Dynamically optimized departure and arrival funnels considering aircraft noise and weather conditions

• Applicant: Professor Dr.-Ing. Hartmut Fricke
• Subject Area: Traffic and Transport Systems, Intelligent and Automated Traffic
• Term: since 2025
• Project identifier: Deutsche Forschungsgemeinschaft (DFG) - Project number 553753829

For competitive/environmental reasons, the aviation industry must constantly reduce operating costs and minimize delays and emissions. A -multi-criteria - trajectory optimization therefore plays a key role in ATM research. Since the optimization goals compete, their mode of action must be well understood and costs quantified. For the cruise phase phase, this has already found its way into daily operations, e.g. with the introduction of free route airspace or weather-dependent, optimal routes in the North Atlantic. For arrivals and departures to/from major airports, however, the high volume of traffic and airspace restrictions often force aircraft to take inefficient altitudes and detours. Furthermore, aircraft noise at altitudes below 5,000 ft is a dominating social acceptance factor, the minimization of which is therefore also essential. In this application, so-called traffic flow funnels are developed. The funnels are designed as procedural spaces that extend current approach and departure routes by a 3D space. The aircraft flies its optimum altitude profile and flight path within the funnel. This allows the current weather and the interests of the airspace users to be taken into account to a much greater extent than before. However, the freedom of movement is also limited by the funnels to maintain predictability for air traffic control and thus safe and efficient operations. The funnels are generated with the TOolchain for Multi-criteria Aircraft Trajectory Optimization (TOMATO), an open access software designed by the applicant, which takes the various target functions into account, with the new focus now on aircraft noise. Current legally fixed methods for noise calculation do not yet allow the integration of trajectory optimization, mainly because the weather (wind, temperature, relative humidity) is not considered. To overcome this, different machine learning approaches are applied to a large set of noise measurement and weather data to predict the noise under given conditions. Using a cost-based noise function, many optimized trajectories are being generated using a Monte Carlo simulation, considering the uncertainty of the weather and operational constraints such as the aircraft mass. These trajectories are used to generate the funnels using a multi-stage clustering approach. As the number of operational funnels (DFS) that can be generated this way was classified as high in previous initial studies, various approaches with a small number for operational restriction using time windows or for reducing the spatial extent will be examined. The aim is to develop an operational concept based on funnels tending to overall optimality while ensuring adherence to existing air traffic control standards.

• DFG Programme: Research Grants