Prediction of roadheader performance by artificial neural network

Performance prediction of the roadheaders is one of the main subjects in determining the economics of the underground excavation projects. During the last decades, researchers have focused on developing performance prediction models for roadheaders. In the first stage of this study, the performance of a roadheader used in Kucuksu sewage tunnel (Istanbul) was recorded in detail and the instantaneous cutting rate (ICR) of the machine was determined. The uniaxial compressive strength (UCS) and rock quality designation (RQD) are used as input parameters in previously developed empirical models in order to point out the efficiency of these models, and the relationships between measured and predicted ICR for different encountered formations. In the second stage of the study, Artificial Neural Network (ANN) technique is used for predicting of the ICR of the roadheader. A data set including UCS, RQD, and measured ICR are established. It is traced that a neural network with two inputs (RQD and UCS) and one hidden layer can be sufficient for the estimation of ICR. In addition, it is determined that increase in number of neurons in hidden layer has positive optimizing on the performance of the ANN and a hidden layer larger than 10 neurons does not have a significant effect on optimizing the performance of the neural network. Furthermore, probability of memorizing is being recognized in this situation. Based on this study, it is concluded that the prediction capacity of ANN is better than the empirical models developed previously.     

Effect of cutting tool blunting on the performances of various mechanical excavators used in low- and medium-strength rocks

Full-scale laboratory cutting tests that measure the specific energy (SE) are widely used to evaluate rock cuttability by mechanical excavators, and in particular roadheaders fitted with radial or drag-type bits. Radial or drag-type bits are often changed during operation as they wear and become blunt. In this study, full-scale cutting tests were carried out on different rock types using bits with varying degrees of wear in order to assess the impacts of pick bluntness on cutting forces and the SE. The relationships between wear flats and cutting forces, SE, and various rock properties such as uniaxial compressive strength, tensile strength, indentation index, Shore hardness, Schmidt hammer hardness, and density were examined and are discussed in this paper. The mean cutting force increased 2- to 3-fold and the cutting SE rose 4- to 5-fold with a 4-mm wear flat as compared to a sharp pick. Critical wear flats were plotted for different rock property values, and 25 MJ/m³ was considered the threshold SE above which cutting performance was considered to be poor. Best-fit predictive models based on statistical analysis of the laboratory cutting test results are introduced as a means to estimate SE as a function of bit type, wear condition, and various mechanical properties of the rock. These models can be used to predict the performances of mechanical excavators that use radial tools, especially roadheaders, continuous miners, and longwall drum shearers.     

1500—2000m^3／h斗轮挖掘机切割阻力的研究

文章介绍了采用缩小比例的模拟斗齿对小龙潭露天矿的原型煤、泥龙岩等进行垂直层理的直线型切割，通过对参与切割的试样物理、力学参数、切割参数的因次分析，建立无量纲方程，进行相似理论的模型试验和对诸参数组织正交试验，来预测原型斗轮的切割阻力，为设计开发斗轮挖掘机提供理论依据。最后，由整机性能测试来加以验证。     

Application of neural networks for the prediction of the unconfined compressive strength (UCS) from Equotip hardness

This paper presents the application of a neural network for the prediction of the UCS from hardness tests on rock samples. To investigate the suitability of this approach, the results of the network are compared to predictions obtained by conventional statistical relations.The network was trained to predict the UCS based on the hardness, porosity, density, grain size and rock type information of a rock sample. A dataset containing 194 rock sample records, ranging from weak sandstones to very strong granodiorites, was used to train the network with the Levenberg–Marquardt training algorithm. Two sets, each containing 17 rock samples, were used to validate the generalization and prediction capabilities of the network.     

利用神经元网络预测岩石或岩石工程的力学性态

本文将人工智能中的神经元网络引入岩石力学领域,用以预测岩石或岩石工程的力学性态。其用法类似经验公式,但其优点是影响岩体力学性态的各种描述性地质因素,均可做为变量输入,故可求得离散性较小的结果。文中附有两个实例,说明此方法的实用性。     

Machine Learning in Engineering Analysis and Design: An Integrated Fuzzy Neural Network Learning Model

Applying neural network computing to structural engineering problems has received increasing interest, with particular emphasis placed on a supervised neural network with the backpropagation (BP) learning algorithm. In this article, we present an integrated fuzzy neural network (IFN) learning model by integrating a newly developed unsupervised fuzzy neural network (UFN) reasoning model with a supervised learning model in structural engineering. The UFN reasoning model is developed on the basis of a single-layer laterally connected neural network with an unsupervised competing algorithm. The IFN learning model is compared with the BP learning algorithm as well as with a counterpropagation learning algorithm (CPN) using two engineering analysis and design examples from the recent literature. This comparison indicates not only a superior learning performance in solved instances but also a substantial decrease in computational time for the IFN learning model. In addition, the IFN learning model is applied to a complicated engineering design problem involving steel structures. The IFN learning model also demonstrates superior learning performance in a complicated structural design problem with a reasonable computational time.     

