Conference Paper
Improving Air Mobility for Pre-Disaster Planning with Neural Network Accelerated Genetic Algorithm
An IEEE ITSC conference paper on using neural-network acceleration to make genetic-algorithm-based pre-disaster air mobility planning faster.
Abstract
Weather disaster related emergency operations pose a great challenge to air mobility in both aircraft and airport operations, especially when the impact is gradually approaching. We propose an optimized framework for adjusting airport operational schedules for such pre-disaster scenarios. We first, aggregate operational data from multiple airports and then determine the optimal count of evacuation flights to maximize the impacted airport’s outgoing capacity without impeding regular air traffic. We then propose a novel Neural Network (NN) accelerated Genetic Algorithm (GA) for evacuation planning. Our experiments show that integration yielded comparable results but with smaller computational overhead. We find that the utilization of a NN enhances the efficiency of a GA, facilitating more rapid convergence even when operating with a reduced population size. This effectiveness persists even when the model is trained on data from airports different from those under test.
In brief
What This Work Does
The paper uses neural networks to speed up a genetic algorithm for planning evacuation flights before weather disasters, helping airports respond faster under constrained operating conditions.
Research impact
Why It Matters
The work supports faster decision-making for air mobility operations when severe weather events are approaching.
Paper at a glance
Research Scope
Pre-Disaster Air Mobility Optimization Framework
Aggregate Operations
Uses multi-airport operational data to estimate available non-commercial capacity for evacuation-support flights.
Formulate Planning
Models schedule adjustment as an optimization problem under regular traffic and emergency constraints.
Search Schedules
Uses a genetic algorithm to explore feasible evacuation-flight schedule changes.
Accelerate Search
Uses neural network approximation to reduce computational overhead during optimization.
Key Contributions
- Proposes a pre-disaster scheduling framework that increases outbound airport capacity while maintaining normal airspace operations.
- Uses non-critical airport capacity to support evacuation flights before disaster impact.
- Introduces a neural-network-accelerated genetic algorithm for evacuation planning.
- Shows that NN assistance can reduce computational overhead and speed GA convergence.
- Evaluates generalization when the NN is trained on airports different from the test airports.
Method Components
Genetic Algorithm
Searches over possible schedule adjustments for pre-disaster air mobility planning.
Neural Network Acceleration
Speeds GA convergence even with reduced population size and limited NN training.
Airport Capacity Modeling
Estimates outgoing evacuation capacity while preserving regular operations as much as possible.
Cross-Airport Generalization
Tests whether an NN trained on one airport can still support GA search for other airports.
Research Gaps
- Real-time optimization
- Emergency operations
- Capacity uncertainty
- Multi-airport coordination
- Weather disruption
- Cross-airport transfer
Publication Details
- Type
- Conference Paper
- Venue
- IEEE ITSC 2024
- Year
- 2024
- Pages
- 1290-1295
Research Topics
Links and Access
Citation
@inproceedings{acharya2024improving,
author={Acharya, Kamal and Velasquez, Alvaro and Liu, Y. and Liu, D. and Sun, Liang and Song, Houbing},
title={Improving Air Mobility for Pre-Disaster Planning with Neural Network Accelerated Genetic Algorithm},
booktitle={IEEE ITSC 2024},
year={2024},
pages={1290--1295},
doi={10.1109/ITSC58415.2024.10920105}
}