A research team from the School of Engineering and Materials Science at Queen Mary University of London (QMUL), has been awarded Third Place in the Traditional Evolutionary Algorithm Group of the Competition on Expensive Optimization for Mixed-Variable Optimization Problems at the 2026 IEEE World Congress on Computational Intelligence (WCCI 2026), held in Maastricht, the Netherlands.

WCCI is one of the world's leading conferences in computational intelligence, bringing together researchers from academia and industry in evolutionary computation, neural networks, and fuzzy systems. The competition challenged participants to develop algorithms for mixed-variable expensive optimisation, where problems involving both continuous and categorical (or discrete) decision variables, where evaluating a candidate solution is computationally or financially expensive.

A representative application is combustion research. Engineers aim to optimise fuel compositions and operating conditions to improve combustion efficiency while reducing emissions. Traditionally, this has relied on repeated physical experiments, which are both costly and time-consuming. Recent advances in artificial intelligence and high-fidelity simulation have made it possible to accurately predict combustion behaviour before conducting real experiments. This creates an opportunity to combine optimisation with simulation: optimisation algorithms can efficiently search for promising fuel mixtures and operating configurations using the simulation model, allowing only the most promising candidates to be validated through physical experiments. By significantly reducing the number of costly experiments required, this approach has the potential to accelerate scientific discovery while lowering research costs.

The team developed a novel optimisation algorithm, ASMLS-MiSACO, which demonstrated strong performance across the competition benchmark suite and was recognised with Third Place. The team consisted of Yinghao Qin, a PhD student in the Centre for Intelligent Transport, and Dr Jun Chen, Reader in Computational Intelligence at Queen Mary.

Reflecting on the achievement, Yinghao Qin said:

"It is a great honour to receive this award at WCCI 2026. The competition provided an excellent opportunity to benchmark our research against leading international teams and highlighted the growing importance of intelligent optimisation methods for solving real-world engineering problems."

The award highlights Queen Mary's continuing research strengths in computational intelligence and optimisation, and demonstrates how advanced optimisation algorithms can contribute to more efficient scientific research and engineering innovation.