Strength prediction model for water-bearing sandstone based on nearinfrared spectroscopy

The strength of water-bearing rock cannot be obtained in real time and by nondestructive experiments,which is an issue at cultural relics protection sites such as grotto temples.To solve this problem,we conducted a near-infrared spectrum acquisition experiment in the field and laboratory uniaxial compression strength tests on sandstone that had different water saturation levels.The correlations between the peak height and peak area of the nearinfrared absorption bands of the water-bearing sandstone and uniaxial compressive strength were analyzed.On this basis,a strength prediction model for water-bearing sandstone was established using the long short-term memory full convolutional network(LSTM-FCN)method.Subsequently,a field engineering test was carried out.The results showed that:(1)The sandstone samples had four distinct characteristic absorption peaks at 1400,1900,2200,and 2325 nm.The peak height and peak area of the absorption bands near 1400 nm and 1900 nm had a negative correlation with uniaxial compressive strength.The peak height and peak area of the absorption bands near 2200 nm and 2325 nm had nonlinear positive correlations with uniaxial compressive strength.(2)The LSTM-FCN method was used to establish a strength prediction model for water-bearing sandstone based on near-infrared spectroscopy,and the model achieved an accuracy of up to 97.52%.(3)The prediction model was used to realize non-destructive,quantitative,and real-time determination of uniaxial compressive strength;this represents a new method for the non-destructive testing of grotto rock mass at sites of cultural relics protection.

Application of Pore Evolution and Fracture Development Coupled Models in the Prediction of Reservoir "Sweet Spots" in Tight Sandstones

The Chang-63 reservoir in the Huaqing area has widely developed tight sandstone ＂thick sand layers, but not reservoirs characterized by rich in oil＂, and it is thus necessary to further study its oil and gas enrichment law. This study builds porosity and fracture development and evolution models in different deposition environments, through core observation, casting thin section, SEM, porosity and permeability analysis, burial history analysis, and ＂four-property-relationships＂ analysis.

Estimating pore volume compressibility by spheroidal pore modeling of digital rocks

The real pores in digital cores were simplified into three abstractive types,including prolate ellipsoids,oblate ellipsoids and spheroids.The three-dimensional spheroidal-pore model of digital core was established based on mesoscopic mechanical theory.The constitutive relationship of different types of pore microstructure deformation was studied with Eshelby equivalent medium theory,and the effects of pore microstructure on pore volume compressibility under elastic deformation conditions of single and multiple pores of a single type and mixed types of pores were investigated.The results showed that the pore volume compressibility coefficient of digital core is closely related with porosity,pore aspect ratio and volumetric proportions of different types of pores.(1)The compressibility coefficient of prolate ellipsoidal pore is positively correlated with the pore aspect ratio,while that of oblate ellipsoidal pore is negatively correlated with the pore aspect ratio.(2)At the same mean value of pore aspect ratio satisfying Gaussian distribution,the more concentrated the range of pore aspect ratio,the higher the compressibility coefficient of both prolate and oblate ellipsoidal pores will be,and the larger the deformation under the same stress condition.(3)The pore compressibility coefficient increases with porosity.(4)At a constant porosity value,the higher the proportion of oblate ellipsoidal and spherical pores in the rock,the more easier for the rock to deform,and the higher the compressibility coefficient of the rock is,while the higher the proportion of prolate ellipsoidal pores in the rock,the more difficult it is for rock to deform,and the lower the compressibility coefficient of the rock is.By calculating pore compressibility coefficient of ten classical digital rock samples,the presented analytical elliptical-pore model based on real pore structure of digital rocks can be applied to calculation of pore volume compressibility coefficient of digital rock sample.

