Airport runway intrusion risk assessment method and system based on artificial intelligence
By using AI-based multimodal data fusion and risk assessment, the existing technical problems of airport runway intrusion detection have been solved, enabling efficient and comprehensive safety management around the clock, reducing false alarm and missed alarm rates, and improving airport security.
Patent Information
- Application Number
- CN202511679697.4
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2025-11-17
- Publication Date
- 2026-02-03
AI Technical Summary
In existing technologies, airport runway intrusion detection relies on manual patrols and traditional systems, which suffer from problems such as significant weather influence, human error, insufficient blind spot coverage, limited positioning accuracy, and high costs. It cannot achieve all-weather, all-round monitoring, and traditional electromagnetic wave detection is easily interfered with, resulting in a high rate of false alarms and missed alarms.
Using an artificial intelligence-based approach, multimodal data (radar trajectory, ADS-B signal, runway monitoring video, meteorological data, and voice commands) are collected, spatiotemporal alignment and feature extraction and fusion are performed to establish a trajectory risk assessment model, conduct risk level assessment, and issue early warnings and adjustment commands based on the assessment results.
It enables proactive risk identification and early accurate identification of runway incursions, reduces false alarm and missed alarm rates, improves the safety management level of airport runways, and ensures rapid intervention in high-risk situations and reasonable intervention in low-risk situations.
Smart Images
Figure CN121459482A_ABST
Abstract
Description
Technical Field
[0001] This application belongs to the field of airport risk management technology, specifically, it relates to an artificial intelligence-based method and system for assessing airport runway intrusion risks. Background Technology
[0002] With the rapid development of the civil aviation industry, airport safety management has become increasingly important. Runway incursion, namely the unauthorized entry of aircraft, vehicles, or personnel into the runway area, has always been one of the significant safety hazards in airport operations. Runway incursions can lead to serious flight accidents and severely impact airport operations and flight safety.
[0003] Currently, the monitoring and prevention of runway incursions mainly rely on manual patrols and traditional warning systems. These methods have many limitations. Traditional runway incursion monitoring methods mainly rely on security personnel patrols and observations, which are greatly affected by weather and lighting conditions. Controllers are under heavy workloads and prone to human error. They also suffer from high labor costs and insufficient blind spot coverage, making it impossible to achieve all-weather, all-round monitoring. At the same time, traditional electromagnetic wave detection cannot detect aircraft and vehicles deviating from or incorrectly executing instructions. They are susceptible to clutter interference, have blind spots, limited positioning accuracy, and are prone to false alarms or missed alarms. Furthermore, they have high construction and maintenance costs. Summary of the Invention
[0004] To address the aforementioned problems and technical deficiencies, this application adopts the following technical solution: an artificial intelligence-based method for assessing airport runway incursion risks, comprising the following steps: The system acquires radar trajectories, ADS-B signals, runway monitoring videos, meteorological data, and voice commands. It performs spatiotemporal alignment and feature extraction on the acquired multimodal data and performs feature fusion on the extracted features. A trajectory risk assessment model is established, which first predicts the trajectory from the fused feature data, and then assesses the risk level of the predicted trajectory. Based on the risk level assessment results, early warning prompts, audible and visual alarms, information pushes, or mandatory adjustment instructions will be issued.
[0005] Preferably, the radar trajectory features and ADS-B signal features are obtained by using wavelet transform to extract the instantaneous velocity, acceleration, and rate of change of heading angle of the trajectory, and by calculating the target turning features through the trajectory curvature; Video features are extracted by taking the bounding box coordinates, area, and motion vector of the target, and the target's moving speed and direction are calculated by the inter-frame difference method. Meteorological features are key influencing factors extracted from meteorological data using gradient boosting trees, including quantitative features of visibility level, wind speed threshold, and precipitation type. The feature extraction of voice commands involves establishing a control command word model, converting the controller's voice commands into structured intent information in real time, extracting the actual trajectory of the target corresponding to the voice command from radar trajectory, ADS-B signal and surveillance video, calculating the deviation between the actual trajectory and the structured intent information, and extracting command features based on the deviation.
