Variational Quantum Algorithm Applied to Collision Avoidance of Unmanned Aerial Vehicles

Area of Science:

• Quantum Computing
• Aerospace Engineering
• Operations Research

Background:

• Unmanned aerial vehicle (UAV) mission planning requires robust conflict management for airspace safety.
• Noisy Intermediate-Scale Quantum (NISQ) devices offer potential for quantum advantage via Variational Quantum Algorithms (VQA).
• Existing methods struggle with the complexity of multi-UAV collision avoidance.

Purpose of the Study:

• To develop a quantum computing-based mathematical model for UAV collision avoidance.
• To map the collision avoidance problem to a Quadratic Unconstrained Binary Optimization (QUBO) problem.
• To leverage VQA for solving the formulated optimization problem.

Main Methods:

• Formulated UAV collision avoidance as an Ising Hamiltonian.
• Employed Variational Quantum Eigensolver (VQE) and Quantum Approximate Optimization Algorithm (QAOA) to find the ground state.
• Utilized Conditional Value-at-Risk (CVaR) to enhance model performance.

Main Results:

• The proposed model successfully mapped UAV collision avoidance to a QUBO problem.
• VQE and QAOA were effectively applied to solve the Ising Hamiltonian.
• Achieved a probability exceeding 90% for feasible solutions with optimized parameters.
• Demonstrated enhanced efficiency for UAV collision avoidance models.

Conclusions:

• Quantum computing, specifically VQA, presents a viable solution for complex multi-UAV collision avoidance.
• The QUBO formulation and VQA approach offer a promising path towards safe and efficient airspace management.
• The integration of CVaR further improves the reliability and performance of the proposed quantum algorithm.