Modeling Tracer Flow Characteristics in Different Types of Pores: Visualization and Mathematical Modeling

Structure of porous media and fluid distribution in rocks can significantly affect the transport characteristics during the process of microscale tracer flow.To clarify the effect of micro heterogeneity on aqueous tracer transport,this paper demonstrates microscopic experiments at pore level and proposes an improved mathematical model for tracer transport.The visualization results show a faster tracer movement into movable water than it into bound water,and quicker occupancy in flowing pores than in storage pores caused by the difference of tracer velocity.Moreover,the proposed mathematical model includes the effects of bound water and flowing porosity by applying interstitial flow velocity expression.The new model also distinguishes flowing and storage pores,accounting for different tracer transport mechanisms(dispersion,diffusion and adsorption)in different types of pores.The resulting analytical solution better matches with tracer production data than the standard model.The residual sum of squares(RSS)from the new model is 0.0005,which is 100 times smaller than the RSS from the standard model.The sensitivity analysis indicates that the dispersion coefficient and flowing porosity shows a negative correlation with the tracer breakthrough time and the increasing slope,whereas the superficial velocity and bound water saturation show a positive correlation.

Prediction Model of Compressive Strength of Fly Ash-Slag Concrete Based on Multiple Adaptive Regression Splines

Accurate prediction of compressive strength of concrete is one of the key issues in the concrete industry. In this paper, a prediction method of fly ash-slag concrete compressive strength based on multiple adaptive regression splines (MARS) is proposed, and the model analysis process is determined by analyzing the principle of this algorithm. Based on the Concrete Compressive Strength dataset of UCI, the MARS model for compressive strength prediction was constructed with cement content, blast furnace slag powder content, fly ash content, water content, reducing agent content, coarse aggregate content, fine aggregate content and age as independent variables. The prediction results of artificial neural network (BP), random forest (RF), support vector machine (SVM), extreme learning machine (ELM), and multiple nonlinear regression (MnLR) were compared and analyzed, and the prediction accuracy and model stability of MARS and RF models had obvious advantages, and the comprehensive performance of MARS model was slightly better than that of RF model. Finally, the explicit expression of the MARS model for compressive strength is given, which provides an effective method to achieve the prediction of compressive strength of fly ash-slag concrete.

基于拌和生产数据的BP神经网络混凝土抗压强度预测

为解决混凝土生产中抗压强度试验周期长及工程管理存在滞后性的问题,提出了一种基于混凝土拌和生产实时监控数据的BP神经网络混凝土抗压强度预测模型。以混凝土拌和生产中的8项物料生产称重数据和5项生产配比数据作为预测输入变量,建立200组混凝土拌和站生产监控数据和对应的抗压强度试验数据样本集,按照6∶2∶2比例划分为训练集、验证集和测试集;分别以C40配比混凝土拌和生产的8项物料称重数据和全部13项数据作为输入变量,进行混凝土28 d抗压强度预测,将预测结果与实际试验结果进行比较,验证所提出BP神经网络模型的预测效果。结果表明:所提出的BP神经网络混凝土强度预测模型能较好地实时预测混凝土28 d抗压强度,且相对误差优于利用7 d抗压强度试验数据估算值;8项物料称重数据作为输入变量的BP神经网络预测模型预测精度更好,平均绝对百分比误差为0.82%,均方根误差为0.52 MPa;利用不同拌和站C20配比、C30配比混凝土拌和生产监控数据对8项输入变量BP神经网络混凝土抗压强度预测模型进行适应性验证可知,其预测平均绝对误差均在0.5 MPa之内,平均绝对百分比误差均小于2%,与C40配比预测误差一致;该预测模型充分挖掘了混凝土拌和站生产实时监控数据的价值,实现了传统混凝土抗压试验结果提前化,对提高工程建设质量水平具有重要意义。

基于GWO-BP神经网络及粮食压缩实验对粮食孔隙率的预测

孔隙率是影响粮堆内热湿传递的关键参数,为探究粮仓中散装粮堆孔隙率的分布规律,通过开展粮食压缩实验来获取不同的粮食种类在不同水分含量和竖向压力条件下的孔隙率。提出了一种基于灰狼算法优化BP(GWO-BP)神经网络的粮食单元体孔隙率预测模型,并将该模型与BP神经网络模型、随机森林模型的孔隙率预测结果进行对比,最后利用粮食单元箱实验对该模型的泛化能力进行验证。结果表明,GWO-BP神经网络模型的孔隙率预测性能最佳,该模型的评价指标R2为0.960 5、RMSE为0.013 7及MAE为0.0131,均在允许的范围内。本研究为粮食孔隙率的确定提供了一种神经网络预测的方法,为深入开展粮堆多场耦合分析提供了重要基础,为安全储粮提供了理论支持。

