Abstract
Overbreak is an undesirable phenomenon in blasting operations. The causing factors of overbreak can be generally divided as blasting and geological parameters. Due to multiplicity of effective parameters and complexity of interactions among these parameters, empirical methods may not be fully appropriated for blasting pattern design. In this research, artificial neural network (ANN) as a powerful tool for solving such complicated problems is developed to predict overbreak induced by blasting operations in the Gardaneh Rokh tunnel, Iran. To develop an ANN model, an established database comprising of 255 datasets has been utilized. A three-layer ANN was found as an optimum model for prediction of overbreak. The coefficient of determination (R2) and root mean square error (RMSE) values of the selected model were obtained as 0.921, 0.4820, 0.923 and 0.4277 for training and testing, respectively, which demonstrate a high capability of ANN in predicting overbreak. After selecting the best model, the selected model was used for optimization purpose using artificial bee colony (ABC) algorithm as one of the most powerful optimization algorithms. Considering this point that overbreak is one of the main problems in tunneling, reducing its amount causes to have a good tunneling operation. After making several models of optimization and variations in its weights, the optimum amount for the extra drilling was 1.63 m2, which is 47% lower than the lowest value (3.055 m2). It can be concluded that ABC algorithm can be introduced as a new optimizing algorithm to minimize overbreak induced by tunneling.
Similar content being viewed by others
Explore related subjects
Discover the latest articles, news and stories from top researchers in related subjects.Avoid common mistakes on your manuscript.
1 Introduction
The tunnels have been excavated for various purposes such as road construction and water transferring in civil and mining works. Although new mechanized excavation methods such as tunnel boring machine (TBM) have been successfully utilized for tunnel excavation, drilling and blasting as a traditional technique can be still used for excavation of tunnels with different shapes and sizes [1]. In fact, it is a primary excavation technique due to its advantages of high flexibility and low cost [2]. Nevertheless, because of using explosive material for rock mass excavation, damages to the peripheral rock mass around the excavation are inevitable. After the blast operation, the excavation cross-section can have two major problems. These two problems (Fig. 1) are called overbreak and underbreak. The over break is defined as a surplus drilled section of the tunnel and the underbreak is defined as the remainder of the blast operation.
Overbreak phenomenon in the executive process of a tunneling project is always one of the most important issues. Nowadays, according to the progress of industry and entrance of new technologies to tunneling industry and gradual acceptance, the new methods are replaced instead of traditional methods (drilling and blast). Though, the tunneling industry uses advanced equipment, there are still reports of the overbreak phenomenon. The main reason is referred to the variety of gender stone and various geological effects which generally cannot be predicted until approaching time. On the other hand, tunnel projects with low volumes, small to medium scale, employers and contractors are unwilling to invest for entering the new mechanized equipment instead of traditional methods. Therefore, it can be argued that the traditional methods specially drilling and blasting are the most common tunnel excavation method for small-to-medium-scale projects. It should be noted that, as far as writers are concerned, there is no experimental and analytical method on the determination of overbreak in the tunnels.
Recently, artificial intelligence (AI) methods such as artificial neural networks (ANN), fuzzy inference system (FIS), and neuro-fuzzy inference system (ANFIS) are developed to solve geotechnical problems [3,4,5,6,7,8,9,10,11,12,13,14,15,16,17,18]. Several researchers used AI methods to predict the uniaxial compressive strength of the rock [19,20,21,22]. Momeni et al. [23] highlighted the use of ANN technique to predict bearing capacity of the pile. In addition, this method was utilized to solve the problem of ground settlement induced by tunneling in the study carried out by Ocak and Seker [24]. Gordan et al. [25] predicted the stability of homogeneous slopes using a combination of particle swarm optimization (PSO) and ANN.
