img

QQ群聊

img

官方微信

  • CN 62-1112/TF 
  • ISSN 1005-2518 
  • 创刊于1988年
高级检索

黄金科学技术, 2019, 27(5): 740-746 doi: 10.11872/j.issn.1005-2518.2019.05.740

采选技术与矿山管理

基于改进PCA与有序多分类Logistic的充填管道磨损风险评估

王石,, 汤艺,, 冯萧

江西理工大学资源与环境工程学院,江西 赣州 341000

Risk Assessment of Filling Pipeline Wearing Based on Improved PCA and Ordered Multi-class Logistic

WANG Shi,, TANG Yi,, FENG Xiao

School of Resources and Environmental Engineering,Ganzhou 341000,Jiangxi,China

通讯作者: 汤艺(1996-),男,江西南昌人,本科生,从事充填技术研究工作。tylwtfglsm@163.com

收稿日期: 2018-08-14   修回日期: 2018-12-11   网络出版日期: 2019-10-29

基金资助: 国家自然科学基金项目“APAM强化絮网结构后全尾砂料浆流动性能演化机制研究”.  51804134
国家自然科学基金项目“渗流—蠕变耦合作用下全尾砂胶结充填体力学性能演化规律及损伤破坏机制”.  51804135
江西省自然科学基金项目“远距离输送过程中添加絮凝剂的高浓度全尾砂浆颗粒分散机理研究”.  20181BAB216013
江西理工大学博士启动基金项目“阴离子型聚丙烯酰胺对似膏体管道输送稳定性的影响机理研究工作”.  jxxjbs17011

Received: 2018-08-14   Revised: 2018-12-11   Online: 2019-10-29

作者简介 About authors

王石(1987-),男,河南济源人,讲师,从事采矿工艺与充填技术研究工作stonersxx@126.com , E-mail:stonersxx@126.com

摘要

为准确预测矿山充填管道磨损风险,建立改进PCA与有序多分类Logistic回归组合的充填管道磨损风险评估模型。结合实际经验,选取12项指标(9项定量指标和3项定性指标)建立评估模型。依据改进PCA算法,筛除影响力指数小的充填料浆密度和充填料浆腐蚀性2项指标,将优选出的主要指标代入有序多分类Logistic回归模型,依照相应概率大小进行风险性等级判定,最后预测矿山充填管道磨损风险等级概率。该方法摒弃了关联性较低的指标,得到可靠的充填管道磨损风险概率分布,为类似矿山科学预测管道磨损风险及采取有效防护措施提供了理论依据。

