Current status and prospects of research on landslide disasters in mine slopes based on multi-source information fusion
-
摘要: 为克服单一信息源无法精确表征矿山滑坡灾害演化特征的问题,基于多源信息融合技术,从矿山边坡多源信息获取、矿山边坡多源信息融合、矿山边坡位移预测及滑坡风险评价3个方面概述了矿山边坡滑坡灾害研究进展。总结了典型的“天”“空”“地”边坡监测手段及“天−空−地”一体化协同监测方法;梳理了包含数据级、特征级和决策级融合的边坡多源信息融合流程;整理了位移与应力、位移与水文气象及其他不同类型的监测数据融合形式;阐述了基于多源信息融合的边坡位移预测及滑坡风险评价相关研究现状。基于当前矿山边坡滑坡灾害研究存在的灾害分析的准确性严重依赖监测数据质量、对岩石力学机理知识利用不足等问题,指出了矿山边坡滑坡灾害研究发展趋势:统一多源数据采集接入标准;开发监测数据与岩石力学机理融合的矿山边坡滑坡灾害分析方法;优化“天−空−地”多源信息的时空关联挖掘算法;加强基于多源信息融合的矿山边坡滑坡灾害预警平台建设。Abstract: In order to overcome the problem that a single information source cannot accurately characterize the evolution features of mining landslide disasters, based on multi-source information fusion technology, this paper summarizes the research progress of mine slope landslide disasters from three aspects: multi-source information acquisition of mine slopes, multi-source information fusion of mine slopes, and mine slope displacement prediction and landslide risk assessment. The study summarizes typical slope monitoring methods of "sky", "air", and "ground" , as well as integrated collaborative monitoring method of "sky-air-ground". The study sorts out the slope multi-source information fusion process that includes data level, feature level, and decision level fusion. The paper organizes the fusion forms of displacement and stress, displacement and hydrological and meteorological monitoring information, as well as other different types. This paper elaborates on the current research status of slope displacement prediction and landslide risk assessment based on multi-source information fusion. The accuracy of disaster analysis in current research on mine slope landslide disasters heavily depends on the quality of monitoring data and insufficient utilization of knowledge of rock mechanics mechanisms. Based on the above problems, the development trends of research on landslide disasters in mine slopes are pointed out. The multi-source data collection and access standards are unified. The method for analyzing landslide disasters in mine slopes is developed by integrating monitoring data with rock mechanics mechanisms. The spatiotemporal association mining algorithm for multi-source information from the "sky-air-ground" is optimized. The construction of a mine slope landslide disaster warning platform based on multi-source information fusion is strengthened.
-
0. 引言
据统计,近10年来煤矿安全事故率持续下降,但仍在高危行业前列[1]。为了保障井下人员安全生产,矿工在井下作业过程中必须佩戴防护设备。由于部分矿工安全意识低,对于防护设备不是很重视,不能有效佩戴防护设备,保障自身安全。目前煤矿企业主要通过人工及视频监控摄像头查看的方法来监督矿工是否佩戴防护设备。由于井下作业环境的监控摄像头位置固定,其覆盖范围广、拍摄距离远,监控画面中防护设备目标的尺寸较小,导致小型防护设备目标易受到目标尺寸和井下环境变化的影响,检测难度大大增加。因此,对煤矿井下场景的小目标(尺寸小于32×32的目标)进行检测研究,在小型防护设备监测中起至关重要的作用。
