a second-level warning (action restriction) is performed; when the prediction coordinate enters the dangerous area, a third-level warning (emergency stop) is performed.

[0095] The tunnel construction site equipped with the application is monitored for safety, when a high-risk event is predicted or the central and edge communication is limited, the edge node triggers an audible and light warning, issues a speed reduction or emergency stop command to the nearest mechanical equipment, records the response log, and returns the event and evidence to the center to form a closed-loop disposal record and post-analysis.

[0096] ​​​When a dangerous situation occurs, such as a man-machine collision, a target is lost, a tunnel collapses, etc., the system enters the rescue process, and a flowchart of this process is shown in Figure 6 The system automatically plans a safe and priority rescue path according to the current obstacles, traffic capacity and rescue resources, and sends the path and segmented navigation instructions to the management center, rescue personnel terminal and controllable rescue vehicles. At the same time, complete positioning trajectories, early warning events, disposal measures and response time delays are recorded for subsequent backtracking and legal and safety audits.

[0097] Experimental verification

[0098] The tunnel length is set to 1000m, the base station group is spaced 30m along the axis, and the target (worker and machine) motion mode is uniform walking + random acceleration event (average speed of personnel 1.2m / s, average speed of machine 20m / s). The simulation scene verification is performed.

[0099] The channel model is selected as a UWB multipath model (main path + several reflection paths), and the NLOS bias obeys a positive bias distribution (average bias 1.0-3.0m). The SNR distribution is 5-25dB (varies with position / obstruction)

[0100] The test sample is set to 2000 independent Monte Carlo experiments (different noise / obstruction distribution) per scene, and the test results are shown in Table 3:

[0101] Table 3: Monte Carlo experiment test results in the experimental scene

[0102]

[0103] Each embodiment in the specification is described in a related manner, and the same or similar parts between each embodiment can be referred to each other. Each embodiment focuses on the difference from other embodiments. In particular, for the system embodiment, since it is basically similar to the method embodiment, the description is relatively simple, and the related parts can be referred to the part of the method embodiment.

[0104] The above only describes the preferred embodiments of the present application, and is not used to limit the protection scope of the present application. Any modification, equivalent replacement, improvement, etc. made within the spirit and principles of the present application is included in the protection scope of the present application.

Claims

1. A method for measuring three-dimensional coordinates in a tunnel based on multiple sensors, characterized in that the steps include... include: Step S1: Collect tunnel data, complete the hardware deployment of the positioning system, and perform data collection and calibration to build the basic environment for the digital twin system; Step S2: Introduce the measurement quality factor Q to comprehensively score the reliability of each observation; Step S3: Calculate the comprehensive cost function of the candidate triangular pyramid group, and select the candidate triangular pyramid group with the smallest cost function as the solution input; Step S4: Write the signal module worn by the worker or the airborne signal module into the digital twin system and distribute it. Perform network-wide time synchronization and start the tunnel three-dimensional positioning system. Based on the covariance matrix in S1 and the comprehensive cost function in S3, select four base stations from two groups of positioning base stations and the signal module to form a composite triangular pyramid and calculate the three-dimensional coordinates of the signal module. Step S5: Edge data processing and early warning.

2. The method for measuring three-dimensional coordinates in a tunnel based on multiple sensors according to claim 1, characterized in that, The specific steps of S1 are as follows: S11, Deploy the positioning base station network based on tunnel data, and arrange positioning base station groups along the tunnel axis; S12, after deployment, record the unique tag, installation height, orientation and surrounding environment data of each base station one by one, and enter them all into the digital twin system to complete the initial state modeling; S13, test the installed positioning system, estimate the relative coordinates between base stations and the clock deviation between them, and output the base station coordinate table and geometric accuracy factor distribution map.

3. The method for measuring three-dimensional coordinates in a tunnel based on multiple sensors according to claim 2, characterized in that, In step S13, the geometric precision factor is calculated as follows: In the three-dimensional space of the tunnel, the target location is first determined as follows: and the The location of each base station is The ranging observation equations were listed and linearized at approximate locations to obtain the observation matrix for locating the target in the tunnel. ; Then, based on the variance of the single ranging error Calculate the covariance matrix of the location estimate. And the geometric precision factor GDOP, the formula is as follows: ； ； In the formula, Indicates transpose. Let be the trace of the matrix.

4. The method for measuring three-dimensional coordinates in a tunnel based on multiple sensors according to claim 1, characterized in that, In step S2, the edge nodes convert the CIR into a two-dimensional time-delay-energy spectrum, which is then classified by a lightweight convolutional neural network to output the probability of the existence of a direct path. If the signal is determined to be LOS, the SAGE algorithm is used to iteratively estimate the first path arrival delay; if it is determined to be NLOS, the extracted delay is corrected using the Bayesian shrinkage compensation method, and the signal angle of arrival and its confidence level are obtained. Finally, the measurement quality factor Q is calculated using the following formula: ； in, For signal-to-noise ratio, The standard deviation of the angle of arrival. Due to historical bias, , , , These are weighting coefficients. It is the natural logarithm. This is the signal-to-noise ratio smoothing coefficient. This represents the reference variance value measured under ideal direct viewing conditions. , This represents the probability that a direct path exists.

5. The method for measuring three-dimensional coordinates in a tunnel based on multiple sensors according to claim 1, characterized in that, In S3, the cost function is calculated using the following formula: ； in, , These are weighting coefficients. For the candidate base station set, The number of base stations in the candidate base station set. Geometric precision factor, For the first Measurement quality factor of the base station group.

6. The method for measuring three-dimensional coordinates in a tunnel based on multiple sensors according to claim 1, characterized in that, The specific steps of S5 are as follows: S51, edge nodes preprocess the raw data and generate compressed observation packages with quality labels, which are then reported to the central processing layer; S52, based on the covariance matrix obtained in step S1 and the positioning terminal coordinates output by the center in step S4, the original positioning result is smoothed, and the behavior of collision or entering the danger zone is predicted. Then, a graded warning is given according to the set threshold.

7. The method for measuring three-dimensional coordinates in a tunnel based on multiple sensors according to claim 6, characterized in that, The specific steps of S52 are as follows: S52a, Smooth 3D Coordinates Based on Signal Module The velocity vector is calculated from the coordinate changes. The system calculates the short-time predicted coordinates of the target based on the velocity vector. When the target exhibits significant acceleration changes, an acceleration compensation term is automatically introduced. The calculation formula is as follows: ； in, The predicted coordinates of the target. Accelerate towards the target To predict the step size; S52b, based on prediction points Calculate the minimum distance between the predicted target point and the boundary of the danger zone and other moving targets. If the minimum distance is less than the safe distance threshold, it is judged as a potential risk event and an early warning is issued. A Level 1 warning is issued when the predicted coordinates enter the threshold boundary; a Level 2 warning is issued when the actual coordinates enter the threshold boundary; and a Level 3 warning is issued when the predicted coordinates enter the danger zone.

Images