Applications of Artificial Intelligence in Mining and Geotechnical Engineering

Front Cover
Applications of Artificial Intelligence in Mining and Geotechnical Engineering
Copyright
Contents
Contributors
Editors’ biography
Preface
Chapter 1: The role of artificial intelligence in smart mining
	1. Industry 4.0 and smart mining
	2. Implementation levels of a smart mining site
	3. Role of artificial intelligence in smart mining
	4. Future perspectives
	Acknowledgments
	References
Chapter 2: Application of artificial neural networks and UAV-based air quality monitoring sensors for simulating dust emi ...
	1. Introduction
		1.1. Motivations
		1.2. Related works
		1.3. Contributions
	2. Proposed UMS-AM system
		2.1. UAV platform
		2.2. Sensor networks
			Optical sensors (Inspire 2s camera parameters)
			Air quality monitoring system
			Communication and networking subsystems
				Packaging the data from the sensor to send to the ground station
				Hardware interconnection
	3. Study site
	4. Data monitoring measurement and methodology
		4.1. Air quality monitoring measurement
		4.2. Multilayer perception neural network
	5. Results
	6. Conclusions
	Conflicts of Interest
	References
Chapter 3: Application of machine learning and metaheuristic algorithms for predicting dust emission (PM2.5) induced by d ...
	1. Introduction
	2. Methodology
		2.1. Gradient boosting machine (GBM)
		2.2. Differential evolution (DE) algorithm
		2.3. Particle swarm optimization (PSO)
		2.4. Integration of DE, PSO and GBM model
		2.5. Performance metrics for evaluation
	3. Data acquisition and preparation
	4. Results and discussion
	5. Conclusion
	Acknowledgments
	References
Chapter 4: Deep neural networks for the estimation of granite materials compressive strength using non-destructive indices
	1. Introduction
	2. Granite materials through history-A short overview
	3. Materials and methods
		3.1. Artificial neural networks
		3.2. Experimental database
		3.3. Performance indexes
	4. Results and discussion
		4.1. Splitting of database datasets
		4.2. Hyperparameters of ANN models used in this study
		4.3. Assessment of the trained and developed ANN models
		4.4. Prediction accuracy comparisons
	5. Limitations and future works
	6. Conclusions
	References
Chapter 5: Estimating the Cd2+ adsorption efficiency on nanotubular halloysites in weathered pegmatites using optimized a ...
	1. Introduction
	2. Materials description
	3. Artificial neural network
	4. Optimization algorithms used
		4.1. Slime mold algorithm
		4.2. Particle swarm optimization
		4.3. Differential evolution
	5. Framework of optimized artificial neural networks
	6. Estimation of Cd2+ adsorption efficiency of halloysite
	7. Discussion
	8. Conclusion
	Acknowledgments
	References
Chapter 6: Application of artificial intelligence in predicting slope stability in open-pit mines: A case study with a no ...
	1. Introduction
	2. Methodology
		2.1. Radial basis function neural network (RBFNN)
		2.2. Imperialist competitive algorithm (ICA)
		2.3. Proposing the ICA-RBFNN model
		2.4. Model assessment metrics
	3. Application
		3.1. Data preparation
		3.2. Model development
	4. Results and discussion
	5. Conclusion
	Acknowledgments
	References
Chapter 7: Application of cubist algorithm, multi-layer perceptron neural network, and metaheuristic algorithms to estima ...
	1. Introduction
	2. Dataset used
	3. Methodology
		3.1. Selection of input variables using the cubist algorithm
		3.2. Multi-layer perceptron neural network
		3.3. Metaheuristic algorithm for optimizing the multi-layer perceptron
			Gray wolf optimization (GWO)
			Particle swarm optimization (PSO)
			Genetic algorithm (GA)
	4. Results and discussions
	5. Conclusion
	Acknowledgments
	References
Chapter 8: Application of artificial intelligence in estimating mining capital expenditure using radial basis function ne ...
