DOI: 10.4028/www.scientific.net/kem.416.248
View full text
Zhong Feng Pan, Gui Cheng Wang, Chong Lue Hua, Hong Jie Pei

Abstract: An improved neural network based on L-M algorithm has been applied to the prediction of the grind-hardening parameters against to the slow convergence rate of conventional BP neural network. And the the neural network model for grind-hardening is established. The neural network prediction system for grind-hardening process has been developed based on L-M algorithm. The functions of system is analyzed, particularly and some pivotal technology to realize the system are put forward.