The effective factors of overbreak can be divided into three groups of rock mass characteristics, geometric properties of the explosion pattern and blasting properties [26]. So far, many researches worked on overbreak in mines and tunnels. Monjezi and Dehghani [27] considered the ratio of stemming to burden, charge the last row to total charge, special charge, special charge per delay and the number of explosion rows in each stage in the GOL-GOHAR mine, Iran as the most influencing factors on overbreak. Hyongdoo Jang and Erkan Topal [28] predicted the overbreak of the Gyby tunnel in South Korea with a value of 0.945 for correlation coefficient (R) between the output of model and the actual data, using ANNs. Their model inputs include uniaxial compressive strength of the rock mass, quality index of rock mass, rock weathering conditions, groundwater conditions, and geomechanical classification index values of rock mass to predict over break.
To propose a comprehensive ANN model, Monjezi et al. [29] used parameters of uniaxial compressive strength, especially drilling, underground water content, burden, hole spacing, stemming, the diameter of the hole, stair height, special charge and consumer charge in ever delay as model inputs. They surveyed the sensitivity analysis on the mentioned parameters and concluded that burden and underground water content are the most effective and least important parameters, respectively. Hook and Brown [30] reported that when discontinuities are parallel along the tunnel axis, they have undesirable effects on over break. Ebrahimi et al. [31] introduced burden, spacing and charge per delay as the most influential parameters on overbreak. Developing an ant colony optimization algorithm, Saghatforoush et al. [32] reduced overbreak and flyrock resulting from blasting, 61 and 58%, respectively.
Gates et al. [33] expressed that insufficient delay time and the number of explosive rows are the most effective factors on over break. Esmaeili et al. [34] suggested that the last row of charge and special charge are the most important factors on over break, while the ratio of burden to spacing, stiffness and density are the least important ones. Ibarra et al. [35] mentioned that the charge factor of environment from explosive can create underbreak in tunnels and the reduction of rock quality may create over break. Mandal [26] expressed that in addition to the rock conditions, in situ stress has also a deep effect on overbreak. Several researchers have suggested empirical models to estimate overbreak [36, 37]. Singh and Xavier [38] carried out a series of small experiments on the physical-scale models to predict the blast damage. They considered characteristics of the rock mass and the explosive material as the most effective parameters on overbreak. Recently, Koopialipoor et al. [39] used a hybrid of genetic algorithm (GA)-ANN model for prediction of overbreak in tunnels and successfully demonstrated that their developed models can predict overbreak with high degree of accuracy.
By reviewing mentioned works, it can be concluded that due to the multiplicity of the effective factors as well as the complex relationship between these parameters, there is a need to develop a new technique to predict and control overbreak phenomenon. On the other hand, parameters influencing overbreak are related to the specific condition of the project. Therefore, it is necessary to evaluate overbreak phenomenon for each project before conducting the operations. The present study attempts to predict overbreak phenomenon at Gardaneh Rokh tunnel, Iran, using ANN approach. Then, artificial bee colony (ABC) is considered and proposed to optimize the blast pattern parameters for minimization of overbreak in the tunnel.
This paper is presented in seven sections: after the current section, the studied location together with data collection will be explained in Sect. 2. Sections 3 and 4 are about background of ANN model and how to implement it for prediction of over-break, respectively. Some material regarding structure of ABC can be found in Sect. 5 and in Sect. 6, modelling process of ABC and its effective parameters to optimize over-break will be described. Finally, the conclusions of this paper are given in the last section.
2 Case study
The Gardaneh Rokh tunnel is one of the most important tunnels in West of Iran. The tunnel is the main road of communication between two centers of Esfahan, and Chaharmahal and Bakhtiari provinces. The road to Isfahan–Shahrekord is economically and strategically crucial highways in Iran. The mentioned tunnel which is a communication path was removed to 7 km from the capital of the provinces and about 100 twists and accident-prone points on this axis. The tunnel with a length of 1300 m and width of 13 m was investigated in Chaharmahal and Bakhtiari Province which is located at a distance of 30 km from Shahrekord city. The location of the study area can be seen in Fig. 2. Excavation of this tunnel, as shown in Fig. 3, was performed in two sections: the top one which was excavated using drilling and blasting method, and the bottom one which was excavated using hydraulic hammer.
