Researchers at Shanghai Jiao Tong University, led by Xi-Jun Yuan and Zi-Qiao Chen, have explored the application of quantum computing combined with machine learning to improve the accuracy of solving complex aircraft flow dynamics problems.
In a recent study published in the journal Intelligent Computing, the team focused on preventing aircraft stalls by analyzing the flow of air over airfoils, such as airplane wings. Traditionally, engineers have studied airfoil behavior to detect angles leading to flow separation. The researchers introduced quantum support vector machines, comparing them to classical support vector machines, and observed significant improvements in accuracy.
By utilizing a quantum support vector machine, the classification accuracy for detecting flow separation increased from 81.8% to 90.9%. Similarly, the accuracy of classifying the angle of attack rose from 67.0% to 79.0%. This suggests that quantum computing methods could offer faster and more precise solutions for fluid dynamics problems, particularly given the large datasets involved in such scenarios.
The study's applications extend beyond aircraft design, encompassing areas like underwater navigation and target tracking. The team conducted two classification tasks, one focusing on binary classification to detect flow separation and another classifying the angle of attack into four classes after flow separation.
The quantum-annealing-based supervised machine learning algorithm, known as a support vector machine, was employed in this research, utilizing the D-Wave Advantage 4.1 system, a physical quantum computing device. Quantum annealing, an optimization process leveraging quantum fluctuations, demonstrated superior performance compared to classical counterparts, offering more accurate results by avoiding local minima.
The potential implications of this research are significant, hinting at the broader integration of quantum computing in solving intricate problems related to fluid dynamics, with possible applications in various fields beyond aviation. The study exemplifies the promising intersection of quantum computing and machine learning in addressing complex real-world challenges.