传统的小目标检测方法[2-4]难以有效提取井下小目标特征信息,随着深度学习的兴起,卷积神经网络(Convolutional Neural Networks,CNN)模型[5]逐步代替了传统小目标检测方法,基于CNN的小目标检测方法主要分为两阶段目标检测和单阶段目标检测。两阶段目标检测算法通过生成区域网络(Region Proposal Network,RPN)提取到感兴趣的特征信息后进行分类,例如R−CNN算法(Region with CNN feature)[6]、Fast R−CNN算法[7]及Faster R−CNN算法[8]等,该类算法需生成大量候选区域,检测速度慢,无法满足对小目标实时检测的要求。单阶段目标检测算法将检测归纳为回归问题,实现端到端的检测技术,如单步多框目标检测(Single Shot MultiBox Detector,SSD)算法[9]、YOLO系列算法[10-14],该类检测算法的检测速度较快,但会有一定的误差。基于CNN的小目标检测方法较传统方法有很大提升,但仍然存在召回率低、误检率高的问题。针对上述问题,文献[15]为了准确处理和提取小目标信息特征,在YOLOv3网络模型的特征金字塔网络中自适应融合浅层和深层特征图的局部和全局特征。文献[16]提出了一种轻量型特征提取模块,该模块采用空洞瓶颈和多尺度卷积获得更丰富的图像特征信息,增强了目标特征表达能力。文献[17]在YOLOv5s模型Backbone区域嵌入自校正卷积(Self-Calibrated Convolution,SCConv)作为特征提取网络,可更好地融合多尺度特征信息。文献[18]提出了一种结合通道和空间注意力引导的残差学习方法,用于捕捉目标的关键信息。文献[19]提出了高分辨率表示模块,通过使用多尺度特征来捕捉目标的细节信息,并将其融合到高分辨率表示模块,有助于提高目标的定位准确性。
上述研究虽然提高了小目标检测效果,但针对的多为常规场景,受煤矿井下恶劣环境影响,在检测过程中存在井下小目标特征信息提取困难的问题。针对该问题,本文提出一种基于YOLOv7−SE的井下场景小目标检测方法。首先,将模拟退火(Simulated Annealing,SA)算法与k−means++聚类算法融合,优化YOLOv7网络模型中初始锚框值。然后,在YOLOv7网络模型骨干网络中增加新的检测层,得到井下小目标信息丰富的特征图。最后,在YOLOv7网络模型骨干网络的聚合网络模块之后引入双层注意力机制[20],强化井下小目标特征表示。
1. YOLOv7网络模型
YOLOv7网络模型[21]主要由输入层、骨干网络、检测头3个模块组成,如图1所示。
输入层将图像缩放到固定尺寸640×640,以满足骨干网络对尺寸输入的要求。骨干网络由卷积模块、聚合网络模块和最大池化模块组成,对输入图像进行特征提取。卷积模块主要由1个卷积层、1个批量归一化层和1个SiLU激活函数构成。聚合网络模块包含3个1×1卷积层和4个3×3卷积层。最大池化模块由1个最大池化层、2个1×1卷积层和1个3×3卷积层组成。检测头由路径聚合特征金字塔网络(Path Aggregation Feature Pyramid Network,PAFPN)和检测层组成,PAFPN用于融合不同尺寸的特征图,检测层输出带有检测类别和准确度的结果。
2. YOLOv7−SE小目标检测方法
2.1 融合SA和k−means++聚类算法
YOLOv7网络模型的初始锚框尺寸是针对井下常规目标设计,难以适用于井下小目标检测,因此,使用k−means++聚类算法对YOLOv7网络模型进行聚类分析,生成新的锚框尺寸,若随机选取的初始化聚类中心点在密度小的簇内,会使其相似性较小,聚类结果较差。因此,利用SA算法[22]来确定k−means++聚类算法中最优的首个初始聚类中心点。SA算法是寻找数据集样本中最优值的算法,先初始化SA算法的参数,如初始温度、终止温度和迭代次数。通过迭代的方式,不断更新聚类中心点的值,并计算目标函数的值。根据SA策略判断是否接受当前参数的更新,若接受,则更新聚类中心点的值,否则降低温度继续迭代下一次参数值,直到达到终止温度,返回最优的聚类中心点。
将SA算法与k−means++聚类算法进行融合,并对本文数据集进行聚类分析,得到的锚框尺寸分别为[9,31],[9,17],[15,37],[16,23],[21,29],[26,40],[29,71],[39,55],[50,79]。
2.2 新的目标检测层
YOLOv7网络模型中深层网络提取抽象的语义特征信息,用来反映井下大目标特征信息,而浅层网络提取目标的细节特征,能够保留更多的井下小目标特征信息。YOLOv7网络模型通常通过增加下采样结构以获取更大的目标感受野,但随着下采样结构的增加,井下小目标的特征信息逐渐丢失,对于在图像中特征信息少的井下小目标不友好,不利于井下小目标检测。
为了使YOLOv7网络模型重点提取浅层网络中丰富的井下小目标特征信息,而抑制深层网络的特征提取,本文去掉YOLOv7网络模型骨干网络的32倍下采样结构,在YOLOv7网络模型中增加1层新的目标检测层,获取浅层网络中细节信息更明确的井下小目标特征,减少在下采样过程中丢失井下小目标特征。YOLOv7网络模型从第1个聚合网络模块开始提取特征信息,经过PAFPN输出到检测层,得到新尺寸(160×160)的特征图,该特征图的井下小目标信息相对丰富。
2.3 引入双层注意力机制
针对煤矿复杂环境下小目标图像特征抽取不准确的问题,本文在YOLOv7网络模型的聚合网络模块之后添加双层注意力机制,以强化聚合网络模块对井下小目标的特征提取能力。
双层注意力机制整体结构如图2所示。第1阶段使用重叠块嵌入,将原始的二维输入图像转换为一维的图像块。第2阶段到第4阶段使用合并块模块进行下采样操作,用于调整相应通道数,并降低输入分辨率,同时在每个阶段的操作后采用双层注意力模块(图3)对特征图做特征变换。
通过收集前n个相关区域中的键值对,并利用稀疏性操作忽略最不相关区域的计算来节省模块参数量和计算量,并对收集的键值对进行注意力操作。
$$ {\boldsymbol{O}} = {A_{{\mathrm{ttention}}}}({\boldsymbol{Q}},{{\boldsymbol{K}}^{\mathrm{T}}},{{\boldsymbol{V}}^{\mathrm{T}}}) + {L_{{\mathrm{CE}}}}({\boldsymbol{V}}) $$ (1) 式中:O为输出;Attention(·)为自注意力函数;Q为查询;KT为键的张量;VT为值的张量;LCE(·)为局部上下文增强模块;V为值。
