论文信息 - Grinding precision forecasting in optical aspheric grinding using artificial neural network and genetic algorithm

Grinding precision forecasting in optical aspheric grinding using artificial neural network and genetic algorithm

A new approach based on an artificial neural network (ANN) was presented for the prediction of machining precision of optical aspheric grinding. The ANN model is based on Globally Convergent Adaptive Quick Back Propagation algorithm (GCAOBP). A genetic algorithm (GA) was then applied to the trained ANN model to predict the gridding precision. The integrated GCAOBP-GA algorithm was successful in predicting the Root Mean Square of profile error (RMS) of optical aspheric workpiece in parallel grinding method using machining parameters. The results of experiments have shown that RMS of machined workpiece in parallel grinding can be predicted effectively through this approach.

Chen Jiang | Yinbiao Guo | Qingqing Yang | Chunguang Han

[1] Li Jian-zhen. A learning algorithm of artificial neural network based on genetic algorithm , 2002 .

[2] Kuang-Hua Fuh,et al. Force modeling and forecasting in creep feed grinding using improved bp neural network , 1997 .

[3] Zhang Yi,et al. INFL UENCE OF CHATTER VIBRATION ON THE UL TRAPRECISION MACHINING ACCURACY OF ASPHERIC SURFACE , 2003 .

[4] Qiu Wei. Globally convert adaptive quick back propagation algorithm based on neural network , 2004 .

[5] Sounak Kumar Choudhury,et al. Prediction of wear and surface roughness in electro-discharge diamond grinding , 2007 .

[6] Peter Willett,et al. Neural network detection of grinding burn from acoustic emission , 2001 .