喷射混杂纤维混凝土抗冻性及损伤模型

目的为提高高寒地区喷射混凝土耐久性,改善冻融环境下普通喷射混凝土易开裂、抗冻性能差等问题。方法在喷射混凝土中掺入混合纤维,通过冻融循环试验、超声波检测、压拉性能试验以及气孔结构测试等方法,研究不同冻融循环次数下喷射混杂纤维混凝土的相对纵波波速、相对压拉强度以及试件内部气孔结构变化规律。通过压拉性能损伤量对试件冻融损伤进行多项评价,分析钢纤维和聚丙烯纤维贡献率与喷射混凝土抗压强度之间的关系,并根据钢纤维和聚丙烯纤维贡献率公式建立喷射混凝土冻融损伤模型。结果与单掺钢纤维和聚丙烯纤维试件相比,混合掺入1%钢纤维、0.1%聚丙烯纤维喷射混凝土试件的抗压强度损失分别降低了7.5%、33.3%,劈裂抗拉强度损失分别降低了15.6%、35.0%。在弦长尺寸0.1mm<d<1mm处,混合掺入0.1%聚丙烯和1%钢纤维喷射混凝土气泡弦长频率比单掺1%钢纤维喷射混凝土和单掺0.1%聚丙烯喷射混凝土分别降低了24.9%和27.3%。钢纤维和聚丙烯纤维对抗压强度的贡献率呈现先增长后下降趋势。基于冻融损伤与钢纤维和聚丙烯纤维贡献率建立冻融损伤模型。结论喷射混杂纤维混凝土力学性能损伤量低,不易开裂,抗冻性能好,可为高寒地区喷射混凝土结构提供理论参考。

剑麻纤维加筋水泥土力学性能试验

目的 研究纤维掺量、纤维长度及养护龄期对剑麻纤维加筋水泥土无侧限抗压强度及巴西劈裂抗拉强度的影响规律。方法 设置纤维掺量为0~0.8%,加筋长度为7~19 mm,养护龄期为7 d、28 d及90 d,对不同影响因素下剑麻纤维加筋水泥土进行无侧限抗压强度试验及巴西劈裂试验。结果 剑麻纤维的掺入提升了水泥土的无侧限抗压强度及巴西劈裂抗拉强度,随着纤维掺量及加筋长度的增加呈现出先增大后减小的规律;当加筋长度为11 mm、纤维掺量为0.4%时,加筋效果最显著,无侧限抗压强度提高幅度达30.2%;剑麻纤维加筋水泥土的抗拉强度、拉压比与纤维掺量呈线性关系。结论 剑麻纤维的掺入改善了水泥土的脆性,提高了水泥土的破坏韧性;采用幂函数对试验数据进行拟合,得到了养护龄期与纤维掺量共同作用下剑麻纤维加筋水泥土无侧限抗压强度的预测模型,可以为实际工程提供参考。

碳酸盐岩地层超压预测模型研究及应用

碳酸盐岩地层超压预测是目前国内外尚未解决的基础性难题。以川东北地区超压碳酸盐岩地层为研究对象,通过物理模拟实验研究碳酸盐岩弹性参数随围压和孔压的变化规律,提出了基于岩石孔隙压缩系数的碳酸盐岩地层超压预测模型。实验结果显示,碳酸盐岩孔隙度的增大和岩样弹性参数变化与孔隙压力增量之间存在一定的相关性,据此建立岩石孔隙压缩系数与孔隙压力、孔隙体积的理论关系,并推导出超压预测模型,模型考虑了区域构造应力的影响,参数可以通过实验和测井资料获得;利用密度、声波和中子测井资料开展典型钻井超压预测,预测结果与实测压力吻合良好。新提出的超压预测模型为碳酸盐岩地层超压预测提供了新思路。