[0006] Furthermore, the feature fusion of the extracted features includes: Preprocessing of multimodal data includes format normalization, noise filtering, and missing value repair; A spatiotemporal alignment algorithm is used to synchronously match data from different time and spatial dimensions; Key features of each data point are extracted using a feature extraction algorithm, and feature fusion is performed after unifying the feature dimensions. A weighted voting mechanism is used to conduct a preliminary evaluation of the fused feature data, and the fused feature data is then corrected in real time based on the preliminary evaluation.
[0007] Preferably, the trajectory prediction employs a dynamic Bayesian network and a fusion reinforcement learning model to analyze historical data features and real-time fused feature data. The analysis results are input into a 3D model of the airport runway to simulate and predict the airport runway operation scenario, generating the future motion trajectory of each target.
[0008] Furthermore, the risk assessment includes: Calculate the environmental impact value based on the season and weather data from the fused feature data; Identifying key risk nodes based on predicted trajectories; Calculate the probability and severity of conflict between aircraft, vehicles, personnel, and foreign objects at high altitudes based on key risk points; A dynamic conflict risk value is generated by combining environmental impact value, conflict probability, severity and dynamic correction factor; A risk level assessment matrix is constructed, and the dynamic conflict risk value is compared with the risk level assessment matrix to obtain the risk level assessment result.
[0009] Furthermore, the dynamic correction factor is obtained by establishing a risk attenuation compensation model, acquiring the quantitative trend of performance attenuation of multimodal data acquisition equipment, and incorporating it as a dynamic correction factor into the risk assessment trajectory model for calculation.
[0010] Furthermore, when the risk level is low or relatively low, an early warning is output to the airport operation management system; when the risk level is medium, an audible and visual alarm is output to the controller terminal, and the target's current location and predicted trajectory are simultaneously pushed; when the risk level is high or high, a forced adjustment command is output to the controller terminal and the target vehicle.
[0011] An artificial intelligence-based airport runway incursion risk assessment system includes: The information acquisition module is used to collect radar trajectories, ADS-B signals, runway monitoring videos, meteorological data, and voice commands. It performs spatiotemporal alignment and feature extraction on the collected multimodal data and performs feature fusion on the extracted features. The risk assessment module is used to establish a trajectory risk assessment model. First, the trajectory is predicted based on the fused feature data, and then the risk level of the predicted trajectory is assessed. The warning and risk avoidance module is used to provide early warning prompts, audible and visual alarms, information push notifications, or forced adjustment command outputs based on the risk level assessment results.
[0012] An electronic device includes a memory, a processor, and a computer program stored in the memory and executable on the processor. When the processor executes the program, it implements the content of the artificial intelligence-based airport runway intrusion risk assessment method described above.
[0013] A computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, implements the content of the artificial intelligence-based airport runway intrusion risk assessment method described above.
[0014] Compared to existing technologies, the beneficial effects of this application are as follows: (1) This application transforms controller voice commands into quantifiable structured intentions and performs real-time deviation analysis with the actual trajectory of the target, thereby proactively identifying potential conflicts caused by communication misunderstandings or execution errors, realizing the shift from passively perceiving the situation to proactively understanding and verifying the intentions, and improving the earlyness and accuracy of risk identification. (2) This application constructs an operational status profile by deeply fusing multimodal data such as radar, ADS-B, video, meteorology and voice commands, and adopts technologies such as spatiotemporal alignment, weighted voting and real-time correction. It uses a hybrid model combining dynamic Bayesian network and reinforcement learning to predict trajectories in a three-dimensional scene, and calculates the environmental impact and equipment attenuation factor to reduce false alarm and false alarm rates. (3) This application constructs a risk level assessment matrix and triggers progressive response measures from early warning prompts, audible and visual alarms to mandatory adjustment instructions, thereby ensuring rapid mandatory intervention in high-risk situations and avoiding unnecessary interference in low-risk situations, enhancing the practicality of the system and the efficiency of human-machine collaboration, and improving the level of proactive safety management of airport runways. Attached Figure Description
[0015] In the attached diagram: Figure 1 This is a schematic diagram of the method steps in an embodiment of this application; Figure 2This is a schematic diagram of the system structure according to an embodiment of this application. Detailed Implementation
[0016] To make the objectives, technical solutions, and advantages of the embodiments of this application clearer, the technical solutions of the embodiments of this application will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are some embodiments of this application, but not all embodiments. Generally, the components of the embodiments of this application described and shown in the accompanying drawings can be arranged and designed in various different configurations. Example
