Multi-objective prediction and optimization of large-diameter slurry shield posture based on CatBoost-MOEAD

doi:10.16265/j.cnki.issn1003-3033.2024.10.1718

Abstract

Abstract:

To avoid abnormal attitude problems such as serpentine and axis deviation during shield tunneling affecting construction safety, a large-diameter slurry shield attitude control method combining CatBoost algorithm and MOEAD. A shield posture prediction model was developed with 19 input parameters and 6 output parameters, and the CatBoost algorithm was used to develop a nonlinear mapping relationship between input and output parameters. The SHAP was used to analyze the effects of input parameters on shield posture. The CatBoost-MOEAD shield posture multi-objective optimization model was coupled with the multi-objective optimization algorithm. Then the proposed model performance was validated against the Wuhan Yangtze River large-diameter slurry shield tunnel project. The results showed that the CatBoost prediction model can efficiently predict the posture of large-diameter mud-water shields. The determination coefficients of the six shield posture objectives ranged from 0.931 to 0.974, the root-mean-square errors ranged from 0.030 to 0.880, and the errors ranged from 0.039 to 1.057. The thrust of the propulsion group has the most significant impact on shield attitude among the major construction parameters. The proposed CatBoost-MOEAD multi-objective optimization method for shield attitude had a great performance in optimization effect with a maximum value of 38.86%.

Key words: categorical boosting (CatBoost), multi-objective evolutionary algorithm based on decomposition (MOEAD), large-diameter slurry shield, shield posture, multi-objective optimization, Shapley additive explanations (SHAP)

CLC Number:

X948
U455

WU Xianguo, LIU Jun, WANG Jingyi, QIN Yawei. Multi-objective prediction and optimization of large-diameter slurry shield posture based on CatBoost-MOEAD[J]. China Safety Science Journal, 2024, 34(10): 50-57.

Figures/Tables 9

Table 1

Initial value

变量	数据			变量	数据
变量	最大值	最小值	平均值	变量	最大值	最小值	平均值
$x 1$ / (mm·min^-1)	32.71	28.12	30.55	x₁₄ ×10⁵ Pa	8.59	5.46	7.04
$x 2$ / (r·min^-1)	1.21	0.91	1.12	x₁₅ ×10⁵ Pa	8.94	6.43	7.70
$x 3$ /kN	94 708	81 493	87 378	$x 16$ /m	0.16	0	0.06
$x 4$ /kN	11 440	6 218	9 296	x₁₇ ×10⁵ Pa	3.98	3.62	3.82
$x 5$ /kN	11 917	6 710	8 329	$x 18$ /kN	30.00	26.00	27.99
$x 6$ /kN	23 872	13 623	18 721	$x 19$ /(°)	16.00	8.00	11.98
$x 7$ /kN	15 135	11 901	13 559	$f 1$ /mm	-18	-54	-39
$x 8$ /kN	20 121	12 368	16 119	$f 2$ /mm	54	20	37
$x 9$ /kN	26 715	16 386	21 353	$f 3$ /mm	-44	-61	-52
$x 10$ / (t·m^-3)	1.43	1.36	1.39	$f 4$ / mm	50	39	45
$x 11$ / (t·m^-3)	1.49	1.41	1.45	$f 5$ / (mm· m^-1)	-24	-29	-27
$x 12$ / (m³·h^-1)	3 026	2 142	2 578	$f 6$ / (mm· m^-1)	4.65	3.03	3.93
$x 13$ / (m³·h^-1)	3 104	2 332	2 657	—	—	—	—

Table 1

Table 2

Hyperparameter optimization results of CatBoost under different shield posture prediction objectives

超参数	$f 1 /$ mm	$f 2 /$ mm	$f 3 /$ mm	$f 4 /$ mm	$f 5 /$ (mm·m^-1)	$f 6 /$ (mm·m^-1)
L2正则化参数	4	3	8	7	9	3
学习率	0.146	0.100	0.066	0.171	0.080	0.124
树的最大深度	6	8	6	6	6	8