老年骨质疏松性胸腰椎压缩骨折患者术后残余腰背痛影响因素及预测模型的构建分析

目的分析老年骨质疏松性胸腰椎压缩骨折患者术后残余腰背痛的影响因素,并构建预测模型。方法选取2019年9月~2022年9月术后残余腰背痛、未发生残余腰背痛的老年骨质疏松性胸腰椎压缩骨折患者各85例(前者为研究组,后者为对照组),收集两组临床资料,采用单因素、多因素Logistic回归分析对术后残余腰背痛的影响因素进行分析,构建预测模型,绘制受试者工作特征曲线(receiver operating characteristic,ROC)分析预测效能。结果合并筋膜损伤、骨密度T值低、骨水泥分布(O型)、短期并发症为患者术后残余腰背痛的独立危险因素,而实施早期康复干预、康复干预依从性良好为保护因素(P<0.05)。构建风险预测模型:Log(P)=2.084×合并筋膜损伤+2.187×骨密度T值+1.793×骨水泥分布+1.796×短期并发症-3.099×实施早期康复干预-3.734×康复干预依从性-27.052,该模型预测老年骨质疏松性胸腰椎压缩骨折术后残余腰背痛的ROC曲线下面积为0.888(95%CI:0.837~0.938,P<0.05);预测模型阈值=16.98时的敏感度、特异度最佳,分别为0.867、0.864;该模型的Hosmer-Lemeshow检验无差异,有较好的符合程度。结论合并筋膜损伤、骨密度T值低、骨水泥分布(O型)、短期并发症,是影响老年骨质疏松性胸腰椎压缩骨折患者术后残余腰背痛的独立危险因素,实施早期康复干预、康复干预依从性良好为保护因素;可据此建立模型进行预测,并开展相应干预措施,减轻患者疼痛。

LOLIMOT模型在CNG发动机NO_(x)排放预测试验中的应用

为解决在选择性催化还原技术(selective catalytic reduction,SCR)的控制策略开发中局部线性模型树(local linear model tree,LOLIMOT)排放模型预测精度不足的问题,提出一种通过优化空间边界,将原模型的超矩形输入空间约束在物理意义范围内的改进LOLIMOT模型。通过某天然气发动机的辨识试验,从分布特征和计算原理角度,分析了该方法对预测结果的影响。结果表明:与原算法相比,改进算法的线性相关度R2提升了1.9%,验证了改进策略的有效性。改进LOLIMOT算法具备较高的收敛速度和稳定性,在排放模型领域具备一定的应用优势。

细观初始孔隙缺陷对混凝土力学性能的影响

采用不同尺寸的EPS颗粒预制混凝土的初始孔隙率及初始孔径尺寸,系统开展4种初始孔隙率和5种初始孔径尺寸下混凝土的轴心抗压强度和弹性模量试验研究,并基于数值统计原理分析含初始缺陷混凝土力学性能的影响规律.结果表明:混凝土的轴心抗压强度和弹性模量降幅随初始孔隙率的增加呈近似线性上升趋势,混凝土的轴心抗压强度降幅随着初始孔径尺寸的增加呈近似对数上升趋势.同时考虑初始孔隙率和初始孔径尺寸的影响,混凝土在初始小孔径、高孔隙率下的损伤度低于其在初始大孔径、低孔隙率下的损伤度.相较初始孔径尺寸,初始孔隙率对混凝土轴心抗压强度和弹性模量的损伤更为显著.

P型橡胶卡箍的拉压疲劳性能测试和寿命预测

P型橡胶卡箍是航空发动机外部管路系统中连接管路与支架、机匣之间的重要连接件,其主要的失效形式为机体振动和管内高压液体脉动共同作用下的疲劳断裂。卡箍的疲劳性能极大地制约了管路系统的性能,因此对于卡箍的疲劳性能研究具有重要意义。本文首先设计并进行了卡箍的拉压疲劳试验,得到了不同加载位移下的疲劳寿命。后续的断口分析表明卡箍疲劳破坏呈现出两种不同的形式:一是卡箍的金属箍带发生由外侧向内侧扩展的疲劳断裂,二是卡箍的橡胶垫圈严重的磨损现象。其次,结合卡箍疲劳试验的加载过程,建立了对应的有限元模型,并通过与贴片位置的应变对比验证了构建模型的预测精度。针对卡箍箍带疲劳断裂的失效形式,通过卡箍箍带所采用的不锈钢材料的疲劳试验结果,建立了卡箍的SWT(Smith-Watson-Topper)、FS(Fatemi-Socie)以及WB(Wang-Brown)临界平面疲劳寿命模型。最后,结合有限元分析得到的最大应力应变以及卡箍拉压疲劳试验得到的疲劳寿命结果,验证了提出的寿命模型均处于3倍分散带内,对于卡箍的疲劳寿命具有良好的估计精度。

