An efficient hybrid differential evolution–Jaya algorithm for enhancing vibration behaviour in automotive turbocharger systems

Abstract

Enhancing the vibration behaviour of rotating systems while meeting conflicting design needs is a difficult task, which is often treated as an inverse issue and formulated as a nonlinear optimization problem with constraints. To tackle this intricate challenge, a novel hybrid algorithm (HDE-Jaya) has been developed, combining the differential evolution (DE) and Jaya methods. The Jaya model is integrated into DE as a mutation operator to boost exploration and exploitation. In addition, the crossover probability is generated randomly at each iteration to preserve diversity. This makes HDE-Jaya a parameter-free algorithm. The performance of HDE-Jaya was validated by optimizing six common mechanical engineering design problems and further evaluated on an automotive turbocharger design, aiming to reduce vibrations, minimize binding stress and improve dynamic stability. The results indicate that HDE-Jaya is superior to other algorithms, including DE and the Jaya algorithm, in terms of solution accuracy and computational efficiency.