$$ {A_{{\mathrm{ttention}}}}({\boldsymbol{Q}},{\boldsymbol{K}},{\boldsymbol{V}}) = S\left(\frac{{{\boldsymbol{Q}}{{\boldsymbol{K}}^{\mathrm{T}}}}}{{\sqrt d }}\right){\boldsymbol{V}} $$ (2) 式中:K为健;S(·)为归一化函数;d为缩放因子;QKT为Q和K之间的相似程度。
对YOLOv7进行上述改进,将改进后的网络模型命名为YOLOv7−SE。
3. 实验结果与分析
3.1 数据集
本文数据集来源于煤矿井下多个场景的视频监控摄像头拍摄的图像视频,包括采煤工作面、胶带机头工作面、井下巷道、井下站台、煤壁面、井下候车点等场景,如图4所示。
为提升数据集的多样性,通过水平、垂直翻转及随机方向旋转等方法扩充数据。整理后共有5 622张图像,将数据集按照8∶1∶1的比例划分为训练集、验证集和测试集。同时,使用labelImg工具对数据集进行标注,在满足小目标尺寸的条件下,标注类别有安全帽和自救器。
为解决井下环境中煤尘对图像的干扰,在YOLOv7网络模型的数据预处理阶段增加图像处理模块,对数据集进行预处理,如图5所示。
首先,采用暗通道去雾方法[23]对原始图像进行去雾,以减少井下煤尘对图像的影响。然后,采用高斯函数对去雾后的图像进行锐化,以突出图像细节,提高井下场景目标边缘与周围像素之间的反差。最后,使用卷积模块作为调优器,利用其反向传播的特性对图像处理方法中的去雾程度和锐化强度进行优化,以达到更好的增强效果。
3.2 评价指标
常用的小目标评价指标包括准确率P、召回率R、平均精度(Average Precision,AP)、所有类别的平均精度值(mean Average Precision,mAP)及每秒传输帧数( Frames Per Second,FPS)。
$$ P = \frac{{{N_{{\text{TP}}}}}}{{{N_{{\text{TP}}}} + {N_{{\text{FP}}}}}} $$ (3) $$ R = \frac{{{N_{{\text{TP}}}}}}{{N_{{\text{TP}}}^{} + {N_{{\text{FN}}}}}} $$ (4) $$ {\text{AP}} = \int_0^1 {PR{\text{d}}R} $$ (5) $$ {\text{mAP}} = \frac{{\displaystyle\sum {{\text{AP}}} }}{M} $$ (6) 式中:NTP为预测正确的正样本数量;NFP为预测错误的正样本数量;NFN为预测错误的负样本数量;M为目标的类别数。
3.3 实验配置
本文实验在ubuntu20.04操作系统中搭建,具体配置见表1。
表 1 实验环境配置Table 1. Experimental environment configuration实验环境 配置 GPU RTX 3090(24 GiB) CPU 12 vCPU Xeon(R) Platinum 8255C 操作系统 ubuntu20.04 GPU环境 CUDA11.3 cuDNN8.2.1 深度学习框架 Pytorch1.11 编译器 Python3.8 3.4 结果分析
3.4.1 模型训练
在模型训练前需对实验超参数进行设置,迭代次数为200,初始学习率为0.015,批量大小为32,选取640×640的图像作为模型的输入。在模型训练过程中损失函数值随迭代次数变化曲线如图6所示。
由图6可看出在模型训练过程前80次迭代,损失函数值下降十分明显,第80—180次迭代时,损失函数值下降趋势趋于平缓,最后20次迭代的损失函数值已逐渐稳定。模型训练过程中的最终损失函数值低于0.05,说明本文模型的训练参数设置合理,模型学习效果较好。
3.4.2 对比实验
为了衡量YOLOv7−SE网络模型的检测性能,将其与Faster R−CNN,RetinaNet,CenterNet,FCOS(Fully Convolutional One-Stage),SSD,YOLOv5,YOLOv7目标检测模型进行对比,结果见表2。
表 2 各模型对比结果Table 2. Comparison results of each model模型 AP/% mAP/% FPS/(帧·s−1) 安全帽 自救器 Faster R−CNN 58.64 52.11 55.38 19.28 RetinaNet 47.20 44.32 45.76 17.41 CenterNet 56.37 49.26 52.82 30.10 FCOS 59.79 56.13 57.96 22.00 SSD 56.97 45.06 51.02 39.10 YOLOv5 60.91 54.84 57.88 38.50 YOLOv7 60.30 57.10 58.70 71.20 YOLOv7−SE 72.50 64.48 68.49 61.84 由表2可看出,YOLOv7−SE网络模型的安全帽检测AP较Faster R−CNN,RetinaNet,CenterNet,FCOS,SSD,YOLOv5,YOLOv7分别提升了13.86%,25.3%,16.13%,12.71%,15.53%,11.59%,12.20%。YOLOv7−SE网络模型的自救器检测AP较FasterR−CNN,RetinaNet,CenterNet,FCOS,SSD,YOLOv5,YOLOv7分别提升了12.37%,20.16%,15.22%,8.35%,19.42%,9.64%,7.38%。YOLOv7−SE网络模型的mAP较Faster R−CNN,RetinaNet,CenterNet,FCOS,SSD,YOLOv5,YOLOv7分别提升了13.11%,22.73%,15.67%,10.53%,17.47%,10.61%,9.79%。YOLOv7−SE网络模型的FPS较Faster R−CNN,RetinaNe,CenterNet,FCOS,SSD,YOLOv5分别提升了42.56,44.43,31.74,39.84,22.74,23.34帧/s,较YOLOv7下降9.36帧/s。说明YOLOv7−SE网络模型的检测性能更佳。
3.4.3 消融实验
为了验证不同改进方法对YOLOv7网络模型性能的影响,设计了6组实验,改进的骨干网络实验为添加新的目标检测层和引入双层注意力机制,实验结果见表3。