Prediction of blast-induced flyrock in Indian limestone minesusing neural networks

Frequency and scale of the blasting events are increasing to boost limestone production. Mines areapproaching close to inhabited areas due to growing population and limited availability of land resourceswhich has challenged the management to go for safe blasts with special reference to opencast mining.The study aims to predict the distance covered by the flyrock induced by blasting using artificial neuralnetwork （ANN） and multi-variate regression analysis （MVRA） for better assessment. Blast design andgeotechnical parameters, such as linear charge concentration, burden, stemming length, specific charge,unconfined compressive strength （UCS）, and rock quality designation （RQD）, have been selected as inputparameters and flyrock distance used as output parameter. ANN has been trained using 95 datasets ofexperimental blasts conducted in 4 opencast limestone mines in India. Thirty datasets have been used fortesting and validation of trained neural network. Flyrock distances have been predicted by ANN, MVRA,as well as further calculated using motion analysis of flyrock projectiles and compared with the observeddata. Back propagation neural network （BPNN） has been proven to be a superior predictive tool whencompared with MVRA.
2014 Institute of Rock and Soil Mechanics, Chinese Academy of Sciences. Production and hosting byElsevier B.V. All rights reserved.     

淮南矿区11–2煤顶板岩石单轴抗压强度预测模型构建

为了探寻煤系地层泥岩和砂岩类单轴抗压强度和弹性模量间的关系,对顾北煤矿11–2煤顶板岩石进行室内单轴压缩试验,得到120组数据。采用SPSS19.0专业数理统计软件,以单轴抗压强度为因变量、弹性模量为自变量,分别建立线性函数、对数函数、二次函数、三次函数、指数函数和幂函数6种回归模型。经回归方程显著性检验、回归系数显著性检验、回归方程拟合优度、D-W检验、共线性检验和回归方程残差正态性、随机性检验,获得预测岩石单轴抗压强度的一元二次非线性回归预测模型。应用该模型对淮南矿区张集煤矿和潘二煤矿11–2煤顶板岩石单轴抗压强度进行预测与验证,结果表明预测值与试验值相符,该非线性回归预测模型可用于同一煤系地层岩石力学参数的预测。     

位移反分析的进化神经网络方法研究

将人工神经与遗传算法相结合,提出了一种用于位移反分析的进化神经网络方法,这种方法基于正交试验获得的样本进行学习,用遗传算法搜索最优的神经网络结构,并用最佳推广预测学习算法训练此网络,以此训练好的网络描述岩体（土）的力学参数与岩体位移是非线性关系,再应用遗传算法从全局空间上搜索,进行岩体力学参数的最优辩识。作为例子,文中给出了弹性问题的反分析,结果是令人满意的。     

范各庄煤矿砂岩岩体结构数字识别及参数表征

 借助先进的3GSM三维岩体不接触测量技术,对范各庄煤矿12煤底板砂岩岩体结构面进行现场测量,获取一系列真实描述岩体宏观结构的数字图像,提取节理几何形态空间分布信息,建立三维岩体结构面空间分布模型。在此基础上,得到砂岩岩体结构面联通网络,基于离散介质渗流方法和现场水文试验数据,建立岩体水力学参数表征方法,得到宏观岩体水力学参数,并分析参数的尺寸效应,实现岩体结构参数(几何形态)数字信息与水力学参数的定量表征有机衔接,为范各庄煤矿12煤底板砂岩岩体结构稳定性分析和突水危险性评价提供更加真实、可靠的数据支持。     

STUDY ON GENETIC-NEURAL NETWORK METHOD OF DISPLACEMENT BACK ANALYSIS

An evolutionary neural network method for displaceme nt back analysis is proposed by combining the neural network and genetic algorit hm. The samples produced in orthogonal experiment are used to train the neural n etwork whose architecture is determined in global optimum by genetic algorithm. Thus, the neural network with optimal architecture trained by optimal prediction algorithm is used to describe the relationship between the rock mechanical para meters and displacements produced due to excavation. Then genetic algorithm is a dopted again to search the optimal rock mechanical parameters in their globa l ranges. As an example, a back analysis for elastic problem is introduced. The results are satisfactory.     