关键词: 充填管道磨损 ; 改进PCA ; 有序多分类Logistic ; 磨损风险 ; 风险等级概率

Abstract

The filling mining method is mainstream mining method used in major mines today.The safe implementation of filling technology depends on the construction of a good filling pipeline transportation system.Considering the complexity of the filled pipeline system and the large number and variety of influencing factors,in order to accurately predict the wear risk of the filling pipeline of the mine,a wear risk assessment of the filling pipeline based on the improved PCA and the ordered multi-class Logistic regression combination model was built.On the basis of practical experience,a total of 12 items including the volume fraction of the filling slurry,the filling doubled line,the corrosiveness of the filling slurry,and the material of the pipe were selected.Reasonable risk levels were divided according to the characteristics of each indicator and the corresponding evaluation model was constructed.The filling production data of four mines such as Jinchuan Longshou mine and Dahongshan copper mine were taken as samples.The improved PCA algorithm was used to analyze the correlation of each index,and the three principal components with a cumulative contribution rate of 91.125% and factor load of the principal component of each indicator were obtained.In the end,the two indicators with low correlation and small influence index,the density of the filling slurry(27.75) and the corrosiveness of the filling slurry(27.60) were deleted.The preferred main indicators were substituted into the ordered multi-class Logistic regression model,the continuous indicator values were linearly fitted to the discrete risk levels,and the regression coefficients,standard errors and significance levels of each index were calculated,and the equations of probability fluctuations were solved. Finally,the probability of four mines corresponding to different wear risks Ⅰ(not easy to wear),Ⅱ(relatively easy to wear),Ⅲ(easy to wear),Ⅳ(very easy to wear) were obtained. They were: Jinchuan Longshou mine is 0.247,0.440,0.154,0.153; Dahongshan copper mine is 0.179,0.240,0.323,0.258; Hedong gold mine is 0.181,0.227,0.345,0.247; Xincheng gold mine is 0.181,0.227,0.345,0.247.From a theoretical point of view,the risk level corresponding to the probability of the largest value was used as the final judgment level of the wear risk of the mine filling pipeline,and Jinchuan Longshou mine is Ⅱ,Dahongshan copper mine is Ⅲ,Hedong gold mine is Ⅲ,Xincheng gold mine is Ⅲ. The mine also makes corresponding level protection and maintenance measures based on this.From the actual production application,it is said that the probability of the Ⅳ risk level should be paid special attention,and the normal operation of the filling pipeline should be guaranteed to the greatest extent within the scope permitted by technology and funds. The mathematical method combining improved PCA and ordered multi-class Logistic regression avoids the collinearity between the various indicators,reduces the interference of the weak indicators on the evaluation results,and obtains the accurate risk of wear risk of filling the pipeline.It provides a theoretical basis for scientific prediction of pipeline wear risk and the implementation of effective and economical protection measures for similar mines.The mine can construct an appropriate filling pipeline protection system according to its own development.

Keywords: filling pipeline wearing ; improve principal component analysis ; ordered multi-class Logistic regression ; wear risk ; probability of risk level

PDF (617KB) 元数据 多维度评价 相关文章 导出 EndNote| Ris| Bibtex  收藏本文

本文引用格式

王石, 汤艺, 冯萧. 基于改进PCA与有序多分类Logistic的充填管道磨损风险评估[J]. 黄金科学技术, 2019, 27(5): 740-746 doi:10.11872/j.issn.1005-2518.2019.05.740

WANG Shi, TANG Yi, FENG Xiao. Risk Assessment of Filling Pipeline Wearing Based on Improved PCA and Ordered Multi-class Logistic[J]. Gold Science and Technology, 2019, 27(5): 740-746 doi:10.11872/j.issn.1005-2518.2019.05.740

随着矿山机械化的发展,矿山的开采深度也在逐步增加,为了更好地控制深度增加带来的高地压和高地热等问题,近年来大部分矿山选择的主体采矿方法均为充填采矿法[1,2,3]。其中,充填系统的合理构建与稳定运行是充填采矿法成功应用的关键。然而作为充填系统中重要一环的充填管道输送系统却不断出现磨损、堵塞和爆裂等失效事故,不仅给矿山带来了巨大的经济损失[4,5,6],而且严重威胁着工作人员的人身安全。其中,充填管道磨损作为最常见的充填管道失效事故,缺乏科学有效的预测评估体系,亟需采取措施进行风险预测。

为此,国内外学者通过数值分析[7,8]、动量方程推导[9,10,11]、构建组合权重和可变模糊耦合模型[12]等多个角度,从理论层面阐述管道磨损的形式及其成因,定性分析各项指标在充填管道磨损中所起的作用,并对矿山管道磨损风险进行等级评估。前人已经做了大量基础研究,但有关实际生产中存在的、诸多人为或自然的不确定性因素如何干扰充填管道磨损风险评估的研究鲜见报道。本文从风险评估等级概率分布的角度对该问题进行了深入探究。

在实际操作中,即使最优的充填管道输送参数,也依然有可能因为现实存在的未知因素,导致充填管道出现重大磨损事件[13,14]。因此,充填管道磨损风险与其影响指标之间的关系并不是一个绝对的静态平衡等式,而应是受到多种不确定性因素干扰的概率波动方程。鉴于此,综合考虑多个影响指标,建立改进PCA与有序多分类Logistic相结合的评估模型。根据改进PCA算法,筛除了影响指数较小的指标,突出了具有代表性的指标在管道磨损中的重要作用。将剩余指标代入有序多分类Logistic回归模型,得到充填管道磨损风险等级及其相应概率,展现充填管道磨损风险与其各指标之间的动态映射关系,建立了科学、系统且准确的管道磨损风险评估体系。