老年骨质疏松性胸腰椎压缩性骨折患者PKP术后再发骨折的风险分析及列线图预测模型的构建

目的探讨老年骨质疏松性胸腰椎压缩性骨折患者经皮椎体后凸成形术(PKP)后再发骨折的风险因素并构建列线图预测模型。方法选取2016年1月~2019年11月行PKP治疗的182例老年骨质疏松性胸腰椎压缩性骨折患者为研究对象,术后3年持续跟踪随访,根据术后恢复情况将患者分为再发骨折组(n=36)与无再发骨折组(n=146)。收集两组患者的临床资料;对计量指标行接受者操作特性(ROC)曲线分析;经Logistic回归分析影响PKP术后再发骨折的独立危险因素;R语言软件4.0“rms”包构建列线图预测模型,校正及决策曲线对列线图预测模型进行内部验证及临床预测效能评估。结果两组患者在骨密度、损伤椎体数、单节骨水泥注入量、骨水泥分布类型、骨水泥渗漏、PKP术前后椎体高度差、后凸角变化方面的差异具有统计学意义(P<0.05)。骨密度、损伤椎体数、单节骨水泥注入量、PKP术前和术后椎体高度差、后凸角变化的曲线下面积(AUC)分别为0.772、0.732、0.722、0.801、0.813,最佳截断值分别为-3.1、2个、3.9 ml、0.4 mm、8.7°。骨密度、损伤椎体数、单节骨水泥注入量、骨水泥渗漏、PKP术前后椎体高度差、后凸角变化是影响老年骨质疏松性胸腰椎压缩性骨折患者PKP术后再发骨折的独立危险因素。列线图预测模型的校正曲线与原始曲线及理想曲线接近,C-index为0.818(95%CI:0.762~0.883),模型拟合度高;列线图预测模型的阈值>0.22,可提供临床净收益,且临床净收益均高于独立预测因子。结论骨密度、损伤椎体数、单节骨水泥注入量、骨水泥渗漏、PKP术前和术后椎体高度差、后凸角变化是影响老年骨质疏松性胸腰椎压缩性骨折患者PKP术后再发骨折的独立危险因素,并构建了预测老年骨质疏松性胸腰椎压缩性骨折患者PKP术后再发骨折的列线图模型。

建立骨质疏松性椎体压缩骨折术后再骨折预测模型

目的探讨骨质疏松性椎体压缩骨折(osteoporotic vertebral compression fracture,OVCF)术后再骨折风险,构建风险预测模型,确定有效防治措施。方法选取2021年8月至2022年6月北京积水潭医院收治的119例OVCF患者作为研究对象,根据术后再骨折与否分为再发组和非再发组,其中再发组22例,男11例,女11例;年龄55~86岁,平均(72.02±5.58)岁。非再发组97例,男50例,女47例;年龄55~86岁,平均(70.79±6.81)岁。统计两组一般资料,采用Lasso-Logistic回归模型筛选OVCF术后再发骨折自变量,采用赤池信息准则(Akaike’s information criterion,AIC)、贝叶斯信息准则(Bayesian information criterion,BIC)比较全变量Logistic回归、逐步Logistic、Lasso-Logistic回归预测效能,构建诺莫图模型,采用受试者工作特征曲线(receiver operating characteristic,ROC)、校准曲线分析OVCF术后再发骨折诺莫图模型效能。结果术后随访8~20个月,平均(12.00±2.40)个月。单因素分析显示,再发组身体质量指数(body mass index,BMI)、骨密度T值、抗酒石酸酸性磷酸酶(tartrate-resistant acid phosphatase 5b,TPACP-5b)、核因子kB受体激活因子配体(receptor acti-vator nuclear factor kappa B ligand,RANKL)、骨保护素(osteoprotegrin,OPG)、术后抗骨质疏松治疗、白细胞介素(interleukin,IL)-17、长期糖皮质激素使用史、脊柱畸形指数(spinal deformity index,SDI)值、手术段Cobb角、后凸角度与非再发组比较,差异有统计学意义(P<0.05);Lasso-Logistic回归模型分析显示lambda.lse值0.049为最优模型,此时进入模型的变量涉及骨密度、SDI值、IL-17、术后抗骨质疏松治疗,经BIC、AIC验证表明所构建模型拟合和预测效果相对较好;诺莫图模型的ROC下面积(area under the curve,AUC)为0.865,敏感度及特异度分别为95.45%、68.04%,且校准曲线显示,其预测效能与实际吻合较好。结论OVCF术后再骨折的发生受围手术期多方面影响,涉及骨密度T值、SDI值、IL-17、术后抗骨质疏松治疗,基于以上因素可有效预测患者再骨折风险,为临床防治再骨折提供参考依据。

