In the context of urban air traffic, there is currently no mature solution for real-time avoidance of electric Vertical Take-Off and Landing (eVTOL) aircraft in dynamic and uncertain environments. To address this challenge, we propose a risk-aware and efficient motion planning method for real-time eVTOL avoidance. The problem is formulated as a Model Predictive Control (MPC) framework with Chance Constraints MPC (CC-MPC), incorporating both collision avoidance and geo-fencing constraints. To efficiently handle the chance constraints, we reformulate them using Big-M and confidence ellipsoids, transforming the CC-MPC problem into a Mixed-Integer Programming (MIP) problem. To efficiently solve the MIP, we employ an iterative convexification-based optimization method, complemented by a global search algorithm that serves as a front-end warm-start mechanism. Finally, all components are integrated within a receding horizon control framework to enable fast and dynamic trajectory generation for eVTOLs. Simulation experiments across various flight scenarios demonstrate the effectiveness of the proposed approach.

Radar direction finding is an essential cooperative monitoring method for aircraft, and the directional parameters of the direction-finding antenna determine the accuracy of direction finding. A modified amplitude comparison direction finding model is proposed based on the traditional principle of adjacent amplitude comparison direction finding and the error of amplitude comparison direction finding. Based on the modified amplitude comparison direction finding model, an error correction genetic simulated annealing (COR-GSA) algorithm is proposed to identify unknown directional antenna directionality parameters by determining the antenna directionality parameters that need to be identified. Identification experiments were conducted using a self-designed dual channel DF receiver, and 1000 sets of aircraft real azimuth data were used for verification. The results showed that using the COR-GSA algorithm to identify antenna directional parameters had the highest DF accuracy. Finally, the identified antenna directionality parameters were used to track the aircraft’s azimuth, and the tracking error was reduced by 18.3% compared to the tracking error without azimuth correction.

In order to address the issue of the traditional variable structure multiple model (VSMM) algorithm's lengthy computation time and difficulty meeting real-time constraints, we developed the intent variable structure multiple model (INT-VSMM) algorithm. This algorithm uses the target state and flight intent data collected by automatic dependent surveillance-broadcast (ADS-B) as the model’s priori data, combining it with the VSMM theoretical framework. The motion pattern of the target in the flight phase of the flight path is decomposed, and a complete set of motion models is established. According to the principle of directed graph switching, a model set switching method is designed, which is mainly based on "hard" switching and supplemented by "soft" switching. The INT-VSMM algorithm is used to track the trajectory of the target aircraft, and short-term trajectory extrapolation is performed based on the target state estimation. According to the simulation results, the INT-VSMM algorithm outperforms the compared current approaches in terms of target tracking performance and computing time. Additionally, the extrapolated trajectory has a lower short-term inaccuracy, which can satisfy conflict detection requirements.

In view of the fact that the high altitude wind, a random factor, often leads to poor robustness of the longitudinal separation between two aircraft in the longitudinal autonomous separation maintenance problem under the random disturbance of track, this paper proposed a longitudinal autonomous separation control method based on model predictive control (MPC). Firstly, the linear time-varying prediction model was developed by establishing the nonlinear kinematics differential equation of the longitudinal separation and the wind field difference between the two aircraft. The longitudinal separation and route deviation distance of the two aircraft was selected as the optimization objectives, the vacuum speed and yaw angle of the front aircraft were taken as the measurable disturbances, and the high-altitude wind was the random disturbance. Terminal equality constraints were added to the air safety and aircraft performance constraints to maintain the stability of the system. To verify the effectiveness of the proposed method, within the specified 120-second simulation time, this article set three sets of different expected separations of 12 km, 13 km, and 14 km. Through the design of an MPC controller, the vacuum speed and yaw angle of the following aircraft were controlled during the rolling time domain cycle. The separation curve between the two aircraft is relatively smooth and always not less than the minimum safety separation of 10 km. It stabilized at the expected target separation in the 74th second, 90th second, and 118th second, and returned to the route starting from the 58th second, 74th second, and 95th second. Two sets of wind field control groups were set up. Two times as much wind was forecast in one group, while eight times as much turbulent wind was disturbed in the other. Both groups were able to establish the expected interval of 12 km smoothly and stably in the 61th second and 72th second, respectively.

The prediction of aircraft taxiing trajectory helps to solve operational problems such as taxiing conflicts and long waiting times at airports, ensuring surface safety while improving service levels and increasing airport throughput. A model is proposed to predict the taxiing trajectory of a surface aircraft based on an attention mechanism that combines gated recurrent units (GRU) and an improved Kalman filter algorithm (IKF). This addresses the current situation where the performance of machine learning models depends on good data sets. In order to better extract data discrepancy features and learn input-to-output mapping relationships, three independent networks of gated recurrent units are first used to capture the future moment motion states and temporal dependencies of the aircraft. An enhanced extended Kalman filter is then fused with the neural network outputs to integrate them into the state prediction and update process, ultimately improving the predicted trajectory sequence accuracy. Finally, the validity of the model was verified using real aircraft taxi trajectories at Lukou Airport. The simulation results show that the proposed model can effectively and accurately predict aircraft taxi trajectories at the field with an overall mean square error of approximately 0.00128. Compared with the single recurrent neural network (RNN), long and short-term memory network (LSTM) and GRU model, the root mean square error (RMSE) is reduced by 72.9%, 54.7% and 39.9% respectively, and the prediction time is 40 ms, which could accurately and quickly predict the taxiing trajectory and provide assistance in reducing the operating load of the airport surface management system.