1 改进PCA与有序多分类Logistic回归原理

改进PCA[15,16]就是设法将数量众多、具有一定相关性的指标重新组合成几个相互无关的综合指标,并且尽可能多地反映原来的信息,再结合影响力指数大小,筛除关联性较小的指标,突出具有代表性的指标。

1.1 改进PCA原理

设有P项指标影响研究对象,以X1X2,…,XP表示,并以这P项指标的观测数据构成了P维随机向量X=(X1X2,…,XP)。同时对各个指标的数据进行标准化,消除各数据在量纲和数量等级上的差异。依据处理后的数据构建协方差矩阵R,它是反映各个指标数据之间相关性程度的统计指标,其中Rij ij=1,2,…,p)为指标XiXj的相关系数,其计算公式为

Rij=k=1n(Xkj-Xi)(Xkj-Xk)k=1n(Xkj-Xi)2(Xkj-Xj)2

再以此为基准,计算特征根λi及其相应特征向量ei,并求得各个主成分的贡献率Gi和因子载荷aij

Gi=λik=1pλk
aij=λieij(i,j=1,2,,p)

最后得出各个主成分Zi的表达式:

Z1=a11X1+a12X2+...+a1pXpZ2=a21X1+a22X2+...+a2pXpZm=am1X1+am2X2+...+ampXp

在此基础上,构建影响力数值Bj来表示每个指标的综合影响力,通过计算各指标Xjj=1,2,…,p)的影响力数值大小,筛除影响力数值小、与研究对象关联性较弱的指标。其计算公式为

Bj=i=1mGi×|aij|

1.2 有序多分类Logistic回归数学模型

已知研究变量yJ个有序离散取值水平,记为y=1,2,...,J。假定连续变量f与影响因素x存在线性关系[17,18,19]

fi=β0+β1x1+β2x2++βmxm+εi   =β0+k=1mβjxj+εi, i=1,2,,n

式中:n表示观测数据组个数;εi(i=1,2,,n)为随机变量。继续假设,连续变量f的变化表示研究变量y取值状态,即存在α1,α2,,αn使得

fα1:y=1:α1<fα2:y=2::f>αJ-1,y=J

则,yj的概率,即pj(yj)应满足

lnpj(yj)1-pj(yj)=lnl=1jPll=j+1JPl=αj-(β0+k=1mβkxik)

将式(7)变形得到

pj(yj)=eαj-(β0+k=1mβkxik)1+eαj-(β0+k=1mβkxik)

最后得到y=j时的概率为

pj(y=j)=pj(yj)-pj(yj-1)=eαj-(β0+k=1mβkxik)1+eαj-(β0+k=1mβkxik)-eαj-1-(β0+k=1mβkxik)1+eαj-1-(β0+k=1mβkxik)

2 充填管道磨损风险评估指标体系

充填管道磨损风险评估是一项庞大复杂的系统工程,科学建立评估指标体系以及正确选择相应评估指标是实现充填管道磨损风险合理评估的根本。有关充填管道的磨损研究,涉及的影响因素众多,但主要还是与充填颗粒性质、充填料浆流速、充填压力和充填管道自身参数等相关。例如,充填骨料形状及骨料中粗颗粒所占比重,骨料形状越不规则,粗颗粒占比越大,料浆流动时对管道产生的摩擦损坏更大。其原因是充填料浆流速快,动量大,对管道产生的冲击力强,导致管道内壁碎屑材料脱落且混入料浆中,增大了管道内壁的不平整度,提高了料浆颗粒与管道内壁的碰撞概率,进一步加剧了管道磨损。综合考虑各类因素的作用机理及涉及方面,最终整理出12个具有代表性的指标。