In excavation operations of top section, the relatively constant explosive pattern was used to drill the tunnel. In this project, there are some minor changes in the pattern of explosions, but the changes were not varied enough to be considered in different arrangements for each stage of the explosion. Table 1 shows general specifications of excavation at top section of the tunnel with a period of explosive design parameter changes in different conditions of rock mass.
In this study, eight parameters were selected as inputs of model for prediction of overbreak, which included 255 datasets of RMR, advanced length, special charge, the holes periphery burden, the end row burden, periphery spacing, end row spacing, and number of applied delay. A simple description of these parameters (input and output) is shown in Table 2.
3 Artificial neural network
Artificial neural network (ANN) was developed by McCulloch and Pittsin [40]. This flexible technique is a type of artificial intelligent (AI) sytem which can solve problems faster with a high degree of accuracy. Furthermore, it can be used to solve non-linear nature issues where input and output parameters are considered as unknown [41]. The ANN is an imitation of the mechanism of data analysis of biological cells. The brain is a high complex network which can act as a parallel processor. Such networks are designed mainly for a series of non-linear mapping between input and outputs. ANNs are learned from previous experiences and generalized using the training samples. They are able to change their behavior based on the environment and are appropriate for the required algorithms for mapping. In ANN systems, the data used to create models are known as training data. In other words, ANNs use training data to learn patterns in the data which can prepare them to achieve the different outputs and results [42]. The structure of ANNs is created by processor units (neurons or nodes), which are responsible for the organization. These neurons can be combined with each other to form the layer. There are different ways to link neural in ANN. Feedforward (FF)-back-propagation (BP) is a common procedure for application in ANNs that its successful use has been reported by many researchers [14, 23, 43,44,45,46]. Each neuron has multiple inputs. These inputs are combined and then the combination of them provides an output after processing. Network cells are connected to each other which output of each cell is considered as the next cell input. The first layer on the left side of the input layer does not play any role in processing and, inputs only import in this section, through existing communications sent to the next layer to the process. The end layer (layer right) is an output layer that provides network response. The layers between input and output layer are called hidden or intermediate layers [47].
One of the most widely learning algorithms in ANN is learning algorithm of error back propagation [48]. The algorithm works on the basis of the error correction learning law, which can be considered extended algorithm of at least average. In general, learning propagation consists of two steps: forward step and back step. In forward stage, the inputs are forward layer by layer in the network, and finally a series of network output will be obtained as predicted values. During the forward stage, synaptic weights will be achieved. On the other hand, in the backward process, the weights are set the error by regulating laws. The difference between predicted response and network response (expected), which is called the error signal, will be released in the opposite direction of network connections and the weights change in a way that predicted response becomes closer to favorable response. Since the recent distribution is made in contrast to the weighted connections, error back propagation is chosen to explain the modification behavior of network. Different performance indices can be used to evaluate system results [42, 49,50,51,52,53,54]:
-
A.
The correlation coefficient (R2)
-
B.
The root mean square error (RMSE)
4 Developing ANN model
In the current study, the perceptron ANN model, which consists of three layers, was used to predict overbreak. Herein, three different learning algorithms were used to learn the ANN. These three algorithms include Levenberg–Markvart (LM), one-step secant (OSS) and scaled conjugate gradient (SCG), which can be compared between common functions for choosing the best learning function. Some researchers have proven that three layers can solve various and non-linear issues (e.g., Hornik et al. [55]). For this, the number of neurons in the first layer is equal to the number of input data (nine neurons). In addition, since the goal is an outlet, a neuron is also determined in the output layer. Finally, for determining hidden layer, several research studies have been conducted to select the number of hidden layer neurons in which they suggested appropriate numbers. There is a need to conduct trial and error methods to obtain the appropriate values for hidden neuron number. In Table 3, several researchers have proposed relationships to select the number of neurons in which Ni is the number of inputs, no is the number of outputs of the model.