Table 2

Fig.1

Table 3

Comparison of shield posture prediction accuracy

机器学习算法	预测精度	盾构姿态预测目标
机器学习算法	预测精度	$f 1$	$f 2$	$f 3$	$f 4$	$f 5$	$f 6$
CatBoost	R²	0.931	0.951	0.966	0.956	0.974	0.970
	RMSE	0.429	0.467	0.880	0.705	0.054	0.030
	MAE	0.535	0.578	1.057	0.824	0.065	0.039
XGBoost	R²	0.928	0.942	0.961	0.948	0.964	0.957
	RMSE	0.436	0.512	0.972	0.755	0.064	0.037
	MAE	0.547	0.631	1.145	0.903	0.076	0.046
RF	R²	0.924	0.935	0.954	0.942	0.953	0.937
	RMSE	0.446	0.551	1.058	0.805	0.074	0.045
	MAE	0.562	0.669	1.241	0.948	0.087	0.056
SVM	R²	0.916	0.925	0.946	0.930	0.944	0.926
	RMSE	0.470	0.589	1.126	0.878	0.080	0.050
	MAE	0.589	0.716	1.336	1.044	0.095	0.060
BPNN	R²	0.908	0.920	0.937	0.913	0.929	0.916
	RMSE	0.494	0.606	1.245	0.985	0.091	0.053
	MAE	0.615	0.739	1.446	1.163	0.107	0.065

Table 3

Fig.2

Table 4

Input parameter range

输入参数	参数范围	输入参数	参数范围
$x 2 /$ (r·min^-1)	[0.9, 1.3]	$x 8$ /kN	[12 000, 20 200]
$x 3 /$ kN	[81 500, 94 800]	$x 9$ /kN	[16 300, 26 800]
$x 4 /$ kN	[6 200, 12 000]	$x 10$ /(t·m^-3)	[1.3, 1.5]
$x 5 /$ kN	[6 200, 12 000]	$x 11$ /(t·m^-3)	[1.3, 1.5]
$x 6 /$ kN	[13 600, 239 000]	$x 12$ /(m³·h^-1)	[2 140, 3 030]
$x 7 /$ kN	[12 000, 15 200]	x₁₇×10⁵ Pa	[3.3, 3.4]

Table 4

Table 5

Adjustment parameters of each scenario

场景	调整参数
场景1	$x 2 、 x 3 、 x 4 ~ x 9 、 x 10 、 x 11$
场景2	$x 2 、 x 3 、 x 4 ~ x 9 、 x 10 、 x 11 、 x 17$
场景3	$x 2 、 x 3 、 x 4 ~ x 9 、 x 10 、 x 11 、 x 12 、 x 17$

Table 5

Table 6

Shield posture before and after optimization

盾构姿态指标	初始值	优化值	优化占比/%
$f 1$ /mm	-39.42	-23.36	40.75
$f 2$ /mm	37.07	21.81	41.17
$f 3$ /mm	-51.97	-34.17	34.26
$f 4$ /mm	44.67	28.95	35.19
$f 5$ /(mm·m^-1)	-26.79	-18.63	30.48
$f 6$ /(mm·m^-1)	3.93	3.60	8.39

Table 6

Table 7

Optimization results of different scenarios

场景		$f 1$ /mm	$f 2$ /mm	$f 3$ /mm	$f 4$ /mm	$f 5$ /(mm·m^-1)	$f 6$ /(mm·m^-1)	平均优化率/%
场景1	优化后数值	-23.356	21.811	-34.166	28.951	-18.627	3.604	36.35
场景1	改进占比/%	40.75	41.17	34.26	35.19	30.48	8.39	36.35
场景2	优化后数值	-22.205	21.705	-33.423	28.905	-18.583	3.538	37.41
场景2	改进占比/%	43.67	41.45	35.69	35.30	30.64	10.07	37.41
场景3	优化后数值	-21.662	21.354	-32.148	28.572	-18.367	3.391	38.86
场景3	改进占比/%	45.05	42.40	38.14	36.04	31.45	13.80	38.86
平均值/%		43.16	41.67	36.03	35.51	30.86	10.75	37.54

