Enhancing Vibration Performance Through an Efficient Hybrid DEJA Algorithm for Rotor System Design Parameter Optimization

The design of rotating systems presents a formidable challenge due to the multitude of conflicting design considerations that must be addressed. In this investigation, the design of an automotive turbocharger functioning as a rotor system is conceptualized using the finite element method and framed as a nonlinearly constrained optimization dilemma. The objective function is articulated as the total mass of the system. The constraint functions, which are imperative for compliance, are devised to mitigate vibration magnitude, confine bending stress within the shaft, and maintain the stability of the rotor system. To tackle this intricate difficulty, a new Hybrid DEJA algorithm is introduced, leveraging the fusion of Differential Evolution (DE) and Jaya (JA) methodologies. The JA framework is combined with a mutation operator into DE algorithm, to extend both exploration and exploitation throughout search process. Additionally, the crossover probability is stochastically generated for each iteration to sustain population diversity. A notable advantage of the proposed Hybrid DEJA lies in its freedom from specific control parameters. Furthermore, numerical findings underscore the superior performance of DEJA in terms of solution precision and computational efficiency compared to conventional DE and JA algorithms.