基于超声横波测试的Na基膨润土充填体强度预测模型研究

充填体作为充填采矿法的核心单元,其强度是保障安全开采的重要指标。Na基膨润土的亚甲基蓝吸附量和生坯抗压强度较高,是制备充填体的优质添加剂,然而鲜有其掺量与充填体强度之间关系的研究成果。本文基于超声波测试技术和单轴压缩试验,分析不同Na基膨润土掺量充填体的横波波速、主频幅值、幅值衰减系数、波形分形维数在各龄期的变化规律,结合敏感性分析遴选出对抗压强度变化最敏感的声学参数。建立了不同Na基膨润土掺量下充填体强度的预测模型,并结合显著性检验及对比分析,对强度预测模型进行了验证。研究成果可为开展充填体单轴抗压强度预测的相关理论研究和工程应用提供参考和帮助。

基于BP神经网络的固化红土抗压强度预测

为分析不同掺量的偏高岭土与石灰共同掺入玄武岩残积红土中对土体的改良效果,本试验选取偏高岭土的掺量分别为0%、2%、4%、6%和8%,石灰的掺量分别为0%、2.5%、5.0%、7.5%和10.0%,同时掺入玄武岩残积红土中,制作25组不同固化红土,对其进行28 d无侧限抗压强度正交试验,并用MATLAB软件建立神经网络预测模型,预测固化红土养护28 d的抗压强度。研究结果表明:本模型预测误差最大为4.56%,拟合度为0.997,且本方法比常规回归分析法更简单、更准确,可预测不同固结材料和掺量的固化红土抗压强度,提高试验效率。

考虑软矿物纵横比的页岩岩石物理建模及其应用

以往针对页岩储层的岩石物理建模常忽略孔隙类型与软矿物纵横比对弹性模量的影响。本文同时考虑孔隙类型、孔隙形状与软矿物纵横比,建立横向各向同性页岩岩石物理模型:将固体矿物分为硬矿物与软矿物两类,考虑软矿物的各向异性特征与形状变化;基于储层实际情况将孔隙分为粒内孔、有机孔与粒间孔三类,孔隙形状分为硬孔隙与软孔隙两种;最后,利用粒子群法反演输入参数,进一步计算纵横波速度、各向异性参数与岩石力学参数。结合实际资料应用,用已知测井横波速度与各向同性岩石力学计算结果进行对比,结果表明该模型的应用效果较好。

透水混凝土性能的影响因素研究

随着城市地表硬化程度越来越高,城市内涝和排水系统超载问题日益严重。国家大力建设海绵城市,透水混凝土作为用量最大的透水铺装材料发挥着重要作用。通过对大量文献调研和统计分析,归纳总结了透水混凝土的水灰比、骨料级配、骨料类型及粉煤灰掺合料对其抗压强度和透水性能的影响,并阐述其作用机理。在此基础上,对其进行回归分析,建立透水混凝土多因素下的抗压强度和透水性能的预测模型,为透水混凝土配合比设计和材料选择等方面提供参考。