An enhanced GA technique for system training and prognostics

The commonly used genetic algorithm (GA)-based methods have some shortcomings in applications such as time-consuming and slow convergence. A novel enhanced genetic algorithm (EGA) technique is developed in this paper to overcome these problems in classical GA methods so as to provide a more efficient technique for system training and optimization. Two approaches are proposed in the EGA technique: Firstly, a novel group-based branch crossover operator is suggested to thoroughly explore local space and speed up convergence. Secondly, an enhanced MPT (Makinen–Periaux–Toivanen) mutation operator is proposed to promote global search capability. The effectiveness of the developed EGA is verified by simulations based on a series of benchmark test problems. The EGA technique is also implemented to train a neural-fuzzy predictor for real-time gear system monitoring. Test results show that the branch crossover operator and enhanced MPT mutation operator can effectively improve the convergence speed and global search capability. The EGA technique outperforms other related GA methods with respect to convergence speed and global search capability.