In addition, some studies established general nonlinear models or used a two-step optimisation approach. Among mathematical programming methods, most studies have employed mixed-integer nonlinear programming (MINLP) or mixed-integer linear programming (MILP) to establish mathematical models. The CR schemes obtained using optimal control methods are generally a trajectory composed of continuous position points of the aircraft, which is quite different from the scheme of ATCOs. further considered constructive weather areas and proposed a stochastic optimal control algorithm for collision avoidance. considered wind certainty and solved the CR problem using the stochastic near-optimal control method. considered fuel-optimal conflict-free trajectory planning as a hybrid optimal control problem. The main approaches to solving CR problems include optimal control methods, mathematical programming methods, swarm intelligence optimisation or search methods, and machine learning methods. Reviews of CR problems can be found in several reports. Therefore, this work uses the deep reinforcement learning (DRL) method to implement a core function in decision support systems, i.e., conflict resolution, to improve the CR model’s intelligence level.Īcademic efforts are underway to address CR problems. The higher level of ATC automation requires automation to adapt to a broad range of scenarios and tasks while giving the human operator appropriate action advice. Existing decision support systems need to be improved in terms of handling specific tasks and scenarios. Assisting ATCOs is concerned with acceptance of schemes by ATCOs and air traffic control (ATC) regulators. Different assisted decision-making targets have different requirements for mechanisms and algorithms. Another category is airborne systems (primarily designed to aid in aircraft decision making), such as the Airborne Separation Assurance System (ASAS), the Traffic Collision Avoidance System (TCAS), the Airborne Collision Avoidance System (ACAS), and ACAS Xu. One category is ground-based systems (primarily designed to assist ATCOs in their decision making), such as the Centre/TRACON Automation System (CTAS), the User Request Evaluation Tool (URET), and En Route Automation Modernization (ERAM). This research can be applied to intelligent decision-making systems for air traffic control.ĭecision support systems with conflict detection and resolution functions can be broadly classified into two categories. The conflict resolution rate decreased slightly to 81.2% when the airspace density was increased by a factor of 1.4. Results show that for 1000 test samples, the trained TCS could resolve 87.1% of the samples. A DRL environment was developed with the actual airspace structure and traffic density of the air traffic operation simulation system. Considering the uncertainty in a real-life situation, this study characterised the deviation of the aircraft’s estimated position to improve the feasibility of conflict resolution schemes. The reward function is designed in accordance with air traffic control regulations. The agent’s actions are determined by the ATCOs’ instructions, such as altitude, speed, and heading adjustments. The tactical conflict solver (TCS) was developed based on deep reinforcement learning (DRL) to train a TCS agent with the actor–critic using a Kronecker-factored trust region. To assist air traffic controllers (ATCOs) in resolving tactical conflicts, this paper proposes a conflict detection and resolution mechanism for handling continuous traffic flow by adopting finite discrete actions to resolve conflicts.
0 Comments
Leave a Reply. |
AuthorWrite something about yourself. No need to be fancy, just an overview. ArchivesCategories |