由于在实际生产中部分指标并没有准确的量化数据,所以结合相关文献资料和专家意见,对各定性指标进行合理赋值,使评价结果更为准确。参考分级标准法将每个指标划分为4个等级,评估等级集为{D1D2D3D4},即划分为Ⅰ级、Ⅱ级、Ⅲ级和Ⅳ级,分别表示管道不易磨损、较易磨损、容易磨损和极易磨损,分级结果见表1~表2。12个评价指标分别为:充填料浆体积分数(I1)、充填料浆密度(I2)、粗颗粒占比(I3)、管道内径(I4)、管道铺设不平整度(I5)、管道绝对粗糙程度(I6)、充填倍线(I7)、料浆流速与临界流速之比(I8)、管道使用年限(I9)、充填骨料形状(I10)、充填料浆腐蚀性(I11)和管道材料(I12)。其中I1~I9为定量指标,I10~I12为定性指标。

表1   定量指标评估等级取值范围

Table 1  Range of values for quantitative indicator evaluation level

磨损等级充填料浆体积分数/%充填料浆密度/(t·m-³)粗颗粒占比/%管道内径/mm管道铺设不平整度/%管道绝对糙度/μm充填倍线料浆流速与临界流速之比管道使用年限/a
I1<30I2<1.5I3<20I4>200I5<1.0I6≤100I7≥7.0I8<1.0I10<2
30≤I1<401.5≤I2<1.720≤I3<40150< I4≤2001.0≤I5<3.0I6≤1005.0≤I7<7.01.0≤I8<1.22≤I10<5
40≤I1<501.7≤I2<1.940≤I3<60100< I4≤1503.0≤I5<5.0300≤I6<5003.0≤I7<5.01.2≤I8<1.55≤I10<10
I1≥50I2≥1.9I3≥60I4≤100I5≥5.0I6≥5001.0≤I7<3.0I8≥1.5I10≥10

新窗口打开| 下载CSV


表2   定性指标评估等级取值范围

Table 2  Range of values for qualitative indicator evaluation level

磨损等级充填骨料形状充填料浆腐蚀性管道材料
圆形或椭圆形中性并且不含有能与充填管道发生反应的成分双金属耐磨复合管
方形特定条件下pH值发生变化与充填管道发生反应钢塑复合管
多棱角形弱酸、弱碱以及一系列能与充填管道发生轻微反应的成分陶瓷复合管
极不规则强酸、强碱或一系列容易与充填管道发生反应的成分单一复合材料

新窗口打开| 下载CSV


3 工程实例

以具体矿山为例,采用改进PCA与有序多分类Logistic回归相结合的方法,对国内使用充填采矿法的4个典型矿山充填管道磨损风险进行预测,由文献[1,12,20]和矿山实际生产资料得到4个矿山关于这12项指标的具体数值,详细数据见表3

表3   矿山充填管道磨损影响指标调查数据

Table 3  Survey data on the impact indicators of mine filling pipelines

矿山名称I1I2I3I4I5I6I7I8I9I10I11I12
金川龙首矿561.98381992.723003.21.38122
大红山铜矿331.69421600.985009.63.07332
河东金矿241.6852820.561005.21.610323
新城金矿521.94331071.272005.83.56431

新窗口打开| 下载CSV


3.1 数据主成分分析

运用SPSS对表3中的数据进行PCA分析处理,得到各个成分的相应贡献率Gi。取其中贡献率占比最大且累计贡献率达91.125%的3个成分Z1Z2Z3作为该组数据的主成分,得到各主成分的因子aij载荷,再计算得到各个指标的影响力指数Bi,见表4

表4   主成分因子载荷及影响力指数

Table 4  Factor load of principal component and index of influence of various factors

主成分I1I2I3I4I5I6I7I8I9I10I11I12
Z1-0.65-0.180.229-0.39-0.460.5510.7170.2020.2340.3750.0930.262
Z2-0.140.840.2240.4110.2370.2060.7910.437-0.23-0.39-0.100.137
Z30.54-0.010.3710.6460.2260.0850.0970.3010.1380.5990.4350.239
Bi42.0627.7537.4435.7534.6139.7738.1430.4737.8936.7127.6033.71

