Application of neural network method to process planning in ship pipe machining

Abstract

Based on artificial neural network for process planning decision in ship pipe manufacturing, a novel method is established by analyzing process characteristics of the ship pipe machining. The process knowledge of pipe machining is shifted from the expression of the external rules to the description of the internal net weight value in order for the net inferring engine to decide the process route of pipe machining rapidly and rightly. Simulation shows that the method can resolve problems of process decision, and overcome the drawbacks of “matching difficulty” and “combination explosion” in traditional intelligent CAPP based on symbol reasoning.

Keywords

computer aided process planning (CAPP) / shipbuilding / pipe machining / back-propagation neural network (BPNN)

References

[1] Wei Chang, Wen-jian Liu, Zhi-wei Xu, et al. Expression of process knowledge based on neural network[J]. Journal of Harbin Institute of Technology, 2000, 32(3): 132-136 (in Chinese)

[2] Jia-xiang Dong, Shu-chun Wang, Xiao-qing Tang. The application of ANN to CAPP tool[J]. Journal of Astronautics, 1994, 15(10): 1212-1218 (in Chinese)

[3] Dan Wu, Xian-kui Wang, Jian Pu. Research on some key techniques of intelligent process planning system[J]. Journal of Tsinghua University (Sci & Tech), 1996, 36(4): 54-59 (in Chinese)

[4] Bao-rong Qin, Ning-sheng Wang. Research on Neural-Network-Based decision method for part shape machining process, 2002, 13(15): 1292-1295 (in Chinese)

[5] Wei-yue Xiao. Intelligent information processing technology in CAPP[M]. 2002, Changsha: National University of Defense Technology Press, (in Chinese)

[6] Jian-gang Yang. Tutorial of artifial neural network [M]. 2001, Hangzhou: Zhejiang University Press, (in Chinese)