Please use this identifier to cite or link to this item:
Title: Risk-based route planning and decision making for unmanned aircraft system operations in urban environments
Authors: Pang, Bizhao
Keywords: Engineering::Aeronautical engineering::Aviation
Engineering::Industrial engineering::Operations research
Issue Date: 2023
Publisher: Nanyang Technological University
Source: Pang, B. (2023). Risk-based route planning and decision making for unmanned aircraft system operations in urban environments. Doctoral thesis, Nanyang Technological University, Singapore.
Abstract: Unmanned aircraft system (UAS) has been fast developing by leveraging advanced technologies such as power and flight control systems. The applications of UAS in urban environments also show a significant increase, with prominent use cases such as parcel delivery, security patrol, aerial photography, and data collection, which significantly facilitate our daily lives and public services. However, the rapid boom of UAS operations in and around urban airspace poses challenges for aviation regulatory bodies and brings various safety risks to the public. Specifically, drone operations may fail and crash into people, causing ground fatality risk. Airborne collisions between drones yield air risk, and the falling of collided drones causes knock-on ground risk. These safety risks could result in severe casualties to third-party people and considerable property loss to urban infrastructures, which need immediate research efforts to assess and mitigate the safety risks. This research problem is further complicated due to various uncertainties, such as weather and population movement in low-altitude urban environments. To solve these problems under uncertainties, mathematical models and algorithms are needed. However, these models and algorithms are lacking in the emerging UAS operational problems, which requires considerable efforts to bridge these research gaps. Therefore, this thesis focuses on developing a Third-party Risk Assessment and Mitigation System (TRAMS) by leveraging mathematical optimization and machine learning techniques. TRAMS considers both ground risk and air risk with the objective of minimizing safety risks and improving the efficiency and resilience of UAS operations. First, ground risk assessment and mitigation problems are solved by a novel UAV flight path optimization method that considers an integrated cost assessment model. The assessment model incorporates fatality risk, property damage risk, and noise impact, which is an extension of current third-party risk indicators at modeling and assessment levels. To solve the proposed optimization problem, a hybrid estimation of distribution algorithm (EDA) and CostA* (named as EDA-CostA*) algorithm is proposed, which provides global and local heuristic information for path searching in cost-based environments. Simulation results demonstrate the effectiveness and reliability of the developed optimization model in reducing risk costs. Second, aiming at handling uncertainties in ground risk studies due to people's movement, a two-stage stochastic route optimization model is proposed to jointly make flight approval and execution decisions. In the first stage, the approval decision is made by checking the feasibility of flight missions based on the target level of safety and flight duration to eliminate unsafe flight plans. The route selection and departure time decisions are made in the second stage by considering stochastic risk conditions to further reduce the ground risk. Numerical studies demonstrate that our proposed stochastic route optimization model performs the best in fatality risk reduction compared to distance-based and conventional risk-based methods. Third, air risk assessment and mitigation are conducted to prevent airborne flight conflicts. A decision-making framework is proposed to adaptively select the most suitable resolution strategy for different types of flight conflicts. The proposed framework is formulated as a double-layer optimization problem considering scheduling, speed adjustment, and rerouting strategies for conflict resolution. The framework's first layer is established to decide which strategy should be selected for what type of conflict. The second layer is developed as a mixed-integer nonlinear programming (MINLP) model to optimize the decision variables of the strategies selected by the first layer. The model is solved by a novel meta-heuristic stochastic fractal search (SFS) algorithm. Simulation results show that the proposed framework can resolve all conflicts with little cost of flight delays. Finally, to cope with uncertainties (e.g., flight delay and trajectory deviation) in conflict resolution decisions, a chance-constrained UAM traffic flow management (UTFM) optimization model is proposed to generate robust conflict-free trajectories and reduce flight delays. The model includes a probabilistic constraint of separation requirement to cope with trajectory time deviations caused by environmental disturbances, such as wind. To solve the model, we convert the probabilistic constraint into a deterministic constraint with a risk-bounded separation guarantee. Numerical studies demonstrate the proposed model's effectiveness, robustness, and scalability. The proposed chance-constrained UTFM model forms an additional layer of safety assurance in the pre-tactical phase, which enables fast flight disruption recovery in the UTFM system under delay and trajectory uncertainties.
DOI: 10.32657/10356/170275
Schools: School of Mechanical and Aerospace Engineering 
Research Centres: Air Traffic Management Research Institute 
Rights: This work is licensed under a Creative Commons Attribution-NonCommercial 4.0 International License (CC BY-NC 4.0).
Fulltext Permission: open
Fulltext Availability: With Fulltext
Appears in Collections:MAE Theses

Files in This Item:
File Description SizeFormat 
PhD Thesis_MAE_Bizhao PANG.pdf9.53 MBAdobe PDFThumbnail

Page view(s)

Updated on Mar 4, 2024

Download(s) 50

Updated on Mar 4, 2024

Google ScholarTM




Items in DR-NTU are protected by copyright, with all rights reserved, unless otherwise indicated.