表 3 消融实验结果Table 3. Results of ablation experiment模型 AP/% mAP/% FPS/(帧·s−1) 安全帽 自救器 YOLOv7 60.30 57.10 58.70 71.20 YOLOv7+改进的k−means++ 63.21 60.70 61.95 74.20 YOLOv7+改进骨干网络 70.70 62.32 66.51 63.18 YOLOv7−SE 72.50 64.48 68.49 61.84 由表3可看出,使用改进的k−means++方法重新聚类分析锚框值,安全帽检测AP、自救器检测AP、mAP、FPS分别为63.21%,60.7%,61.95%,74.2帧/s,较YOLOv7分别提升了2.91%,3.6%,3.25%,3帧/s;改进YOLOv7网络模型骨干网络后,安全帽检测AP、自救器检测AP、mAP分别为70.7%,62.32%,66.51%,较YOLOv7分别提升了10.4%,5.22%,7.81%,FPS下降8.02帧/s,为63.18帧/s;YOLOv7−SE网络模型的安全帽AP、自救器AP、mAP分别为72.5%,64.48%,68.49%,较YOLOv7分别提升了12.2%,7.38%,9.79%,FPS为61.84帧/s,说明YOLOv7−SE模型在保证检测速度的同时,有效强化了YOLOv7−SE网络模型对井下小目标的特征提取能力。
3.4.4 检测效果对比分析
为更加直观地体现YOLOv7−SE网络模型的优越性,在井下采煤工作面、胶带机头工作面、井下巷道及井下站台等场景中与YOLOv7网络模型对比,对比结果如图7所示。
由图7可看出,在对安全帽和自救器的检测中,YOLOv7网络模型出现漏检和误检的问题,而YOLOv7−SE网络模型有效改善了该问题,提高了检测精度。因此,YOLOv7−SE网络模型可满足井下小目标检测任务。
4. 结论
1) 针对煤矿井下场景中目标尺寸较小、环境存在大量煤尘导致小目标特征提取困难等问题,将SA算法与k−means++聚类算法融合,在YOLOv7骨干网络中增加新的目标检测层,同时将双层注意力机制嵌入聚合网络模块之后。YOLOv7−SE网络模型安全帽检测AP、自救器检测AP、mAP分别为72.50%,64.48%,68.49%,较YOLOv7网络模型分别提升了12.2%,7.38%,9.79%,FPS为61.84帧/s。
2) 将YOLOv7−SE网络模型与Faster R−CNN,RetinaNet,CenterNet,FCOS,SSD,YOLOv5,YOLOv7进行对比,实验结果表明,YOLOv7−SE网络模型对安全帽和自救器的检测精度最高。
3) 在对安全帽和自救器的检测中,YOLOv7−SE网络模型有效改善了漏检和误检问题,提高了检测精度。
【编者按】煤矿灾害一旦发生,其影响范围和严重程度非常大;不同灾害之间可能存在相互影响和关联,使得灾害预防和应对变得复杂;煤矿灾害影响一般不会随着灾害的结束而立即消除;预防和应对煤矿灾害需要专业知识和技能等。针对煤矿灾害的多数工作都是建立在有效感知灾害的前提之上。但通过灾害数据和现象来感知灾害十分困难,特别是在数据不全、及时性和实时性差等条件下更为困难。随着智能感知技术的成熟、应用和实践,为实现煤矿灾害感知提供了智能方法。为介绍和推动智能感知技术在煤矿灾害预测中的应用,交流相关理论方法和研究成果,《工矿自动化》编辑部特邀沈阳理工大学崔铁军教授担任客座主编,于2024年第6期组织出版“煤矿灾害智能感知新技术与实践”专题。在专题出刊之际,衷心感谢各位专家学者的大力支持! -
[1] 李红毅,李永华,高会有. 白云鄂博矿区凸形边坡多因素影响下滑坡破坏模式研究[J]. 中国安全生产科学技术,2023,19(增刊1):98-103. LI Hongyi,LI Yonghua,GAO Huiyou. Study on landslide failure mode under the influence of multiple factors on convex slopes in Baiyunebo mining area[J]. Journal of Safety Science and Technology,2023,19(S1):98-103.
[2] 李树建,张义平. 晋宁磷矿采场边坡灾害及其防治研究[J]. 化工矿物与加工,2014,43(8):39-42. LI Shujian,ZHANG Yiping. Hazards of pit slope in Jinning Phosphate Mine and its prevention[J]. Industrial Minerals & Processing,2014,43(8):39-42.
[3] 冯巩,夏元友,王智德,等. 基于位移信息融合的露天矿边坡动态预警方法[J]. 中国安全科学学报,2022,32(3):116-122. FENG Gong,XIA Yuanyou,WANG Zhide,et al. Dynamic early warning method of open-pit mine slopes based on integrated displacement information[J]. China Safety Science Journal,2022,32(3):116-122.
[4] 曹兰柱,孙成亮,王东,等. FLAC3D的露天矿边坡变形破坏数值模拟分析[J]. 辽宁工程技术大学学报(自然科学版),2016,35(7):679-682. CAO Lanzhu,SUN Chengliang,WANG Dong,et al. Simulation analysis of damaged degeneration of open-pit mine slope based on FLAC3D value[J]. Journal of Liaoning Technical University (Natural Science),2016,35(7):679-682.
[5] 宋卫东,杜建华,杨幸才,等. 深凹露天转地下开采高陡边坡变形与破坏规律[J]. 北京科技大学学报,2010,32(2):145-151. SONG Weidong,DU Jianhua,YANG Xingcai,et al. Deformation and failure of a high steep slope due to transformation from deep open-pit to underground mining[J]. Journal of University of Science and Technology Beijing,2010,32(2):145-151.