According to the values of relationships presented in Table 3, for all three neural network learning algorithms from a range between 2 and 18 neurons for hidden neuron number was considered. Considering the importance of R2 and RMSE of each series of training and testing systems, a comparison was made between them to select the best model. This comparison is based on a proposed technique by Zorlu et al. [63], where each section is evaluated assigning a score. Based on this method, every performance index (R2 or RMSE) was calculated in its own class and best of them got highest ratin/ranking. For instance, values of 0.913, 0.913, 0.904, 0.908, 0.902, 0.912, 0.912, 0.913, 0.913, 0.902, 0.898, 0.931, 0.925, and 0.915 were obtained for section of R2 training dataset for models 1–14, respectively. Ranking results of the mentioned 12 models were, respectively, obtained as 31, 31, 27, 28, 26, 30, 30, 31, 31, 26, 25, 40, 37, and 32. It should be noted that scores are generated for 42 models, the highest score of which is assigned to the best section, and if the two sections are the same, the same score is awarded to them. Finally, the score of each row is aggregated from the models and is considered as a total score. Table 4 presents the results of this neural network. As shown in Table 4, Model No. 21 created with the LM learning algorithm was selected as the best model based on the highest score.
Figures 4 and 5 show the results of training and testing stages for the selected model (model number 21). As it can be seen, R2 values of 0.921 and 0.923 for training and testing show ability of ANN model in predicting overbreak. In fact, ANN can provide a high-level prediction capacity for the estimation of overbreak with a low error.
In the following, basis of an optimization algorithm, namely, artificial bee colony (ABC), in optimizing overbreak and its effective parameters are described. After that, minimization process of overbreak and input parameters will be presented and discussed.
5 Artificial bee colony
In this research, one of the new optimization algorithms called ABC algorithm has been used. This algorithm is based on the life of bees and is introduced by Karaboga [64]. In this algorithm, the bees form a colony together. In Clooney, the bees are just as simple components of the whole collection, which can be used to explore and search resources (answers). The same thing causes use of this algorithm to find the answers of the various problems. There are three major groups in each colony that work together to look for the best answer. The first group, known as scout bees, is used in the environment (search space) to seek resources (goals or problem answers). After returning the bees to the hive and exchanging information with the second-group bees (employed bees), the discovered resources begin to extract. Finally, the third group of bees (onlooker bees) in the hive uses resource information to evaluate responses in terms of fitness and provide the best sources (answers) to the hive (system). In this research, algorithm coding is implemented in MATLAB environmental software. The general flowchart of this algorithm, which is used to optimize overbreak in tunnels, is presented in detail in Fig. 6.
Recently, this algorithm has been applied in various engineering fields [31, 65,66,67,68,69]. Its major applications are to optimize engineering issues. Furthermore, some researchers have recently used this algorithm to improve the performance of ANN [3, 44]. More details regarding ABC structure and how to work, can be found in other studies [64, 70, 71].
6 Optimization of overbreak by ABC
In this research, after selecting the best ANN network (No. 21), the ABC algorithm was used to minimize overbreak results of the tunnel. Considering the amount of RMR which is varies from 30 to 39, the highest number of samples was selected with a RMR of about 36, and an optimization for this range was obtained from the rock mass of the tunnel pathway. As the complete explanation of this algorithm is given above, this search continues to find the minimum amount of over break. Several models of the ABC algorithm were implemented with different bees. In Fig. 7, several results of the algorithm show the best cost of over break.
As shown in Fig. 7, the bee impacts were evaluated considering total iteration number of 300 and number of bees in the range of 15–60. As a result, generally, system performance would be better by increasing number of bees. Nevertheless, after number of bees = 40 there was a very small difference between results of system. Hence, 40 was selected as the appreciate bee number. Focusing on iteration number, it was also found that after iteration number = 150, there was almost no changes, so this number was selected and utilized as the optimum one.
