Improving Maglev Performance with Machine Learning

For centuries, the transportation sector has been one of the largest contributors to global carbon emissions. To address this issue, many research groups are now actively investigating how technology can be used to make transportation systems more sustainable and environmentally friendly.

One of the most promising technologies currently being researched is magnetic levitation (maglev), which uses powerful magnets to push trains along the rails. By virtually eliminating the friction caused by tracks, maglev trains can quickly accelerate to incredibly high speeds with very little energy.

But while magnetic levitation systems have long been widely studied, it remains difficult for researchers to precisely measure the forces imparted by magnets. This lack of precision is one of the reasons why magnetic levitation has yet to replace traditional, carbon-intensive modes of transportation.

Through published research Physics C: Superconductivity and its ApplicationsNow, a team led by Ercan Caner Özkat at Recep Tayyip Erdogan University in Rize, Turkey, shows how artificial intelligence can help to improve the accuracy of measuring these forces. Their findings bring the widespread deployment of magnetic levitation systems one step closer to reality and could ultimately help reduce carbon emissions from the global transportation sector.

“With both the population and energy consumption growing rapidly, energy-efficient systems are crucial,” Ozcutt said. “Superconducting magnetic levitation systems serve this purpose due to their characteristic of frictionless motion.”

In magnetic levitation systems, the magnets consist of high-temperature superconductors, which allow them to transmit electric current without resistance, even at outdoor ambient temperatures. Two different forces have a particularly strong influence on their performance: levitation forces lift the train vertically, reducing friction with the track below, while lateral forces act perpendicular to the track and cause the train to rock from side to side.

To ensure sufficient load capacity and safe movement, it is important to enhance both the levitation force and the lateral guidance force of the magnetic levitation system. So far, multi-faceted interactions between superconductors and permanent magnet tracks have proven effective. However, mechanical constraints have made it difficult to experimentally measure the system so far.

Ozcutt's team investigated how this could be achieved with the help of artificial intelligence: “We were driven by the idea that some of the research challenges associated with magnetic levitation systems could be overcome through machine learning,” Ozcutt explains.

Machine learning is a branch of artificial intelligence that makes accurate predictions about a system's characteristics based on past experience with real-world data. Through their analysis, the researchers identified machine learning models that performed particularly well in predicting both the forces based on the properties of superconducting magnets, and the lateral and horizontal movements of maglev trains.

Based on this model, the researchers developed a step-by-step process for evaluating how to tune a magnetic levitation system to maximize levitation force while keeping lateral forces within certain tolerances. “We believe that the generalizable methodology we have presented can serve as a guide and a useful problem-solving tool for researchers working in this field,” says Ozcutt.

They now hope that their approach can help improve the efficiency and performance of real magnetic levitation systems, making them more practical and economical — which could accelerate their adoption in transportation systems around the world and bring the goal of net-zero carbon emissions one step closer to reality.