	1. Introduction
	2. Methodology
		2.1. Radial basis function neural network (RBFNN)
		2.2. Metaheuristic algorithms
			Brief principle of GA algorithm
			Brief principle of PSO algorithm
			Brief principle of MFO algorithm
			Brief principle of the HHO algorithm
		2.3. Proposing the metaheuristics-based RBFNN models for estimating CAPEX
		2.4. Performance metrics for evaluation
	3. Data preparation
	4. Results and discussions
	5. Conclusions
	Acknowledgments
	References
Chapter 9: Application of deep learning techniques for forecasting iron ore prices: A comparative study of long short-ter ...
	1. Introduction
	2. Methodology
		2.1. Long short-term memory neural network (LSTM)
		2.2. Convolutional neural network (CNN)
	3. Dataset used
	4. Results and discussion
	5. Conclusion
	Acknowledgments
	References
Chapter 10: Optimization of large mining supply chains through mathematical programming
	1. Overview
		1.1. Mining supply chain
		1.2. Optimization using mathematical programming
		1.3. Solvers
	2. Modeling challenges
	3. Mining companies applying advanced analytics
	4. Optimization model
		4.1. Indices and sets
		4.2. Parameters
		4.3. Decision variables
			Linear variables
			Integer variables
		4.4. Objective function
		4.5. Constraints
			Maximum capacity constraints
			Flow balance constraints
			On any day at any distributor, only one split option is active
			Proportional relationship between two outflows of distributors
			Relationship between linear and integer variables
			SAG stockpile capacity constraints
			Blending grade constraints
			Non-negativity constraints
	5. Case study
		5.1. Simplified operations of the equipment network
		5.2. Results and discussions
			Results and comparison
			Throughput limit-Modified mine plan
			Impacts of distributors options
			Optimizer response to planned maintenance
	6. Conclusions
	References
Chapter 11: Underground mine planning and scheduling optimization: Opportunities for embracing machine learning augmented ...
	1. Introduction
	2. Applications of machine learning in mine planning and scheduling
		2.1. Accuracy of schedule parametric inputs
		2.2. Symbiotic resemblance of model to production operations (model framing effectiveness)
		2.3. Suitability of the optimization objective function
		2.4. Dynamic capability of models to adjust to changing operating environments
	3. Conclusions
	References
Chapter 12: Application of artificial intelligence in distinguishing genuine microseismic events from the noise signals i ...
	1. Introduction
	2. Database and statistical analysis
	3. Methods
		3.1. Meta-heuristic algorithm
		3.2. Linear discriminant analysis
		3.3. Model construction
		3.4. Classification performance
	4. Summary and conclusions
	Appendix
	References
Chapter 13: The implementation of AI-based modeling and optimization in mining backfill design
	1. Introduction
	2. The use of AI in backfill design
	3. Case studies
		3.1. Predictive modeling practice
			Material
			Method
				Non-linear multiple regression modeling
				ANFIS model
			Results
		3.2. Optimization practices
			Cemented paste backfill optimization practice #1
				Materials
				Methods
					Development of objective functions using ML
					Optimization
				Results
			Cemented paste backfill optimization practice #2
				Materials
				Methods
					Development of objective functions using ML
					Optimization
				Results
	4. Conclusions
	References
Chapter 14: Application of artificial intelligence in predicting blast-induced ground vibration
	1. Introduction
	2. Case study
	3. Methodology
		3.1. Particle swarm optimization
		3.2. Backpropagation neural network
		3.3. Support vector machine
		3.4. Empirical techniques
		3.5. Development of various models
		3.6. Statistical evaluation of model performance
	4. Results and discussion
		4.1. PSO results
		4.2. BPNN and PSO-BPNN models formed
		4.3. SVM and PSO-SVM models formed
		4.4. Empirical models formed
		4.5. Comparison of all formed models for the prediction of blast-induced ground vibration
	5. Conclusion
	References
Chapter 15: Application of an expert extreme gradient boosting model to predict blast-induced air-overpressure in quarry  ...