After analyzing the output parameters, the optimized parameters are given in Table 5. According to the ABC algorithm, the optimized particle for overbreak in tunnels, which is executed by drilling and blasting method, was 1.63 m2. As indicated in Table 2, the minimum amount for overbreak was about 3.055 m2, which was reduced as 47% compared to the original state using the optimization ABC algorithm. Moreover, the values of 4, 0.921, 1.796, 1.695, 1.242, 1.235 and 3.927 in the sequence for the number of delays, the load of the last row of the chess, the row of the first row of the hull, the intervals of the last row headers, the distance to the front end hulls, the special spending, and the length of the advance were achieved by ABC algorithm. The results indicate that by developing an ABC algorithm, the optimum values can be obtained and the overbreak values resulting from drilling and blasting in tunell can be minimized. Considering that overbreak is one of the main problems in tunneling, the reduction of this amount can contribute to have a good tunneling operation and its stability.
7 Conclusions
In the present study, using the help of AI models, prediction and optimization of overbreak in tunnel were conducted. After identifying the effective parameters in the overbreak phenomenon, eight input parameters were used to create a neural network of three types of learning function. After selecting the best model based on scoring, the selected model was used for optimization. The R2 and RMSE values of the selected model were 0.921, 0.4820, 0.923 and 0.4277 for training and testing, respectively. The ABC algorithm, one of the new optimization algorithms, was used to optimize these parameters of the explosion pattern. Considering that overbreak is one of the main problems in tunneling, the reduction of this amount can contribute to have a good tunneling operation and its stability. After making several models of optimization and variations in its weights, the optimum amount for the extra drilling was 1.63 m2, which is 47% lower than the lowest value (3.055 m2). Finally, this method can obtain the optimal pattern minimizing the amount of overbreak in the tunnels. It can be concluded that the developed algorithms in this study can be used in industry and practice considering ranges of model inputs with caution.
References
Bhandari S (1997) Engineering rock blasting operations. A A Balkema, Amsterdam, p 388
Raina AK, Murthy V, Soni AK (2014) Flyrock in bench blasting: a comprehensive review. Bull Eng Geol Environ 73:1199–1209
Koopialipoor M, Armaghani DJ, Hedayat A et al (2018) Applying various hybrid intelligent systems to evaluate and predict slope stability under static and dynamic conditions. Soft Comput. https://doi.org/10.1007/s00500-018-3253-3
Hasanipanah M, Armaghani DJ, Amnieh HB et al (2018) A risk-based technique to analyze flyrock results through rock engineering system. Geotech Geol Eng 36:2247. https://doi.org/10.1007/s10706-018-0459-1
Asadi A, Moayedi H, Huat BBK et al (2011) Artificial neural networks approach for electrochemical resistivity of highly organic soil. Int J Electrochem Sci 6:1135–1145
Moayedi H, Hayati S (2018) Applicability of a CPT-based neural network solution in predicting load-settlement responses of bored pile. Int J Geomech 18:6018009
Hasanipanah M, Jahed Armaghani D, Khamesi H et al (2016) Several non-linear models in estimating air-overpressure resulting from mine blasting. Eng Comput. https://doi.org/10.1007/s00366-015-0425-y
Hasanipanah M, Monjezi M, Shahnazar A et al (2015) Feasibility of indirect determination of blast induced ground vibration based on support vector machine. Meas J Int Meas Confed. https://doi.org/10.1016/j.measurement.2015.07.019
Hasanipanah M, Shahnazar A, Bakhshandeh Amnieh H, Jahed Armaghani D (2017) Prediction of air-overpressure caused by mine blasting using a new hybrid PSO–SVR model. Eng Comput. https://doi.org/10.1007/s00366-016-0453-2
