Quantum computing and machine learning enhance airplane performance and safety

In the quest to enhance airplane safety and performance with quantum computing, understanding the behavior of air over wing surfaces, known as airfoils, is crucial.

This is particularly true when it comes to preventing stalls, a scenario where an airplane loses the lift necessary to maintain flight.

Recently, a novel approach to this challenge emerged from the labs of Shanghai Jiao Tong University, where researchers Xi-Jun Yuan and Zi-Qiao Chen have turned to the burgeoning field of quantum computing.

Fusion of quantum computing and machine learning

Quantum computing, when fused with machine learning, has opened new vistas in various scientific domains.

In the realm of airplanes and aerodynamics, the quantum realm offers a more precise and efficient way to solve complex problems, especially those related to fluid dynamics.

The team at Shanghai Jiao Tong University has pioneered the use of a quantum support vector machine, a tool far superior to its classical counterpart.

In their research, the use of this quantum method resulted in a significant increase in predictive accuracy. The classification accuracy for detecting flow separation on airfoils jumped from 81.8% to 90.9%.

Similarly, the accuracy in determining the angle of attack — a critical parameter in aircraft performance — saw an improvement from 67.0% to 79.0%.

Quantum applications beyond airplane design

The implications of these findings are vast.

Quantum computing’s ability to handle large datasets with greater speed and accuracy makes it an ideal candidate for tackling complex fluid dynamics problems, such as those involved in airplane performance.

Beyond aircraft design, this technology has potential applications in fields like underwater navigation and target tracking.

Task one: Binary classification

Delving into the specifics, the researchers conducted two sets of classification tasks. The first involved a binary classification using a small dataset, aimed at detecting the presence or absence of flow separation.

This choice of a small dataset highlighted the challenge of achieving high accuracy in limited data scenarios.

The data were gathered using pressure sensors on an airfoil in a wind tunnel, under varying airspeeds and angles of attack.

This dataset, comprising 45 multidimensional points, was split into two parts for training and testing purposes.

Task two: Complex classification of angle of attack

The second task was more intricate, focusing on classifying the angle of attack post-flow separation into four distinct categories.

This involved breaking down the problem into four separate binary classification problems, each determining whether the angle of attack fell into a specific class.

Simulation-generated data formed the basis of this dataset, which consisted of 63 multidimensional points.

Quantum computing, airplanes, and fluid dynamics

The research team employed a quantum-annealing-based support vector machine for these tasks, using the D-Wave Advantage 4.1 system.

Quantum annealing, an optimization process that leverages quantum fluctuations, proved more effective than traditional optimization algorithms.

It offered a higher likelihood of finding the global minimum among potential solutions, thereby enhancing accuracy.

In summary, this study demonstrates the superior performance of quantum annealing implementations over classical methods, while also underscoring the potential of quantum computing in transforming how we solve complex real-world problems.

As quantum technology continues to evolve, its applications could redefine the frontiers of various scientific and technological fields, with aerodynamics being just the beginning.

The full study was published in the journal Intelligent Computing.

—–

Like what you read? Subscribe to our newsletter for engaging articles, exclusive content, and the latest updates.

Check us out on EarthSnap, a free app brought to you by Eric Ralls and Earth.com.

—–