[0017] like Figure 1 As shown, an artificial intelligence-based method for assessing airport runway incursion risk includes the following steps: The system acquires radar trajectories, ADS-B signals, runway monitoring videos, meteorological data, and voice commands. It performs spatiotemporal alignment and feature extraction on the acquired multimodal data and performs feature fusion on the extracted features. Radar trajectory features and ADS-B signal features are obtained by using wavelet transform to extract the instantaneous velocity, acceleration, and rate of change of heading angle of the trajectory, and by calculating the target turning features through the trajectory curvature. Video features are extracted by taking the bounding box coordinates, area, and motion vector of the target, and the target's moving speed and direction are calculated by the inter-frame difference method. Meteorological features are key influencing factors extracted from meteorological data using gradient boosting trees, including quantitative features of visibility level, wind speed threshold, and precipitation type. The feature extraction of voice commands involves establishing a control command word model, converting the controller's voice commands into structured intent information in real time, extracting the actual trajectory of the target corresponding to the voice command from radar trajectory, ADS-B signal and surveillance video, calculating the deviation between the actual trajectory and the structured intent information, and extracting command features based on the deviation.
[0018] Feature fusion of extracted features includes: Preprocessing of multimodal data includes format normalization, noise filtering, and missing value repair; A spatiotemporal alignment algorithm is used to synchronously match data from different time and spatial dimensions; Key features of each data point are extracted using a feature extraction algorithm, and feature fusion is performed after unifying the feature dimensions. A weighted voting mechanism is used to conduct a preliminary evaluation of the fused feature data, and the fused feature data is then corrected in real time based on the preliminary evaluation.
[0019] A trajectory risk assessment model is established, which first predicts the trajectory from the fused feature data, and then assesses the risk level of the predicted trajectory. Trajectory prediction employs a dynamic Bayesian network and a fusion reinforcement learning model to analyze historical data features and real-time fusion feature data. The analysis results are input into a 3D model of the airport runway to simulate and predict the airport runway operation scenario and generate the future motion trajectory of each target.
[0020] Risk assessment includes: Calculate the environmental impact value based on the season and weather data from the fused feature data; Identifying key risk nodes based on predicted trajectories; Calculate the probability and severity of conflict between aircraft, vehicles, personnel, and foreign objects at high altitudes based on key risk points; A dynamic conflict risk value is generated by combining environmental impact value, conflict probability, severity and dynamic correction factor; A risk level assessment matrix is constructed, and the dynamic conflict risk value is compared with the risk level assessment matrix to obtain the risk level assessment result.
[0021] The dynamic correction factor is obtained by establishing a risk attenuation compensation model, acquiring the quantitative trend of performance attenuation of multimodal data acquisition equipment, and then using it as a dynamic correction factor to calculate the risk assessment trajectory model.
[0022] Based on the risk level assessment results, early warning prompts, audible and visual alarms, information pushes, or mandatory adjustment instructions will be issued.
[0023] When the risk level is low or relatively low, an early warning is issued to the airport operations management system; when the risk level is medium, an audible and visual alarm is issued to the controller terminal, and the target's current location and predicted trajectory are simultaneously pushed; when the risk level is high or high, a forced adjustment instruction is issued to the controller terminal and the target vehicle. Example
[0024] like Figure 2 As shown, an artificial intelligence-based airport runway intrusion risk assessment system includes: The information acquisition module is used to collect radar trajectories, ADS-B signals, runway monitoring videos, meteorological data, and voice commands. It performs spatiotemporal alignment and feature extraction on the collected multimodal data and performs feature fusion on the extracted features. Radar trajectory features and ADS-B signal features are obtained by using wavelet transform to extract the instantaneous velocity, acceleration, and rate of change of heading angle of the trajectory, and by calculating the target turning features through the trajectory curvature. Video features are extracted by taking the bounding box coordinates, area, and motion vector of the target, and the target's moving speed and direction are calculated by the inter-frame difference method. Meteorological features are key influencing factors extracted from meteorological data using gradient boosting trees, including quantitative features of visibility level, wind speed threshold, and precipitation type. The feature extraction of voice commands involves establishing a control command word model, converting the controller's voice commands into structured intent information in real time, extracting the actual trajectory of the target corresponding to the voice command from radar trajectory, ADS-B signal and surveillance video, calculating the deviation between the actual trajectory and the structured intent information, and extracting command features based on the deviation.