Amiri M, Amnieh HB, Hasanipanah M, Khanli LM (2016) A new combination of artificial neural network and K-nearest neighbors models to predict blast-induced ground vibration and air-overpressure. Eng Comput 32:631–644
Koopialipoor M, Nikouei SS, Marto A et al (2018) Predicting tunnel boring machine performance through a new model based on the group method of data handling. Bull Eng Geol Environ. https://doi.org/10.1007/s10064-018-1349-8
Armaghani DJ, Mahdiyar A, Hasanipanah M et al (2016) Risk assessment and prediction of flyrock distance by combined multiple regression analysis and Monte Carlo simulation of quarry blasting. Rock Mech Rock Eng 49:1–11. https://doi.org/10.1007/s00603-016-1015-z
Armaghani DJ, Hasanipanah M, Mohamad ET (2016) A combination of the ICA-ANN model to predict air-overpressure resulting from blasting. Eng Comput 32:155–171. https://doi.org/10.1007/s00366-015-0408-z
Hasanipanah M, Noorian-Bidgoli M, Jahed Armaghani D, Khamesi H (2016) Feasibility of PSO-ANN model for predicting surface settlement caused by tunneling. Eng Comput. https://doi.org/10.1007/s00366-016-0447-0
Hasanipanah M, Jahed Armaghani D, Bakhshandeh Amnieh H et al (2016) Application of PSO to develop a powerful equation for prediction of flyrock due to blasting. Neural Comput Appl. https://doi.org/10.1007/s00521-016-2434-1
Khandelwal M, Armaghani DJ (2016) Prediction of drillability of rocks with strength properties using a hybrid GA-ANN technique. Geotech Geol Eng 34:605–620. https://doi.org/10.1007/s10706-015-9970-9
Shams S, Monjezi M, Majd VJ, Armaghani DJ (2015) Application of fuzzy inference system for prediction of rock fragmentation induced by blasting. Arab J Geosci 8:10819–10832
Jahed Armaghani D, Mohd Amin MF, Yagiz S et al (2016) Prediction of the uniaxial compressive strength of sandstone using various modeling techniques. Int J Rock Mech Min Sci. https://doi.org/10.1016/j.ijrmms.2016.03.018
Raina AK, Haldar A, Chakraborty AK et al (2004) Human response to blast-induced vibration and air-overpressure: an Indian scenario. Bull Eng Geol Environ 63:209–214
Hajihassani M, Jahed Armaghani D, Monjezi M et al (2015) Blast-induced air and ground vibration prediction: a particle swarm optimization-based artificial neural network approach. Environ Earth Sci 74:2799–2817. https://doi.org/10.1007/s12665-015-4274-1
Marto A, Hajihassani M, Jahed Armaghani D et al (2014) A novel approach for blast-induced flyrock prediction based on imperialist competitive algorithm and artificial neural network. Sci World J 2014:643715
Jahed Armaghani D, Tonnizam Mohamad E, Hajihassani M et al (2016) Evaluation and prediction of flyrock resulting from blasting operations using empirical and computational methods. Eng Comput. https://doi.org/10.1007/s00366-015-0402-5
Momeni E, Jahed Armaghani D, Hajihassani M, Mohd Amin MF (2015) Prediction of uniaxial compressive strength of rock samples using hybrid particle swarm optimization-based artificial neural networks. Meas J Int Meas Confed. https://doi.org/10.1016/j.measurement.2014.09.075
Ocak I, Seker SE (2013) Calculation of surface settlements caused by EPBM tunneling using artificial neural network, SVM, and Gaussian processes. Environ Earth Sci 70:1263–1276
Gordan B, Jahed Armaghani D, Hajihassani M, Monjezi M (2016) Prediction of seismic slope stability through combination of particle swarm optimization and neural network. Eng Comput https://doi.org/10.1007/s00366-015-0400-7
Mandal SK, Singh MM (2009) Evaluating extent and causes of overbreak in tunnels. Tunn Undergr Sp Technol 24:22–36
Monjezi M, Dehghani H (2008) Evaluation of effect of blasting pattern parameters on back break using neural networks. Int J Rock Mech Min Sci 45:1446–1453
Jang H, Topal E (2013) Optimizing overbreak prediction based on geological parameters comparing multiple regression analysis and artificial neural network. Tunn Undergr Sp Technol 38:161–169
