Wei Zhou
Current Organizations / Institutes
The organizations/institutes from the scholar's latest article
Research Subjects
Wei Zhou has a total of 14 co-authors. Here are the top co-authors by number of shared publications:
Anouk Stahnke
German Aerosp Ctr DLR, Inst Flight Guidance, Braunschweig, Germany
1 shared publicationDaniel Delahaye
1 shared publication- Imen Dhief
Air Traffic Management Research Institute
1 shared publication - Leo Julius Materne
Institute of Flight Guidance, German Aerospace Center (DLR), Lilienthalplatz 7, 38108 Braunschweig, Germany
1 shared publication - Lorenz Losensky
Institute of Flight Guidance, DLR German Aerospace Center, 38108 Braunschweig, Germany
1 shared publication Sameer Alam
1 shared publication
Latest Publications
- Open Access
- 7 September 2026
- Published by Elsevier in SSRN Electronic Journal
Distributionally robust multi-aircraft trajectory planning with finite convective weather and wind ensembles
Multi-aircraft trajectory planning under convective weather requires aircraft to avoid evolving storm regions while maintaining separation under wind uncertainty. Finite forecast ensembles provide a limited representation of future convective weather and wind conditions and may underrepresent adverse realizations, leading to overly optimistic empirical risk estimates. In this paper, we formulate a distributionally robust optimal control problem that accounts for uncertainty in both convective weather and wind distributions through Wasserstein ambiguity sets and conditional value-at-risk (CVaR) constraints. The resulting problem is infinite-dimensional, nonlinear and nonconvex. We derive a tractable approximation using an affine model to predict aircraft motion and supporting halfspaces to construct upper bounds on losses associated with convective weather intrusion and insufficient separation between aircraft. We then project uncertainty onto the corresponding supporting directions to reformulate the distributionally robust CVaR (DR-CVaR) constraints for both losses as finite-dimensional linear constraints. We solve the resulting convex quadratic program within a receding-horizon model predictive control (MPC) framework. Numerical experiments using a historical storm event and ensemble wind data evaluate performance under unseen realizations and increasing traffic complexity. In the tested scenarios, our proposed distributionally robust CVaR (DR-CVaR) method yields lower convective weather intrusion risk than sample average approximation (SAA) and robust optimization (RO) benchmarks, and lower loss of separation rates as aircraft interactions increase. It also achieves higher operational efficiency than RO. Sensitivity analysis shows that the ambiguity radius provides a practical way to adjust robustness.
- 7 September 2026
- Published by Elsevier in SSRN Electronic Journal
Open AccessDistributionally robust multi-aircraft trajectory planning with finite convective weather and wind ensembles
Multi-aircraft trajectory planning under convective weather requires aircraft to avoid evolving storm regions while maintaining separation under wind uncertainty. Finite forecast ensembles provide a limited representation of future convective weather and wind conditions and may underrepresent adverse realizations, leading to overly optimistic empirical risk estimates. In this paper, we formulate a distributionally robust optimal control problem that accounts for uncertainty in both convective weather and wind distributions through Wasserstein ambiguity sets and conditional value-at-risk (CVaR) constraints. The resulting problem is infinite-dimensional, nonlinear and nonconvex. We derive a tractable approximation using an affine model to predict aircraft motion and supporting halfspaces to construct upper bounds on losses associated with convective weather intrusion and insufficient separation between aircraft. We then project uncertainty onto the corresponding supporting directions to reformulate the distributionally robust CVaR (DR-CVaR) constraints for both losses as finite-dimensional linear constraints. We solve the resulting convex quadratic program within a receding-horizon model predictive control (MPC) framework. Numerical experiments using a historical storm event and ensemble wind data evaluate performance under unseen realizations and increasing traffic complexity. In the tested scenarios, our proposed distributionally robust CVaR (DR-CVaR) method yields lower convective weather intrusion risk than sample average approximation (SAA) and robust optimization (RO) benchmarks, and lower loss of separation rates as aircraft interactions increase. It also achieves higher operational efficiency than RO. Sensitivity analysis shows that the ambiguity radius provides a practical way to adjust robustness.
- 17 September 2025
- Published by Elsevier in Journal of Air Transport Management
- Vol. 131
Evaluating risk-based hazard corridors in air traffic controller decisions during space launch failures
- 17 September 2025
- Published by Elsevier in Journal of Air Transport Management
- Vol. 131
Evaluating risk-based hazard corridors in air traffic controller decisions during space launch failures
- 30 October 2024
- Published by Elsevier in Acta Astronautica
- Vol. 226, 728-759
A review of international best practices in integrating space launch activities with air traffic management: Guidance for future space launch in the Singapore region
- 30 October 2024
- Published by Elsevier in Acta Astronautica
- Vol. 226, 728-759
A review of international best practices in integrating space launch activities with air traffic management: Guidance for future space launch in the Singapore region