[6] 郭子钰,刘向峰,王来贵,等. 抚顺西露天矿开挖作用下边坡变形破坏滑动规律研究[J]. 金属矿山,2021(12):190-199. GUO Ziyu,LIU Xiangfeng,WANG Laigui,et al. Study on the slope deformation and failure sliding law under the excavation of Fushun west open-pit mine[J]. Metal Mine,2021(12):190-199.
[7] 邓李政,袁宏永,张鸣之,等. 滑坡变形监测预警技术研究进展[J]. 清华大学学报(自然科学版),2023,63(6):849-864. DENG Lizheng,YUAN Hongyong,ZHANG Mingzhi,et al. Research progress on landslide deformation monitoring and early warning technology[J]. Journal of Tsinghua University (Science and Technology),2023,63(6):849-864.
[8] 叶江,李才艺,高红旗,等. 基于多时相光学遥感影像亚像素相关的边坡监测——以溪洛渡电站为例[J]. 测绘通报,2024(1):38-43,108. YE Jiang,LI Caiyi,GAO Hongqi,et al. Sub-pixel correlation-based slope monitoring using multi-temporal optical remote sensing images:a case study of Xiluodu hydropower station[J]. Bulletin of Surveying and Mapping,2024(1):38-43,108.
[9] 李振洪,宋闯,余琛,等. 卫星雷达遥感在滑坡灾害探测和监测中的应用:挑战与对策[J]. 武汉大学学报(信息科学版),2019,44(7):967-979. LI Zhenhong,SONG Chuang,YU Chen,et al. Application of satellite radar remote sensing to landslide detection and monitoring:challenges and solutions[J]. Geomatics and Information Science of Wuhan University,2019,44(7):967-979.
[10] 李如仁,葛永权,李梦晨,等. 基于InSAR−COMSOL的露天矿边坡稳定性分析及形变预测[J]. 金属矿山,2024(3):172-182. LI Ruren,GE Yongquan,LI Mengchen,et al. Stability analysis and deformation prediction of open-pit mine slopes based on InSAR and COMSOL[J]. Metal Mine,2024(3):172-182.
[11] 万忠明,王亚文,范子义. 无人机倾斜摄影技术在边坡监测中的应用[J]. 测绘通报,2022(6):170-172. WAN Zhongming,WANG Yawen,FAN Ziyi. Application of UAV tilt photography technology in slope monitoring[J]. Bulletin of Surveying and Mapping,2022(6):170-172.
[12] 孙威,朱成峰,张玉婷. 基于GNSS的露天采场边坡监测系统构建[J]. 现代矿业,2022,38(6):213-215,219. SUN Wei,ZHU Chengfeng,ZHANG Yuting. Construction of open-pit stope slope monitoring system based on GNSS[J]. Modern Mining,2022,38(6):213-215,219.
[13] 钟小宇,衣瑛,亢建民,等. 北斗GNSS技术在露天采场边坡监测中的应用[J]. 中国矿山工程,2022,51(1):56-60. ZHONG Xiaoyu,YI Ying,KANG Jianmin,et al. Application of Beidou GNSS technology in open-pit mine slope monitoring[J]. China Mine Engineering,2022,51(1):56-60.
[14] 尹永明,邹江湖,李华汐,等. 南方雨季环境下地基合成孔径雷达在露天矿山边坡监测中的应用[J]. 中国安全生产科学技术,2023,19(增刊1):55-59. YIN Yongming,ZOU Jianghu,LI Huaxi,et al. Application of ground-based synthetic aperture radar for monitoring of open pit mine slopes in southern rainy season environment[J]. Journal of Safety Science and Technology,2023,19(S1):55-59.
[15] 秦宏楠,马海涛,于正兴,等. 地基雷达干涉测量动态高频次数据用于滑坡早期预警方法研究[J/OL]. 武汉大学学报(信息科学版):1-9[2024-04-19]. https://doi.org/10.13203/j.whugis20220152. QIN Hongnan,MA Haitao,YU Zhengxing,et al. Landslide early warning method based on dynamic high frequency data of ground-based radar interferometry[J/OL]. Geomatics and Information Science of Wuhan University:1-9[2024-04-19]. https://doi.org/10.13203/j.whugis20220152.
[16] 王德军,万田宝,孙晓东,等. 基于三维激光扫描的植被覆盖边坡监测[J]. 测绘通报,2023(12):112-115. WANG Dejun,WAN Tianbao,SUN Xiaodong,et al. Monitoring of vegetation covered slopes based on 3D laser[J]. Bulletin of Surveying and Mapping,2023(12):112-115.
[17] 张峰,裴华富. 一种用于滑坡位移监测的OFDR测斜仪研发[J]. 中国测试,2023,49(1):119-125. ZHANG Feng,PEI Huafu. Development of an OFDR inclinometer for landslide displacement monitoring[J]. China Measurement & Test,2023,49(1):119-125.
[18] 许强,董秀军,李为乐. 基于天−空−地一体化的重大地质灾害隐患早期识别与监测预警[J]. 武汉大学学报(信息科学版),2019,44(7):957-966. XU Qiang,DONG Xiujun,LI Weile. Integrated space-air-ground early detection,monitoring and warning system for potential catastrophic geohazards[J]. Geomatics and Information Science of Wuhan University,2019,44(7):957-966.