Table 7

References 10

[1]	黄威, 任梦, 陈培帅, 等. 盾构水平姿态的理论分析模型[J]. 隧道建设:中英文, 2022, 42(1): 83-89.
	HUANG Wei, REN Meng, CHEN Peishuai, et al. Theoretical analysis model of shield horizontal attitude[J]. Tunnel Construction, 2022, 42(1): 83-89.
[2]	SUGIMOTO M, SRAMOON A. Theoretical model of shield behavior during excavation. I: theory[J]. Journal of Geotechnical and Geoenvironmental Engineering, 2002, 128(2): 138-155.
[3]	苏栋, 谭毅俊, 沈翔, 等. 软土地层加固对盾构姿态调控及地层变形的影响研究[J]. 现代隧道技术, 2023, 60(2): 138-148,167.
	SU Dong, TAN Yijun, SHEN Xiang, et al. A study on impact of soft soil stratum reinforcement on the attitude regulation of shield machine and stratum deformation[J]. Modern Tunnelling Technology, 2023, 60(2): 138-148,167.
[4]	SUN Wei, YUE Ming, WEI Jian. Relationship between rectification moment and angle of shield based on numerical simulation[J]. Journal of Central South University, 2012, 19(2): 517-21.
[5]	曹化锦. 基于机器学习和非支配排序遗传算法的盾构姿态预测与优化[J]. 铁道建筑, 2023, 63(7): 93-97.
	CAO Huajin. Prediction and optimization of shield posture based on machine learning and non-dominated sorting genetic algorithm[J]. Railway Engineering, 2023, 63(7): 93-97.
[6]	徐进, 林良宇, 章龙管, 等. 基于深度学习的盾构掘进姿态预测模型[J]. 地下空间与工程学报, 2022, 18(增2): 813-821.
	XU Jin, LIN Liangyu, ZHANG Longguan, et al. Prediction model of shield tunneling attitude based on deep learning[J]. Chinese Journal of Underground Space and Engineering, 2022, 18(S2): 813-821.
[7]	SCAVUZZO C M, SCAVUZZO J M, CAMPERO M N, et al. Feature importance: Opening a soil-transmitted helminth machine learning model via SHAP[J]. Infectious Disease Modelling, 2022, 7(1): 262-276. doi: 10.1016/j.idm.2022.01.004 pmid: 35224316
[8]	吴贤国, 冯宗宝, 刘俊, 等. 基于RF-NSGA-Ⅱ的盾构施工地表沉降安全控制多目标优化[J]. 中国安全科学学报, 2022, 32(8): 45-51. doi: 10.16265/j.cnki.issn1003-3033.2022.08.2702
	WU Xianguo, FENG Zongbao, LIU Jun, et al. Multi-objective optimization of surface settlement safety control during shield construction based on RF-NSGA-Ⅱ[J]. China Safety Science Journal, 2022, 32(8): 45-51. doi: 10.16265/j.cnki.issn1003-3033.2022.08.2702
[9]	阮顺领, 韩思淼, 张宁宁, 等. 基于CNN-aGRU融合模型的尾矿坝浸润线预测方法[J]. 中国安全科学学报, 2023, 33(增1): 119-127.
	RUAN Shunling, HAN Simiao, ZHANG Ningning, et al. Prediction method of saturation line of tailings dam based on CNN-aGRU fusion model[J]. China Safety Science Journal, 2023, 33(S1): 119-127. doi: 10.16265/j.cnki.issn1003-3033.2023.S1.2481
[10]	GB 50446—2017,盾构法隧道施工与验收规范[S].
	GB 50446-2017,Code for construction and acceptance of shield tunnelling method[S].

