A carregar...
Publicação
3D flight route optimization for air-taxis in urban areas with evolutionary algorithms
| Nome: | Descrição: | Tamanho: | Formato: | |
|---|---|---|---|---|
| 21.87 MB | Adobe PDF |
Autores
Orientador(es)
Resumo(s)
Electric aviation is being developed as a new mode of transportation for
the urban areas of the future. This requires an urban air space management that
considers these aircraft and restricts the vehicles’ flight routes from passing nofly
areas. Flight routes need to be determined that avoid the no-fly areas and are
also optimally planned in regard to minimize the flight time, energy consumption
and added noise. The no-fly areas and the flight routes can be best modelled as
three-dimensional geographical objects. The problem of finding a good flight route
that suits all three criteria is hard and requires an optimization technique. Yet, no
study exists for optimizing 3D-routes that are represented as geographical objects
while avoiding three-dimensional restricted areas. The research gap is overcome by
optimizing the 3D-routes with the multi-criteria optimization technique called Nondominated
Sorting Genetic Algorithm (II).We applied the optimization on the study
area of Manhattan (New York City) and for two representatives of different electrical
aircraft, the Lilium Jet and the Ehang 184. Special procedures are proposed in the
optimization process to incorporate the chosen geographical representations. We
included a seeding procedure for initializing the first flight routes, repair methods for
invalid flight routes and a mutation technique that relocates points along a sine curve.
The resulting flight routes are compromise solutions for the criteria flight time, energy
emission and added noise. Compared to a least distance path, the optimized flight
routeswere improved for all three objectives. The lowest observed improvementwas a
noise reduction by 36% for the Ehang 184. The highest improvement was an energy
consumption reduction by 90% for the Lilium Jet. The proposed representation
caused high computation times, which lead to other limitations, e.g. a missing
uncertainty analysis.With the proposed methods, we achieved to optimize 3D-routes
with multiple objectives and constraints. A reproducibility self-assessment1 resulted
in 2, 2, 2, 2, 1 (input data, preprocessing, methods, computational environment,
results).
Descrição
Dissertation submitted in partial fulfilment of the requirements for the degree of Master of Science in Geospatial Technologies
Palavras-chave
Electric aviation New mode Urban areas Three-dimensional geographical objects Megacities Geographical Information Systems