新窗口打开| 下载CSV


表4可知,充填料浆密度、充填料浆腐蚀性和料浆流速与临界流速之比这3个指标在各主成分中所占的比值较小,对充填管道稳定性的影响也相对较小。再结合各指标的影响力指数Bi大小,最终确定将充填料浆密度和充填料浆腐蚀性这2个关联性弱的指标剔除,避免了各指标之间的共线性,同时突出了主要指标在充填管道磨损风险评估中的作用。

3.2 有序多分类Logistic回归参数计算及模型检验

结合表4中剩余主要指标及文献[1,12,20]的数据,运用SPSS进行有序多分类logistic回归处理,得到各项指标回归参数、标准误差及显著性水平,见表5

表5   管道磨损风险Logistic回归模型

Table 5  Logistic regression model of pipeline wear risk

变量回归系数标准误差显著性水平
I10.35070.31080.1128
I30.03720.18260.3972
I4-0.10550.26860.0293
I5-0.04690.17820.2833
I60.24160.13060.0633
I70.28170.10260.0375
I80.09830.17720.3985
I90.24780.18120.1755
I10-0.29280.14280.0427
I120.27930.16420.0694
β1.78550.76340.0259

新窗口打开| 下载CSV


将金川龙首矿、大红山铜矿、河东金矿和新城金矿4个矿山的调查数据代入风险评估模型进行计算,对数据结果进行归一化处理,得到4个矿山管道磨损性概率等级,见表6

表6   4个矿山管道磨损风险等级概率

Table 6  Pipeline wear risk rating for four mines

矿山名称各风险等级概率
金川龙首矿0.2470.4400.1540.153
大红山铜矿0.1790.2400.3230.258
河东金矿0.1810.2270.3450.247
新城金矿0.1700.2300.4180.182

新窗口打开| 下载CSV


由上述概率分布可知,金川龙首矿的充填管道磨损风险等级主要为Ⅱ,充填管道较易磨损。大红山铜矿、河东金矿和新城金矿的磨损风险等级主要为Ⅲ,充填管道容易磨损。

为检验本模型的正确性,将管道磨损风险等级评估结果与可变模糊耦合模型[12]和未确知测度理论评估模型[21]得到的评估结果进行比较,结果见表7

表7   不同模型风险评估结果比较

Table 7  Comparison of risk assessment results of different models

矿山名称综合风险等级
基于改进PCA与多分类Logistic回归主客观组合权重与可变模糊模型未确知测度综合评价模型
金川龙首矿
大红山铜矿
河东金矿
新城金矿

新窗口打开| 下载CSV


表7可知,本文评估模型对4个矿山的管道磨损风险评估等级与其他2种评估体系所得结果一致,且本文模型准确预测了风险发生的概率,表明该模型的建立是科学、合理的。

在生产过程中,应根据矿山的实际经济技术水平来制定降低充填管道磨损风险的具体措施。例如,对于产量大、效益高的矿山来说,通过改进PCA分析和有序多分类Logistic计算得到该矿山充填管道Ⅱ级磨损风险等级概率最大,但其Ⅲ级磨损风险等级概率也较大。此时,为了保障安全生产应将矿山的管道维护等级调整为Ⅲ级,以应对突发状况。对于小型矿山来说,若其Ⅰ级磨损风险等级概率最大,Ⅱ级磨损风险等级概率较小,结合安全水平和经济状况,认为此时矿山仅需要保持Ⅰ级的管道维护等级。

4 结论

(1)将改进PCA与有序多分类Logistic回归相结合,构建充填管道磨损风险评价模型,应用该模型筛除了影响指数较小、关联性较弱的指标,并预测了不同磨损风险等级发生的概率。