[1]	WU Xianguo, LIU Jun, CAO Yuan, LEI Yu, LI Shifan, QIN Yawei. Shield attitude prediction and optimization based on CatBoost-NSGA-III algorithm [J]. China Safety Science Journal, 2024, 34(8): 69-77.
[2]	WU Xianguo, LIU Jun, SU Feiming, CHEN Hongyu, FENG Zongbao. Multi-objective prediction optimization for large-diameter slurry shield tunneling construction based on CatBoost-MOEAD [J]. China Safety Science Journal, 2024, 34(6): 57-64.
[3]	MA Tianming, HUANG Chuyuan, CHEN Xianfeng. Heterogeneous vehicle routing problem of hazardous materials transportation considering carbon emissions [J]. China Safety Science Journal, 2024, 34(1): 193-199.
[4]	CHAI Huo, HE Ruichun, JIA Xiaoyan, DAI Cunjie. Vehicle scheduling for hazardous materials transportation based on dissimilar paths [J]. China Safety Science Journal, 2023, 33(9): 181-188.
[5]	ZHANG Yanping, CUI Na, ZHANG Bo. Multi-objective optimization strategy of emergency material supply based on multi-mode transportation [J]. China Safety Science Journal, 2023, 33(7): 230-238.
[6]	WANG Hui, ZHOU Guoqiang, WANG Yujian, YUE Xingqi, ZHANG Yiming. Structural parameter analysis and optimization of pneumatic safety valve in aviation propulsion system [J]. China Safety Science Journal, 2023, 33(2): 48-58.
[7]	ZHANG Minbo, ZHONG Ziyi, YAN Jin, WANG Cuiling, WANG Zichao, LI Chunxin. Model of emergency supplies allocation in epidemic areas considering spread of biohazard diffusion [J]. China Safety Science Journal, 2023, 33(11): 206-213.
[8]	WU Xianguo, FENG Zongbao, LIU Jun, WANG Lei, CHEN Hongyu, LI Xinyi. Multi-objective optimization of surface settlement safety control during shield construction based on RF-NSGA-II [J]. China Safety Science Journal, 2022, 32(8): 45-51.
[9]	HONG Liang, LIU Gang, GE Ruhai, LI Liang, LI Siyuan. Multi-objective optimization of normally-open airbag of school buses under mixed conditions [J]. China Safety Science Journal, 2022, 32(4): 72-79.
[10]	JI Xiaofeng, YANG Chunli, HAO Jingjing, QIN Wenwen. Research prospects and hotspot comparison for emergency logistics in China and foreign countries [J]. China Safety Science Journal, 2021, 31(12): 144-152.
[11]	DU Xiuli, CAO Xiuxiu, ZHONG Zilan, HOU Benwei, WANG Wei. Robust geotechnical design of soil-pile foundation system considering multiple failure modes [J]. China Safety Science Journal, 2020, 30(6): 23-30.
[12]	ZHAO Jianqiang, XIONG Yanming, ZHANG Youjian. Multi-objective optimization design for safety system based on isolation-inoperability competing failure mechanism [J]. China Safety Science Journal, 2020, 30(6): 1-7.
[13]	MA Hui, ZHANG Wenjing, DONG Meihong. Spatial conflict analysis and multi-objective optimization for prefabricated building hoisting construction [J]. China Safety Science Journal, 2020, 30(2): 28-34.
[14]	SONG Yinghua, SU Beibei, HUO Feizhou, NING Jingjing, FANG Danhui. Research on rapid location selection of emergency materials distribution center considering dynamic demand [J]. China Safety Science Journal, 2019, 29(8): 172-177.
[15]	SONG Ye, SONG Yinghua, LIU Dan, WANG Zhe. Earthquake emergency rescue teams assignment model based on time satisfaction and competence [J]. China Safety Science Journal, 2018, 28(8): 180-185.