Predicting uniaxial compressive strength of serpentinites through physical,dynamic and mechanical properties using neural networks

The uniaxial compressive strength(UCS)of intact rock is one of the most important parameters required and determined for rock mechanics studies in engineering projects.The limitations and difficulty of conducting tests on rocks,specifically on thinly bedded,highly fractured,highly porous and weak rocks,as well as the fact that these tests are destructive,expensive and time-consuming,lead to development of soft computing-based techniques.Application of artificial neural networks(ANNs)for predicting UCS has become an attractive alternative for geotechnical engineering scientists.In this study,an ANN was designed with the aim of indirectly predicting UCS through the serpentinization percentage,and physical,dynamic and mechanical characteristics of serpentinites.For this purpose,data obtained in earlier experimental work from central Greece were used.The ANN-based results were compared with the experimental ones and those obtained from previous analysis.The proposed ANN-based formula was found to be very efficient in predicting UCS values and the samples could be classified with simple physical,dynamic and mechanical tests,thus the expensive,difficult,time-consuming and destructive mechanical tests could be avoided.     

Elastic and shear moduli of coal measure rocks derived from basic well logs using fractal statistics and radial basis functions

Gamma ray, density, sonic and core logs obtained from two boreholes drilled over a longwall panel in Southwestern (SW) Pennsylvania were analyzed for formation boundaries, log-derived porosities and densities and for rock elastic properties from sonic transit times. Gamma ray (GR) and density logs (DL) were analyzed using univariate statistical techniques and fractal statistics for similarity and ordering of the log data in depth. A Fourier transformation with low-pass filter was used as a noise elimination (filtering) technique from the original logs. Filtered data was tested using basic univariate and fractal statistics, rescaled range (R/S) and power spectrum (PS) analysis to compare the information characteristics of the filtered logs with the original data. The randomness of log data in depth was analyzed for fractional Gaussian noise (fGn) or fractional Brownian motion (fBm) character.A new prediction technique using radial basis function (RBF) networks was developed to calculate shear and Young's moduli of the formations when sonic logs are not available. For this approach, the filtered logs were used as input to an RBF based upon a combination of supervised and unsupervised learning. The network was trained and tested using rock elastic properties calculated from the sonic log of one of the boreholes. The network was used to predict the elastic and shear moduli of the coal-measure rocks over a longwall coal mine in SW Pennsylvania. This approach demonstrated that it could be used for prediction of elastic and shear moduli of coal-measure rocks with reasonable accuracy.     

Identifying the cutting tool type used in excavations using neural networks

The paper presents results of preliminary research on utilising neural networks to identify excavating cutting tool’s type used in multi-tool excavating heads of mechanical coal miners. Such research is necessary to identify rock excavating process with a given head, and construct adaptation systems for control of excavating process with such a head.     

A Backpropagation Neural Network Model for Semi-rigid Steel Connections

Abstract: The analysis of semirigid steel structure connections based on exact theoretical modeling, which is demanding and time consuming if all the nonlinear parameters of the problem are taken into account, can be avoided provided that enough experimental measurements exist and an appropriate predictor can be constructed from them. A supervised learning backpropagation neural network approach is proposed in this paper for the construction of this model free predictor. A number of experimental momentrotation curves for single-angle and single-plate beam-to-column connections are used in this paper to train the neural network. The trained network provides us with an estimator for the mechanical behavior of the steel structure connection element.     

Intelligent characterisation and diagnosis of the groundwater quality in an urban fractured-rock aquifer using an artificial neural network

This paper addresses the problem of how to diagnose the effect of stormwater infiltration on groundwater quality variables and to capture the complex nonlinear relationships that exist between groundwater quality variables. It is argued that because of the complex nonlinear relationships between the groundwater quality variables, classical linear statistical methods are unreliable and difficult to visualise the results. The application of intelligent techniques, which can analyse the multi-dimensional groundwater quality data with the sophisticated visualisation technique, is vital for sustainable groundwater management.In this paper, the Kohonen self-organising feature maps (KSOFM) neural network is applied to analyse the effect of stormwater infiltration on the groundwater quality, and diagnose the inter-relationship of the groundwater quality variables in a fractured rock aquifer. Based on the pattern analysis visualised in component planes and U-matrix, the inter-relationships among the groundwater quality variables due to the stormwater infiltration are extracted and interpreted. The pattern distribution of groundwater quality variables due to different aquifer conditions is also analysed.It is concluded that the KSOFM technique described in this paper provides an effective analysing and diagnosing tool to understand the dynamic in the groundwater quality and to extract knowledge contained in the multi-dimensional data. Finally it has considerable potential not only in groundwater quality monitoring and diagnosis, but also in other environmental areas.     

