Analysis and Prediction of Bore Pressure Based on BP ANN

Abstract:

The rule of Fire bore pressure and the maximum bore pressure point are very important problems in interior ballistics. This paper analyzed the characteristics of the bore pressure by an improved BP neural network, furthermore predicted the maximum bore pressure by the net-work well trained above. First, an improved three-layer BP neural network were established ,and trained by low frequency (20kHz) pressure data, and then several high frequency (50kHz,100kHz,200kHz) pressure data were fitted by the same network, furthermore the maximum bore pressure and the time of the maximum bore pressure were predicted by the network. Compared with the actual measured data, it was shown that this method also applied to the high frequency pressure data, and had a high fitting accuracy and prediction accurate. Above all, it proved the method is feasible and generic, can be applied to related dynamics problems, and has a value for application.

• Periodical: Advanced Materials Research (Volume 740)
• Pages: 82-85
• DOI: https://doi.org/10.4028/www.scientific.net/AMR.740.82
• Online since: August 2013
• Authors: Xiu Hui Tan, Yan Ping Bai, Da Wei Shen
• Keywords: Bore Pressure, BP ANN, Curve Fitting, Maximum Bore Pressure Prediction

References

[1] Bao Tingyu, Qiu Wenjian: Interior Ballistics M. Beijing: Beijing Institute of Technology Press, (1995)

[2] Cai Weimei, Yi Lianjun, Xu Wanhe. J. Sichuan Ordnance Journal. A New Interior Ballistic Pressure Calculation Method. Vo1 33, No.2, Feb.20l2

[3] Li Xin'e, Zu Jing, Ma Tiehua, Xu Peng. J. Chinese Journal of Scientific. Instrument Design of Capacitance Pressure Sensor for Gun Pressure Testing. Vol.32, No.3, Mar.20ll

[4] Runelbart D, Meclelland J. Parallel Distributed Processing: Explorations in the Microstructure of Cognition. Cambridge: Bradford Books, MIT Press, (1986)

[5] Cong Shuang. Matlab Toolbox for the Neural Network Theory and Application. M. Chinese Science and Technology University Press, 2008, 64