	1. Introduction
	2. Background of case study
		2.1. Study site
		2.2. Data collection
	3. Methodology
		3.1. Extreme gradient boosting
		3.2. Bayesian optimization
		3.3. Optimized extreme gradient boosting model
	4. Results and discussion
		4.1. Evaluation criteria
		4.2. Performance of developed models
		4.3. Importance analysis
	5. Conclusions
	Acknowledgments
	References
Chapter 16: Application of artificial intelligence in predicting rock fragmentation: A review
	1. Introduction
	2. Blasting and fragmentation
	3. Blastability in traditional literature-The empirical approach
	4. Use of AI in blastability
		4.1. Artificial neural networks for predicting rock fragmentation
		4.2. Genetic algorithms for predicting rock fragmentation
		4.3. Machine learning for predicting rock fragmentation
		4.4. Hybrid approaches for predicting rock fragmentation
	5. Challenges and future directions
	6. Conclusion
	Acknowledgments
	References
Chapter 17: Underground stope dilution optimization applying machine learning
	1. Introduction
	2. Applications of machine learning in underground stope dilution optimization
		2.1. Feature range and selection
		2.2. Studies applied AI methods
	3. Conclusions
	References
Chapter 18: Applying a novel hybrid ALO-BPNN model to predict overbreak and underbreak area in underground space
	1. Introduction
	2. Methodologies
		2.1. Backpropagation neural network (BPNN)
		2.2. Ant lion optimizer (ALO)
	3. Data preparation and performance evaluation
	4. Results and discussion
		4.1. Developing a hybrid ALO-BPNN model for predicting overbreak and underbreak area
		4.2. Comparation performance of OUA prediction
		4.3. Sensitively analysis
	5. Conclusion and summary
	References
Chapter 19: Fragmentation by blasting size prediction using SVR-GOA and SVR-KHA techniques
	1. Introduction
	2. Data analysis and pre-processing
	3. Method
		3.1. Support vector regression
		3.2. Grasshopper optimization algorithm (GOA)
		3.3. Krill herd algorithm (KHA)
	4. Model development and discussion
	5. Conclusion
	References
Chapter 20: Application of machine vision in two-dimensional feature characterization of rock engineering
	1. Introduction
	2. Rock mass information acquisition method
	3. Traditional image algorithms
	4. Deep learning algorithms
		4.1. Classification and detection of lithology of rock mass
		4.2. Analysis of rock mass block and particle size
		4.3. Analysis of rock fracture
		4.4. Analysis of other rock mass parameters
	5. Conclusion
	References
Chapter 21: Groundwater potential assessment in Dobrogea region of Romania using artificial intelligence and bivariate st ...
	1. Introduction
	2. Study area
	3. Data
		3.1. Wells inventory
		3.2. Groundwater predictors
	4. Methods
		4.1. Multicollinearity assessment
		4.2. Weights of evidence (WOE)
		4.3. Support vector machine (SVM)
		4.4. ROC curve for validation
	5. Results and discussion
		5.1. Multicollinearity assessment
		5.2. Weights of evidence
		5.3. Groundwater potential
		5.4. Results validation
	6. Conclusions
	References
Chapter 22: Application of artificial intelligence techniques for the verification of pile capacity at construction site: ...
	1. Introduction
	2. Background of soft computing
		2.1. Artificial neural network (ANN)
		2.2. Support vector machine (SVM)
		2.3. Decision tree (DT)
		2.4. Genetic programming and gene expression programming (GP & GEP)
	3. Application of AI for pile capacity prediction
		3.1. Base artificial intelligence (AI) models
		3.2. Hybrid AI models
	4. Discussion
	5. Future perspective
	6. Conclusion
	References
Chapter 23: Landslide susceptibility in a hilly region of Romania using artificial intelligence and bivariate statistics
	1. Introduction
	2. Study area
	3. Data
		3.1. Landslide inventory
		3.2. Landslide predictors
	4. Methods
		4.1. Frequency ratio (FR)
		4.2. Multilayer perceptron
		4.3. ROC curve for results validation
	5. Results and discussions
		5.1. Frequency ratio (FR) analysis
		5.2. Landslide susceptibility mapping
		5.3. Results validation
	6. Conclusions
	References
Chapter 24: Spatial prediction of bridge displacement using deep learning models: A case study at Co Luy bridge
	1. Introduction
	2. Study area and data used
		2.1. Co Luy bridge
		2.2. Data used
	3. Methods
		3.1. Long short-term memory (LSTM)
		3.2. Gated recurrent unit (GRU)
		3.3. Proposed deep learning models
		3.4. Setting parameters
		3.5. Forecasting performance metrics
	4. Results and analysis
	5. Discussions
	6. Conclusions
	References
Index
Back Cover