[19] 许强,朱星,李为乐,等. “天−空−地” 协同滑坡监测技术进展[J]. 测绘学报,2022,51(7):1416-1436. XU Qiang,ZHU Xing,LI Weile,et al. Technical progress of space-air-ground collaborative monitoring of landslide[J]. Acta Geodaetica et Cartographica Sinica,2022,51(7):1416-1436.
[20] 许强,董秀军,朱星,等. 基于实景三维的天−空−地−内滑坡协同观测[J]. 工程地质学报,2023,31(3):706-717. XU Qiang,DONG Xiujun,ZHU Xing,et al. Landslide collaborative observation technology based on real scene 3D view from space-air-ground-interior perspective[J]. Journal of Engineering Geology,2023,31(3):706-717.
[21] 张凯,李全生,戴华阳,等. 矿区地表移动“空天地” 一体化监测技术研究[J]. 煤炭科学技术,2020,48(2):207-213. ZHANG Kai,LI Quansheng,DAI Huayang,et al. Research on integrated monitoring technology and practice of "space-sky-ground" on surface movement in mining area[J]. Coal Science and Technology,2020,48(2):207-213.
[22] 程刚,张昊宇,朱鸿鹄,等. 边坡全维度监测技术与模型试验研究[J]. 高校地质学报,2024,30(2):207-217. CHENG Gang,ZHANG Haoyu,ZHU Honghu,et al. Research of full dimension monitoring technology and model test of slopes[J]. Geological Journal of China Universities,2024,30(2):207-217.
[23] 罗勇,唐华伟,管贵平,等. 基于综合监测技术的大型岩堆边坡防治研究[J]. 中外公路,2016,36(6):9-13. LUO Yong,TANG Huawei,GUAN Guiping,et al. Study on slope prevention and control of large-scale rock piles based on comprehensive monitoring technology[J]. Journal of China & Foreign Highway,2016,36(6):9-13.
[24] 郭延辉,杨溢,高才坤,等. 云南鲁甸地震红石岩堰塞湖右岸特高边坡综合监测及变形特征分析[J]. 中国地质灾害与防治学报,2020,31(6):30-37. GUO Yanhui,YANG Yi,GAO Caikun,et al. Comprehensive monitoring and deformation analysis of extra high slope on the right bank of Hongshiyan Dammed Lake in Ludian Earthquake[J]. The Chinese Journal of Geological Hazard and Control,2020,31(6):30-37.
[25] 马旭峰,王家臣,杨占峰,等. 多源信息融合在露天矿滑体监测中的应用[J]. 有色金属(矿山部分),2008,60(3):32-35. MA Xufeng,WANG Jiachen,YANG Zhanfeng,et al. Application of multi-source information fusion in the landslide monitoring of open pit mine[J]. Nonferrous Metals (Mining Section),2008,60(3):32-35.
[26] PENG M,LI X Y,LI D Q,et al. Slope safety evaluation by integrating multi-source monitoring information[J]. Structural Safety,2014,49:65-74. DOI: 10.1016/j.strusafe.2013.08.007
[27] 刘阳,张建经,李孟芳,等. 基于模糊理论与SVM的边坡地震失稳规模贝叶斯网络评估方法[J]. 岩石力学与工程学报,2019,38(增刊1):2807-2815. LIU Yang,ZHANG Jianjing,LI Mengfang,et al. Fuzzy theory- and SVM-based Bayesian network assessment method for slope seismic instability scale[J]. Chinese Journal of Rock Mechanics and Engineering,2019,38(S1):2807-2815.
[28] 王智伟,王利,黄观文,等. 基于BP神经网络的滑坡监测多源异构数据融合算法研究[J]. 地质力学学报,2020,26(4):575-582. WANG Zhiwei,WANG Li,HUANG Guanwen,et al. Research on multi-source heterogeneous data fusion algorithm of landslide monitoring based on BP neural network[J]. Journal of Geomechanics,2020,26(4):575-582.
[29] 王利,许豪,舒宝,等. 利用互信息和IPSO−LSTM进行滑坡监测多源数据融合[J]. 武汉大学学报(信息科学版),2021,46(10):1478-1488. WANG Li,XU Hao,SHU Bao,et al. A multi-source heterogeneous data fusion method for landslide monitoring with mutual information and IPSO-LSTM neural network[J]. Geomatics and Information Science of Wuhan University,2021,46(10):1478-1488.
[30] 郑海青,赵越磊,宗广昌,等. 融合自注意力机制的Conv−LSTM边坡位移预测方法[J]. 金属矿山,2022(11):193-197. ZHENG Haiqing,ZHAO Yuelei,ZONG Guangchang,et al. Prediction of slope displacement based on conv-LSTM combined with self-attention mechanism[J]. Metal Mine,2022(11):193-197.
[31] LIU Yong,XU Chang,HUANG Biao,et al. Landslide displacement prediction based on multi-source data fusion and sensitivity states[J]. Engineering Geology,2020,271. DOI: 10.1016/j.enggeo.2020.105608.
[32] LIU Chun,SHAO Xiaohang,LI Weiyue. Multi-sensor observation fusion scheme based on 3D variational assimilation for landslide monitoring[J]. Geomatics,Natural Hazards and Risk,2019,10(1):151-167. DOI: 10.1080/19475705.2018.1513871
[33] 鄢好,陈骄锐,李绍红,等. 基于时间序列和GRU的滑坡位移预测[J]. 人民长江,2021,52(1):102-107,133. YAN Hao,CHEN Jiaorui,LI Shaohong,et al. Predicting of landslide displacement based on time series and Gated Recurrent Unit[J]. Yangtze River,2021,52(1):102-107,133.