(2)选取充填料浆体积分数和充填倍线等12项指标建立充填管道磨损评估模型。用改进PCA算法筛除了充填料浆密度和充填料浆腐蚀性这2项影响指数小、关联性弱的指标。最终确定出金川龙首矿、大红山铜矿、新城金矿和河东金矿4个矿山充填管道磨损风险等级的准确概率,该评估结果与可变模糊耦合模型、未确知测度理论算法所得结果一致。

(3)基于改进PCA和有序多分类Logistic的充填管道磨损风险评估模型具有较高的精确度。不仅指明了各个指标的影响程度,而且预测了充填管道磨损风险等级概率,对类似矿山避免充填管道失效事故具有积极意义。

参考文献

WangSLiSQinJ Cet al.

Effect of anionic polyacrylamide on the structural stability of thickened tailings slurry in pipeline transportation

[J].Advances in Materials Science and Engineering,2018. DOI:10.1155/2018/7131487.

[本文引用: 3]

LiuZ XDangW GLiuQ Let al.

Optimization of clay material mixture ratio and filling process in gypsum mine goaf

[J]. International Journal of Mining Science and Technology,2013233): 337-342.

[本文引用: 1]

李夕兵刘冰.

硬岩矿山充填开采现状评述与探索

[J]. 黄金科学技术,2018264): 492-502.

[本文引用: 1]

LiXibingLiuBing.

Review and exploration of current situation of backfill mining in hard rock mines

[J].Gold Science and Technology2018264):492-502.

[本文引用: 1]

SivakuganNVeenstraRNaguleswaranN.

Underground mine backfilling in Australia using paste fills and hydraulic fills

[J]. International Journal of Geosynthetics and Ground Engineering,201512):18.

[本文引用: 1]

丁文军李广辉丁志浩.

某金矿充填系统优化改造

[J]. 黄金科学技术,2016244): 87-91.

[本文引用: 1]

DingWenjunLiGuanghuiDingZhihao.

Optimization of filling system for a gold mine

[J].Gold Science and Technology2016244):87-91.

[本文引用: 1]

GuptaA KPaulB.

A review on utilisation of coal mine overburden dump waste as underground mine filling material:A sustainable approach of mining

[J]. International Journal of Mining and Mineral Engineering,201562): 172-186.

[本文引用: 1]

王贤来郑晶晶张钦礼.

充填钻孔内管道磨损的影响因素及保护措施

[J]. 矿冶工程,2009295): 9-12.

[本文引用: 1]

WangXianlaiZhengJingjingZhangQinli,et al.

Influencing factors on the abrasion of pipelines in backfill drilling and protective measures

[J]. Mining and Metallurgical Engineering2009295): 9-12.

[本文引用: 1]

张钦礼郑晶晶王新民.

充填管道磨损形式及机理分析研究

[J]. 金属矿山,2009395): 115-118.

[本文引用: 1]

ZhangQinliZhengJingjingWangXinminet al.

A study of the abrasion form and mechanism of backfilling pipelines

[J]. Metal Mine2009395): 115-118.

[本文引用: 1]

张德明王新民郑晶晶.

深井充填钻孔内管道磨损机理及成因分析

[J]. 武汉理工大学学报,20103213): 100-105.

[本文引用: 1]

ZhangDemingWangXinminZhengJingjinget al.

Wear mechanism and causes of backfilling drill-holes pipelines in deep mine

[J]. Journal of Wuhan University of Technology20103213): 100-105.

[本文引用: 1]

毛明发王炳文朱家锐.

充填料浆自流输送管道磨损机理研究

[J]. 金属矿山,2018474): 178-184.

[本文引用: 1]

MaoMingfaWangBingwenZhuJiaruiet al.

Study on wear mechanism of gravity transportation pipeline for backfilling slurry

[J]. Metal Mine2018474): 178-184.

[本文引用: 1]

梅晓李兴华孔贺.

膏体充填管道磨损机理分析

[J]. 煤炭技术,2016357): 274-276.

[本文引用: 1]

MeiXiaoLiXinghuaKongHeet al.