Dominant rock properties affecting the performance of conical picks and the comparison of some experimental and theoretical results

Conical picks are the essential cutting tools used especially on roadheaders, continuous miners and shearers and their cutting performance affects directly the efficiency and the cost of rock/mineral excavation. In this study, in order to better understand the effects of dominant rock properties on cutter performance, 22 different rock specimens having compressive strength values varying from 10 to 170 MPa are first subjected to a wide range of mechanical tests. Then, laboratory full-scale linear cutting tests with different depth of cut and cutter spacing values are realized on large blocks of rock specimens using one type of conical pick. Specific energy, cutting and normal force values for relieved and unrelieved cutting modes are recorded using a triaxial force dynamometer with capacity of 50 tonnes and a data acquisition system. Cutter force and specific energy values are correlated with rock properties and theoretical force and specific energy values obtained from widely used theoretical approaches.The results indicate that uniaxial compressive strength among the rock properties investigated is best correlated with the measured cutter performance values, which is in good agreement with previous studies. However, it is also emphasized in this study that Brazilian tensile strength, Schmidt hammer rebound values, static and dynamic elasticity modulus are also dominant rock properties affecting cutter performance.Theoretical specific energy defined by different researchers has a meaningful relationship with the experimental specific energy, which is an essential parameter for predicting the instantaneous cutting rates of mechanical excavation systems. It is also demonstrated that the experimental cutter forces obtained for 5 mm depth of cut are in good agreement with theoretical force values, if the friction angle between rock and cutting tool is included in the theoretical formulation. It is emphasized that, to some extend, laboratory tests can help to minimize high cost of a trial–error approach in the field.     

煤矿深部岩石力学性能试验分析与硬岩巷道快速掘进方法研究

 在RMT–150B岩石力学试验机上进行深部巷道砂岩单轴压缩和不同围压下三轴压缩试验,测得砂岩的力学性质参数。根据岩石应力–应变曲线分析深部岩石的强度和变形特性,岩石抗压强度随围压的增加而提高,围压从10 MPa增加到15 MPa时,抗压强度增幅达到40.3%。深部高应力下,砂岩承载后产生的变形及破坏形态与围压大小密切相关,其主应力差–应变曲线斜率随着围压的增加而明显变陡,破坏荷载增高。依据岩石力学特性,采用分类布孔,掏槽眼宜用直径f42 mm的钻头钻眼,其余的炮眼用直径f32 mm的钻头钻眼,缩短钻眼时间。掏槽眼采用中深孔不同阶微差斜眼掏槽方法,炮眼深度宜采用2.2～2.5 m,有利于巷道的进尺;周边眼采用小直径药卷光面爆破技术,有利于巷道成形。     

深部矿区煤岩体强度测试与分析

 基于钻孔触探法原理,开发出小孔径井下煤岩体强度测定装置。在实验室对34个煤岩样品进行试验：在煤岩块上钻取标准试件,测量单轴抗压强度;在留下的钻孔中,用煤岩体强度测定装置测定探针临界载荷,分析探针破坏钻孔壁煤岩的形态;然后确定煤岩块单轴抗压强度与探针临界载荷的关系。试验表明,探针破坏钻孔壁煤岩的形状、深度及范围与煤岩性质密切相关。煤岩体强度越高,破坏范围、侵入深度越小,破坏形状越规则。结合井下实测数据,回归得出描述探针临界载荷与煤岩体单轴抗压强度关系的公式。同时,分析临界载荷的离散性及控制措施,讨论结构面对煤岩体强度的影响及测试分析方法,并在典型的深部矿区——新汶矿区进行井下原位测试。新汶矿区巷道顶板不同岩性的岩层强度相差很大,不同矿井的岩层强度也存在明显差别。煤层强度由于煤帮出现破碎区、煤层性质不均匀、煤层结构面分布不均匀等原因变化较大,出现明显的波动。基于井下煤岩体强度实测数据的巷道支护设计,符合井下环境中的煤岩体条件,设计的合理性与可靠性显著提高,巷道围岩稳定性与支护状况得到明显改善。最后分析钻孔触探法存在的问题,并提出改进建议。     

Assessment of pick wear based on the field performance of two transverse type roadheaders: a case study from Amasra coalfield

A successful excavation of roadheaders depends on the cutting performance and the tool life of conical picks. Tool life is important in terms of wear rate which is affected by different rock parameters such as equivalent quartz content, mineral grain size, as well as cutting parameters on the cutterhead. In this study, analyses among wear rate, specific energy, advance rate, and cutter consumption were carried out. The wear mechanisms of two different models of conical picks were examined from different aspects depending on rock and machine parameters. Their relation with the mechanical and abrasivity properties of rocks and petrographic analyses were investigated. In addition, the metallurgic content and Rockwell hardness of conical picks were determined to describe the metal alloys and their effects on the wear of cutting tool. The results showed that the metallurgic content, pick positions, and other environmental conditions influence the wear mechanism. Finally, two different models were proposed to estimate the pick consumption in sandstone and siltstone rocks based on actual data obtained from coalfield.