[34] ZHANG Jie,WANG Zipeng,HU Jinzheng,et al. Bayesian machine learning-based method for prediction of slope failure time[J]. Journal of Rock Mechanics and Geotechnical Engineering,2022,14(4):1188-1199. DOI: 10.1016/j.jrmge.2021.09.010
[35] 刘冠洲,张达,张元生,等. 高陡边坡多源监测预警信息一体化平台研究[J]. 冶金自动化,2018,42(6):1-7. LIU Guanzhou,ZHANG Da,ZHANG Yuansheng,et al. Research on multi-source monitoring and early warning information integration platform for high and steep slopes[J]. Metallurgical Industry Automation,2018,42(6):1-7.
[36] 张凌凡,陈忠辉,周天白,等. 基于梯度提升决策树的露天矿边坡多源信息融合与稳定性预测[J]. 煤炭学报,2020,45(增刊1):173-180. ZHANG Lingfan,CHEN Zhonghui,ZHOU Tianbai,et al. Multi-source information fusion and stability prediction of slope based on gradient boosting decision tree[J]. Journal of China Coal Society,2020,45(S1):173-180.
[37] 刘超云,尹小波,张彬. 基于Kalman滤波数据融合技术的滑坡变形分析与预测[J]. 中国地质灾害与防治学报,2015,26(4):30-35,42. LIU Chaoyun,YIN Xiaobo,ZHANG Bin. Analysis and prediction of landslide deformations based on data fusion technology of Kalman-filter[J]. The Chinese Journal of Geological Hazard and Control,2015,26(4):30-35,42.
[38] 朱自强,吴顺川,刘洋,等. 基于自适应Kalman滤波融合技术的边坡变形分析[J]. 矿业研究与开发,2020,40(1):16-21. ZHU Ziqiang,WU Shunchuan,LIU Yang,et al. Slope deformation analysis based on adaptive Kalman filter fusion technology[J]. Mining Research and Development,2020,40(1):16-21.
[39] 吴泽鑫,张成良,张华超,等. 基于SSA−BP的露天矿山边坡位移变形预测[J]. 有色金属工程,2024,14(6):125-133. WU Zexin,ZHANG Chengliang,ZHANG Huachao,et al. Research on slope displacement prediction of open-pit mine based on SSA-BP[J]. Nonferrous Metals Engineering,2024,14(6):125-133.
[40] 宁永香,崔希民. 矿山边坡地表变形的PSO−ELM预测模型[J]. 煤田地质与勘探,2020,48(6):201-206,216. NING Yongxiang,CUI Ximin. PSO-ELM prediction model for surface deformation of mine slope[J]. Coal Geology & Exploration,2020,48(6):201-206,216.
[41] JIANG Song,LIU Hongsheng,LIAN Minjie,et al. Rock slope displacement prediction based on multi-source information fusion and SSA-DELM model[J]. Frontiers in Environmental Science,2022,10. DOI: 10.3389/fenvs.2022.982069.
[42] 李胜,韩永亮. 基于MFOA−SVR露天矿边坡变形量预测研究[J]. 中国安全生产科学技术,2015,11(1):11-16. LI Sheng,HAN Yongliang. Research on forecasting of slope deformation in open-pit mine based on MFOA-SVR[J]. Journal of Safety Science and Technology,2015,11(1):11-16.
[43] 张研,范聪,吴哲康,等. 基于PSO−RVM的矿山边坡变形量预测模型[J]. 金属矿山,2022(10):191-196. ZHANG Yan,FAN Cong,WU Zhekang,et al. Forecast model of mine slope deformation based on PSO-RVM[J]. Metal Mine,2022(10):191-196.
[44] 金爱兵,张静辉,孙浩,等. 基于SSA−SVM的边坡失稳智能预测及预警模型[J]. 华中科技大学学报(自然科学版),2022,50(11):142-148. JIN Aibing,ZHANG Jinghui,SUN Hao,et al. Intelligent prediction and alert model of slope instability based on SSA-SVM[J]. Journal of Huazhong University of Science and Technology (Natural Science Edition),2022,50(11):142-148.
[45] 罗亦泳,张立亭,鲁铁定,等. 露天矿边坡变形预测的协同进化多核相关向量机模型[J]. 中国安全科学学报,2016,26(11):110-114. LUO Yiyong,ZHANG Liting,LU Tieding,et al. Forecasting of slope deformation in open-pit mine based on CEPSO-MK-RVM[J]. China Safety Science Journal,2016,26(11):110-114.
[46] CHEN Bingqian,YU Hao,ZHANG Xiang,et al. Time-varying surface deformation retrieval and prediction in closed mines through integration of SBAS InSAR measurements and LSTM algorithm[J]. Remote Sensing,2022,14(3). DOI: 10.3390/rs14030788.
[47] HE Yi,YAN Haowen,YANG Wang,et al. Time-series analysis and prediction of surface deformation in the Jinchuan mining area,Gansu Province,by using InSAR and CNN-PhLSTM network[J]. IEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing,2022,15:6732-6751. DOI: 10.1109/JSTARS.2022.3198728
[48] 姜长三,曾桢,万静. 多源信息融合研究进展综述[J]. 现代计算机,2023,29(18):1-9,29. JIANG Changsan,ZENG Zhen,WAN Jing. A review of research advances in multi-source information fusion[J]. Modern Computer,2023,29(18):1-9,29.