Analysis on wear mechanism of backfill paste

[J]. Coal Technology2016357): 274-276.

[本文引用: 1]

薛希龙王新民张钦礼.

充填管道磨损风险评估的组合权重与可变模糊耦合模型

[J]. 中南大学学报(自然科学版),20164711): 3752-3758.

[本文引用: 4]

XueXilongWangXinminZhangQinli.

An integrated model of combination weights and variable fuzzy on evaluating backfill pipeline wear risk

[J]. Journal of Central South University(Science and Technology)20164711): 3752-3758.

[本文引用: 4]

丁文军李广辉丁志浩.

某金矿充填系统优化改造

[J].黄金科学技术,2016,244):87-91.

[本文引用: 1]

DingWenjun,LiGuanghui,DingZhihao.

Optimization of filling system for a gold mine

[J].Gold Science and Technology,2016,244):87-91.

[本文引用: 1]

叶永飞张雅楠李士超.

循环载荷下胶结充填体损伤声发射表征

[J]. 黄金科学技术,2018266): 819-825.

[本文引用: 1]

YeYongfeiZhangYananLiShichaoet al.

Acoustic emission characterization of damage of cemented filling under cyclic loading

[J].Gold Science and Technology2018266):819-825.

[本文引用: 1]

唐勇波桂卫华彭涛.

PCA和KICA特征提取的的变压器故障诊断模型

[J].高压电技术,2014402): 557-563.

[本文引用: 1]

TangYongboGuiWeihuaPengTaoet al.

Transfomer fault diagnosis model based on PCA and KICA feature extraction

[J].High Voltage Engineering2014402): 557-563.

[本文引用: 1]

王培良夏春江.

基于PCA-PDBNs的故障检测与自学习辨识

[J]. 仪器仪表学报,2015365): 1147-1154.

[本文引用: 1]

WangPeiliangXiaChunjiang.

Fault detection and self-learning identification based on PCA-PDBNs

[J]. Chinese Journal of Scientific Instrument2015365): 1147-1154.

[本文引用: 1]

赵琳边扬荣建.

基于有序Logistic回归的城市人行道服务水平研究

[J]. 交通运输系统工程与信息,2014144): 131-138.

[本文引用: 1]

ZhaoLinBianYangRongJianet al.

Pedestrian LOS of urban sidewalks based on orderly Logistic regression

[J]. Journal of Transportation Systems Engineering and Information Technology2014144): 131-138.

[本文引用: 1]

白少布.

基于有序Logistic模型的企业供应链融资模型融资风险预警研究

[J]. 经济经纬,20106): 66-71.

[本文引用: 1]

BaiShaobu.

Research on early warning of financing risk of enterprise supply chain financing model based on ordered Logistic model

[J]. Economic Survey20106): 66-71.

[本文引用: 1]

王瑷玲石宁陈欣玉.

基于有序Logistic回归的农村院落建设影响因素研究

[J]. 农林经济管理学报,2017166): 776-782.

[本文引用: 1]

WangAilingShiNingChenXinyuet al.

Factors influencing rural courtyard construction based on ordinal Logistic regression

[J].Journal of Agro-Forestry Economics and Management2017166): 776-782.

[本文引用: 1]

郑晶晶张钦礼王新民.

充填管道系统失效模式与影响分析(FMEA)及失效影响模糊评估

[J]. 中国安全科学学报,2009196): 166-171.

[本文引用: 2]

ZhengJingjingZhangQinliWangXinminet al.

FMEA analysis of backfilling pipeline system and fuzzy evaluation of failure effects

[J].China Safety Science Journal2009196): 166-171.

[本文引用: 2]

王新民王石鄢德波.

基于未确知测度理论的充填管道堵塞风险性评价

[J]. 中国安全科学学报,2012224): 151-156.

[本文引用: 1]

WangXinminWangShiYanDeboet al.

Risk assessment on blocking of filling pipeline based on uncertainty measurement theory

[J]. China Safety Science Journal2012224): 151-156.

[本文引用: 1]

/