Monjezi M, Ahmadi Z, Varjani AY, Khandelwal M (2013) Backbreak prediction in the Chadormalu iron mine using artificial neural network. Neural Comput Appl 23:1101–1107
Hoek E, Brown ET (1980) Underground excavations in rock. CRC Press, Boca Raton
Ebrahimi E, Monjezi M, Khalesi MR, Armaghani DJ (2016) Prediction and optimization of back-break and rock fragmentation using an artificial neural network and a bee colony algorithm. Bull Eng Geol Environ 75:27–36
Saghatforoush A, Monjezi M, Faradonbeh RS, Armaghani DJ (2016) Combination of neural network and ant colony optimization algorithms for prediction and optimization of flyrock and back-break induced by blasting. Eng Comput 32:255–266
Gates WCB, Ortiz LT, Florez RM (2005) Analysis of rockfall and blasting backbreak problems, US 550, Molas Pass, CO. Alaska Rocks 2005, 40th US Symp. Rock Mech
Esmaeili M, Osanloo M, Rashidinejad F et al (2014) Multiple regression, ANN and ANFIS models for prediction of backbreak in the open pit blasting. Eng Comput 30:549–558
Ibarra JA, Maerz NH, Franklin JA (1996) Overbreak and underbreak in underground openings part 2: causes and implications. Geotech Geol Eng 14:325–340
Roth J (1979) A model for the determination of flyrock range as a function of shot conditions. US Department of Commerce NTIS rep no PB81222358, p 61
Lundborg N (1974) The hazards of flyrock in rock blasting. Swedish Detonic Res Found Reports DS 12
Singh SP, Xavier P (2005) Causes, impact and control of overbreak in underground excavations. Tunn Undergr Sp Technol 20:63–71
Koopialipoor M, Armaghani DJ, Haghighi M, Ghaleini EN (2017) A neuro-genetic predictive model to approximate overbreak induced by drilling and blasting operation in tunnels. Bull Eng Geol Environ. https://doi.org/10.1007/s10064-017-1116-2
McCulloch WS, Pitts W (1943) A logical calculus of the ideas immanent in nervous activity. Bull Math Biophys 5:115–133
Garrett JH (1994) Where and why artificial neural networks are applicable in civil engineering. J Comput Civil Eng 8:129–130
Fausett L, Fausett L (1994) Fundamentals of neural networks: architectures, algorithms, and applications. Prentice-Hall, Upper Saddle River
Engelbrecht AP (2007) Computational intelligence: an introduction. Wiley, Hoboken
Ghaleini EN, Koopialipoor M, Momenzadeh M et al (2018) A combination of artificial bee colony and neural network for approximating the safety factor of retaining walls. Eng Comput. https://doi.org/10.1007/s00366-018-0625-3
Monjezi M, Hasanipanah M, Khandelwal M (2013) Evaluation and prediction of blast-induced ground vibration at Shur River Dam, Iran, by artificial neural network. Neural Comput Appl 22:1637–1643
Jahed Armaghani D, Hasanipanah M, Mahdiyar A et al (2016) Airblast prediction through a hybrid genetic algorithm-ANN model. Neural Comput Appl. https://doi.org/10.1007/s00521-016-2598-8
Haykin S, Network N (2004) A comprehensive foundation. Neural Netw 2:41
Jahed Armaghani D, Hajihassani M, Monjezi M et al (2015) Application of two intelligent systems in predicting environmental impacts of quarry blasting. Arab J Geosci 8:9647–9665. https://doi.org/10.1007/s12517-015-1908-2
Koopialipoor M, Fallah A, Armaghani DJ et al (2018) Three hybrid intelligent models in estimating flyrock distance resulting from blasting. Eng Comput. https://doi.org/10.1007/s00366-018-0596-4
Mojtahedi SFF, Ebtehaj I, Hasanipanah M et al (2018) Proposing a novel hybrid intelligent model for the simulation of particle size distribution resulting from blasting. Eng Comput. https://doi.org/10.1007/s00366-018-0582-x
Hasanipanah M, Bakhshandeh Amnieh H, Khamesi H et al (2018) Prediction of an environmental issue of mine blasting: an imperialistic competitive algorithm-based fuzzy system. Int J Environ Sci Technol. https://doi.org/10.1007/s13762-017-1395-y