[49] 钟国强,徐润,宋杰,等. 基于多态模糊贝叶斯网络的桩锚支护边坡失稳可能性评价[J]. 公路,2020,65(2):50-55. ZHONG Guoqiang,XU Run,SONG Jie,et al. Possibility evaluation of slope instability of pile-anchor support based on polymorphic fuzzy Bayesian network[J]. Highway,2020,65(2):50-55.
[50] AHMAD F,TANG Xiaowei,QIU Jiangnan,et al. Prediction of slope stability using Tree Augmented Naive-Bayes classifier:modeling and performance evaluation[J]. Mathematical Biosciences and Engineering,2022,19(5):4526-4546. DOI: 10.3934/mbe.2022209
[51] 徐卫亚,胡业凡,吴伟伟,等. 基于云模型和D−S证据理论的多源信息融合滑坡安全性评价[J]. 河海大学学报(自然科学版),2022,50(1):59-66. DOI: 10.3876/j.issn.1000-1980.2022.01.009 XU Weiya,HU Yefan,WU Weiwei,et al. Landslide safety evaluation by multi-source information fusion based on cloud model and D-S evidence theory[J]. Journal of Hohai University (Natural Sciences),2022,50(1):59-66. DOI: 10.3876/j.issn.1000-1980.2022.01.009
[52] 张化进,吴顺川,李兵磊. 基于改进D-S证据理论选择性集成的边坡稳定性评价[J/OL]. 金属矿山:1-12[2024-04-19]. http://kns.cnki.net/kcms/detail/34.1055.TD.20230818.1631.002.html. ZHANG Huajin,WU Shunchuan,LI Binglei. Slope stability evaluation based on selective ensemble of improved D-S evidence theory[J/OL]. Metal Mine:1-12[2024-04-19]. http://kns.cnki.net/kcms/detail/34.1055.TD.20230818.1631.002.html.
[53] 李哲,刘彤,刘路路,等. 基于证据理论的支挡型黄土高陡边坡稳定性评价[J]. 东南大学学报(自然科学版),2023,53(3):436-444. LI Zhe,LIU Tong,LIU Lulu,et al. Stability evaluation of supported high and steep loess slope based on D-S evidential reasoning[J]. Journal of Southeast University (Natural Science Edition),2023,53(3):436-444.
[54] 李国辉,刘永,招国栋,等. 基于RS-BPNN理论的边坡稳定性预测及应用[J]. 南华大学学报(自然科学版),2015,29(3):122-128. LI Guohui,LIU Yong,ZHAO Guodong,et al. The prediction and application of slope stability based on RS-BPNN[J]. Journal of University of South China (Science and Technology),2015,29(3):122-128.
[55] ZHANG Duo,FENG Dongmei. Mine geological disaster risk assessment and management based on multisensor information fusion[J]. Mobile Information Systems,2022,2022(1). DOI: 10.1155/2022/1757026.
[56] LIU Zaobao,SHAO Jianfu,XU Weiya,et al. An extreme learning machine approach for slope stability evaluation and prediction[J]. Natural Hazards,2014,73(2):787-804. DOI: 10.1007/s11069-014-1106-7
[57] 杨勇,张忠政,胡军,等. 基于随机权重法改进PSO−ELM的露天矿边坡稳定性分析[J]. 有色金属工程,2022,12(5):128-134. DOI: 10.3969/j.issn.2095-1744.2022.05.016 YANG Yong,ZHANG Zhongzheng,HU Jun,et al. Slope stability analysis of open-pit mine based on improved PSO-ELM with random weight method[J]. Nonferrous Metals Engineering,2022,12(5):128-134. DOI: 10.3969/j.issn.2095-1744.2022.05.016
[58] KANG Fei,XU Qing,LI Junjie. Slope reliability analysis using surrogate models via new support vector machines with swarm intelligence[J]. Applied Mathematical Modelling,2016,40(11/12):6105-6120.
[59] KANG Fei,LI Jingshuang,LI Junjie. System reliability analysis of slopes using least squares support vector machines with particle swarm optimization[J]. Neurocomputing,2016,209:46-56. DOI: 10.1016/j.neucom.2015.11.122
[60] TIEN BUI D,MOAYEDI H,GÖR M,et al. Predicting slope stability failure through machine learning paradigms[J]. ISPRS International Journal of Geo-Information,2019,8(9). DOI: 10.3390/ijgi8090395.
[61] 王健伟,徐玉胜,李俊鑫. 基于网格搜索支持向量机的边坡稳定性系数预测[J]. 铁道建筑,2019,59(5):94-97. DOI: 10.3969/j.issn.1003-1995.2019.05.22 WANG Jianwei,XU Yusheng,LI Junxin. Prediction of slope stability coefficient based on grid search support vector machine[J]. Railway Engineering,2019,59(5):94-97. DOI: 10.3969/j.issn.1003-1995.2019.05.22
[62] ZHANG Wengang,LI Hongrui,HAN Liang,et al. Slope stability prediction using ensemble learning techniques:a case study in Yunyang County,Chongqing,China[J]. Journal of Rock Mechanics and Geotechnical Engineering,2022,14(4):1089-1099. DOI: 10.1016/j.jrmge.2021.12.011
[63] QI Chongchong,TANG Xiaolin. Slope stability prediction using integrated metaheuristic and machine learning approaches:a comparative study[J]. Computers & Industrial Engineering,2018,118:112-122.