[0025] Feature fusion of extracted features includes: Preprocessing of multimodal data includes format normalization, noise filtering, and missing value repair; A spatiotemporal alignment algorithm is used to synchronously match data from different time and spatial dimensions; Key features of each data point are extracted using a feature extraction algorithm, and feature fusion is performed after unifying the feature dimensions. A weighted voting mechanism is used to conduct a preliminary evaluation of the fused feature data, and the fused feature data is then corrected in real time based on the preliminary evaluation.
[0026] The risk assessment module is used to establish a trajectory risk assessment model. First, the trajectory is predicted based on the fused feature data, and then the risk level of the predicted trajectory is assessed. Trajectory prediction employs a dynamic Bayesian network and a fusion reinforcement learning model to analyze historical data features and real-time fusion feature data. The analysis results are input into a 3D model of the airport runway to simulate and predict the airport runway operation scenario and generate the future motion trajectory of each target.
[0027] Risk assessment includes: Calculate the environmental impact value based on the season and weather data from the fused feature data; Identifying key risk nodes based on predicted trajectories; Calculate the probability and severity of conflict between aircraft, vehicles, personnel, and foreign objects at high altitudes based on key risk points; A dynamic conflict risk value is generated by combining environmental impact value, conflict probability, severity and dynamic correction factor; A risk level assessment matrix is constructed, and the dynamic conflict risk value is compared with the risk level assessment matrix to obtain the risk level assessment result.
[0028] The dynamic correction factor is obtained by establishing a risk attenuation compensation model, acquiring the quantitative trend of performance attenuation of multimodal data acquisition equipment, and then using it as a dynamic correction factor to calculate the risk assessment trajectory model.
[0029] The warning and risk avoidance module is used to provide early warning prompts, audible and visual alarms, information push notifications, or forced adjustment command outputs based on the risk level assessment results.
[0030] When the risk level is low or relatively low, an early warning is issued to the airport operations management system; when the risk level is medium, an audible and visual alarm is issued to the controller terminal, and the target's current location and predicted trajectory are simultaneously pushed; when the risk level is high or high, a forced adjustment instruction is issued to the controller terminal and the target vehicle. Example
[0031] From a hardware perspective, this application provides an embodiment of an electronic device containing all or part of an artificial intelligence-based airport runway intrusion risk assessment method. The electronic device includes a service processor and a distributed memory. The service processor is connected to the memory, and the distributed memory stores a service self-management program configured to store machine-readable instructions. The service processor executes the service self-management program, and the instructions, when executed by the processor, implement the artificial intelligence-based airport runway intrusion risk assessment method as described above. Example
[0032] The embodiments of this application also provide a computer-readable storage medium capable of implementing the AI-based airport runway intrusion risk assessment method with the execution subject as a server or client in the above embodiments. The computer-readable storage medium stores a computer program that, when executed by a processor, implements all the contents of the AI-based airport runway intrusion risk assessment method with the execution subject as a server or client in the above embodiments.
[0033] The embodiments described above are merely preferred embodiments of this application, and while the descriptions are specific and detailed, they should not be construed as limiting the scope of this application. It should be noted that those skilled in the art can make various modifications, improvements, and substitutions without departing from the concept of this application, and these all fall within the protection scope of this application.
Claims
1. An artificial intelligence-based method for assessing airport runway incursion risk, characterized in that, Includes the following steps: The system acquires radar trajectories, ADS-B signals, runway monitoring videos, meteorological data, and voice commands. It performs spatiotemporal alignment and feature extraction on the acquired multimodal data and performs feature fusion on the extracted features. A trajectory risk assessment model is established, which first predicts the trajectory from the fused feature data, and then assesses the risk level of the predicted trajectory. Based on the risk level assessment results, early warning prompts, audible and visual alarms, information pushes, or mandatory adjustment instructions will be issued.
