Study on Large Deformation Prediction and Control Technology of Carbonaceous Slate Tunnel in Lixiang Railway

Table 4

Later prediction results.

Monitoring cycle (d)	Peripheral convergence (mm)	BP neural network model		GA-BP neural network model
		Estimate (mm)	Relative error (%)	Estimate (mm)	Relative error (%)
41	718.2	732.4	1.97	724.1	0.81
42	720.6	735.9	2.12	727.3	0.93
43	721.4	737.8	2.27	728.9	1.03
44	722.5	738.9	2.27	730.4	1.09
45	723.7	740.1	2.27	730.8	0.98
Average relative error			2.18		0.97