[1] [1] YANG L J, CHANDEL R S, BIBBY M J. The effects of processvariables on the bead height of submerged-arc weld deposits［J］. Canadian Metallurgical Quarterly, 1992, 31(4): 289-297.

[2] [2] YANG Z H, LI X Q, WANG Z M, et al. The bead geometry predicting model based on partial least squares regression［J］. Welding Technology, 2010, 39(3): 6-9.

[3] [3] XU M, LIU S Y, LI Y Q, et al. Optimization of SPCC and 65Mn dissimilar metal laser welding process parameters based on response surface methodology［J］. Applied Laser, 2017, 37(3): 362-366.

[4] [4] HONG Y W, LIU S Y, XU C Y, et al. The calculationof laser-arc hybrid weld bead shape basing on multiple population genetic algorithm and neural-networks［J］. Applied Laser, 2015, 35(6): 677-683.

[5] [5] MURUGAN N, GUNARAJ V. Prediction and control of weld bead geometry and shape relationships in submerged arc welding of pipes［J］. Journal of Materials Processing Technology, 2005, 168(3): 478-487.

[6] [6] RIDINGS G E, THOMSON R, THEWLIS G. Prediction of multiwire submerged arc weld bead shape using neural network modelling［J］. Science and Technology of Welding and Joining, 2002, 7: 265- 279.

[7] [7] DONG Z B, WEI Y H, ZHAN X H, et al. Optimization of mechanical properties prediction models of welded joints combined neural network with genetic algorithm［J］. Transactions of the China Welding Institution, 2007, 28(12): 69-72.

[8] [8] LIU S Y, ZHANG H, SHI Y, et al. Effects of process parameters on droplet transfer and bead shape in CO2-MAG hybrid welding［J］. Chinese Journal of Lasers, 2010, 37(12): 3172-3179.

[9] [9] L Q B, DAI H, TAN K L, et al. Quality prediction of alternating current flash butt welding of rail based on improved back propagation neural network［J］. Transactions of the China Welding Institution, 2005, 26(5): 65-68.