2. The method for assessing airport runway incursion risk based on artificial intelligence according to claim 1, characterized in that, The radar trajectory features and ADS-B signal features are obtained by using wavelet transform to extract the instantaneous velocity, acceleration, and rate of change of heading angle of the trajectory, and by calculating the target turning features through the trajectory curvature. Video features are extracted by taking the bounding box coordinates, area, and motion vector of the target, and the target's moving speed and direction are calculated by the inter-frame difference method. Meteorological features are key influencing factors extracted from meteorological data using gradient boosting trees, including quantitative features of visibility level, wind speed threshold, and precipitation type. The feature extraction of voice commands involves establishing a control command word model, converting the controller's voice commands into structured intent information in real time, extracting the actual trajectory of the target corresponding to the voice command from radar trajectory, ADS-B signal and surveillance video, calculating the deviation between the actual trajectory and the structured intent information, and extracting command features based on the deviation.
3. The method for assessing airport runway incursion risk based on artificial intelligence according to claim 2, characterized in that, The feature fusion of the extracted features includes: Preprocessing of multimodal data includes format normalization, noise filtering, and missing value repair; A spatiotemporal alignment algorithm is used to synchronously match data from different time and spatial dimensions; Key features of each data point are extracted using a feature extraction algorithm, and feature fusion is performed after unifying the feature dimensions. A weighted voting mechanism is used to conduct a preliminary evaluation of the fused feature data, and the fused feature data is then corrected in real time based on the preliminary evaluation.
4. The method for assessing airport runway incursion risk based on artificial intelligence according to claim 1, characterized in that, The trajectory prediction employs a dynamic Bayesian network and a fusion reinforcement learning model to analyze historical data features and real-time fusion feature data. The analysis results are input into a 3D model of the airport runway to simulate and predict the airport runway operation scenario, generating the future motion trajectory of each target.
5. The method for assessing airport runway incursion risk based on artificial intelligence according to claim 4, characterized in that, The risk assessment includes: Calculate the environmental impact value based on the season and weather data from the fused feature data; Identifying key risk nodes based on predicted trajectories; Calculate the probability and severity of conflict between aircraft, vehicles, personnel, and foreign objects at high altitudes based on key risk points; A dynamic conflict risk value is generated by combining environmental impact value, conflict probability, severity and dynamic correction factor; A risk level assessment matrix is constructed, and the dynamic conflict risk value is compared with the risk level assessment matrix to obtain the risk level assessment result.
6. The method for assessing airport runway incursion risk based on artificial intelligence according to claim 5, characterized in that, The dynamic correction factor is obtained by establishing a risk attenuation compensation model, acquiring the quantitative trend of performance attenuation of multimodal data acquisition equipment, and then using it as a dynamic correction factor to calculate the risk assessment trajectory model.
7. The method for assessing airport runway incursion risk based on artificial intelligence according to claim 5, characterized in that, When the risk level is low or relatively low, an early warning message is output to the airport operation management system. When the risk level is medium, an audible and visual alarm is output to the controller terminal, and the target's current location and predicted trajectory are pushed simultaneously; when the risk level is high or medium, a forced adjustment command is output to the controller terminal and the target vehicle.
8. An artificial intelligence-based airport runway incursion risk assessment system, characterized in that, include: The information acquisition module is used to collect radar trajectories, ADS-B signals, runway monitoring videos, meteorological data, and voice commands. It performs spatiotemporal alignment and feature extraction on the collected multimodal data and performs feature fusion on the extracted features. The risk assessment module is used to establish a trajectory risk assessment model. First, the trajectory is predicted based on the fused feature data, and then the risk level of the predicted trajectory is assessed. The warning and risk avoidance module is used to provide early warning prompts, audible and visual alarms, information push notifications, or forced adjustment command outputs based on the risk level assessment results.
9. An electronic device comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, characterized in that, When the processor executes the program, it implements the content of the artificial intelligence-based airport runway intrusion risk assessment method as described in claim 1.
10. A computer-readable storage medium having a computer program stored thereon, characterized in that, When the computer program is executed by the processor, it implements the content of the artificial intelligence-based airport runway intrusion risk assessment method as described in claim 1.