Hasanipanah M, Faradonbeh RS, Amnieh HB et al (2016) Forecasting blast-induced ground vibration developing a CART model. Eng Comput 33:1–10
Hasanipanah M, Faradonbeh RS, Armaghani DJ et al (2017) Development of a precise model for prediction of blast-induced flyrock using regression tree technique. Environ Earth Sci 76:27
Taheri K, Hasanipanah M, Golzar SB, Majid MZA (2017) A hybrid artificial bee colony algorithm-artificial neural network for forecasting the blast-produced ground vibration. Eng Comput 33:689–700
Hornik K, Stinchcombe M, White H (1989) Multilayer feedforward networks are universal approximators. Neural Netw 2:359–366
Hecht-Nielsen R (1989) Kolmogorov’s mapping neural network existence theorem. In: Proc Int Jt Conf Neural Networks. pp 11–14
Ripley BD (1993) Statistical aspects of neural networks. Netw Chaos Stat Probab Asp 50:40–123
Paola JD (1994) Neural network classification of multispectral imagery. Master Tezi Univ, Arizona
Wang C (1994) A theory of generalization in learning machines with neural network application. Ph.D. thesis, The University of Pennsylvania, USA
Masters T (1993) Practical neural network recipes in C++. Morgan Kaufmann, Burlington
Kanellopoulos I, Wilkinson GG (1997) Strategies and best practice for neural network image classification. Int J Remote Sens 18:711–725
Kaastra I, Boyd M (1996) Designing a neural network for forecasting financial and economic time series. Neurocomputing 10:215–236
Zorlu K, Gokceoglu C, Ocakoglu F et al (2008) Prediction of uniaxial compressive strength of sandstones using petrography-based models. Eng Geol 96:141–158
Karaboga D (2005) An idea based on honey bee swarm for numerical optimization. Technical report-tr06, Erciyes university, engineering faculty, computer engineering department
Nozohour-leilabady B, Fazelabdolabadi B (2016) On the application of artificial bee colony (ABC) algorithm for optimization of well placements in fractured reservoirs; efficiency comparison with the particle swarm optimization (PSO) methodology. Petroleum 2:79–89
Ahmad A, Razali SFM, Mohamed ZS, El-shafie A (2016) The application of artificial bee colony and gravitational search algorithm in reservoir optimization. Water Resour Manag 30:2497–2516
Zhang C, Ouyang D, Ning J (2010) An artificial bee colony approach for clustering. Expert Syst Appl 37:4761–4767
Rodriguez FJ, García-Martínez C, Blum C, Lozano M (2012) An artificial bee colony algorithm for the unrelated parallel machines scheduling problem. In: International conference on parallel problem solving from nature. Springer, Berlin, pp 143–152
de Oliveira IMS, Schirru R, de Medeiros J (2009) On the performance of an artificial bee colony optimization algorithm applied to the accident diagnosis in a pwr nuclear power plant. 2009 Int. Nucl. Atl. Conf. (INAC 2009)
Karaboga D, Basturk B (2007) A powerful and efficient algorithm for numerical function optimization: artificial bee colony (ABC) algorithm. J Glob Optim 39:459–471
Gordan B, Koopialipoor M, Clementking A et al (2018) Estimating and optimizing safety factors of retaining wall through neural network and bee colony techniques. Eng Comput. https://doi.org/10.1007/s00366-018-0642-2
Acknowledgements
The authors would like to express their sincere appreciation to the anonymous reviewers for their valuable and constructive suggestions.
Author information
Authors and Affiliations
Corresponding author
Additional information
Publisher’s Note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
Rights and permissions
About this article
Cite this article
Koopialipoor, M., Ghaleini, E.N., Haghighi, M. et al. Overbreak prediction and optimization in tunnel using neural network and bee colony techniques. Engineering with Computers 35, 1191–1202 (2019). https://doi.org/10.1007/s00366-018-0658-7
Received:
Accepted:
Published:
Issue Date:
DOI: https://doi.org/10.1007/s00366-018-0658-7