Using deep neural networks coupled with principal component analysis for ore production forecasting at open-pit mines

Ore production is usually affected by multiple influencing inputs at open-pit mines.Nevertheless,the complex nonlinear relationships between these inputs and ore production remain unclear.This becomes even more challenging when training data(e.g.truck haulage information and weather conditions)are massive.In machine learning(ML)algorithms,deep neural network(DNN)is a superior method for processing nonlinear and massive data by adjusting the amount of neurons and hidden layers.This study adopted DNN to forecast ore production using truck haulage information and weather conditions at open-pit mines as training data.Before the prediction models were built,principal component analysis(PCA)was employed to reduce the data dimensionality and eliminate the multicollinearity among highly correlated input variables.To verify the superiority of DNN,three ANNs containing only one hidden layer and six traditional ML models were established as benchmark models.The DNN model with multiple hidden layers performed better than the ANN models with a single hidden layer.The DNN model outperformed the extensively applied benchmark models in predicting ore production.This can provide engineers and researchers with an accurate method to forecast ore production,which helps make sound budgetary decisions and mine planning at open-pit mines.

Predicting the alloying element yield in a ladle furnace using principal component analysis and deep neural network

The composition control of molten steel is one of the main functions in the ladle furnace(LF)refining process.In this study,a feasible model was established to predict the alloying element yield using principal component analysis(PCA)and deep neural network(DNN).The PCA was used to eliminate collinearity and reduce the dimension of the input variables,and then the data processed by PCA were used to establish the DNN model.The prediction hit ratios for the Si element yield in the error ranges of±1%,±3%,and±5%are 54.0%,93.8%,and98.8%,respectively,whereas those of the Mn element yield in the error ranges of±1%,±2%,and±3%are 77.0%,96.3%,and 99.5%,respectively,in the PCA-DNN model.The results demonstrate that the PCA-DNN model performs better than the known models,such as the reference heat method,multiple linear regression,modified backpropagation,and DNN model.Meanwhile,the accurate prediction of the alloying element yield can greatly contribute to realizing a“narrow window”control of composition in molten steel.The construction of the prediction model for the element yield can also provide a reference for the development of an alloying control model in LF intelligent refining in the modern iron and steel industry.

Application of principal component-radial basis function neural networks (PC-RBFNN) for the detection of water-adulterated bayberry juice by near-infrared spectroscopy

Near-infrared (NIR) spectroscopy combined with chemometrics techniques was used to classify the pure bayberry juice and the one adulterated with 10% (w/w) and 20% (w/w) water. Principal component analysis (PCA) was applied to reduce the dimensions of spectral data, give information regarding a potential capability of separation of objects, and provide principal component (PC) scores for radial basis function neural networks (RBFNN). RBFNN was used to detect bayberry juice adulterant. Multiplicative scatter correction (MSC) and standard normal variate (SNV) transformation were used to preprocess spectra. The results demonstrate that PC-RBFNN with optimum parameters can separate pure bayberry juice samples from water-adulterated bayberry at a recognition rate of 97.62%, but cannot clearly detect water levels in the adulterated bayberry juice. We conclude that NIR technology can be successfully applied to detect water-adulterated bayberry juice.

A Hybrid Neural Network Model for ENSO Prediction in Combination with Principal Oscillation Pattern Analyses

El Niño-Southern Oscillation(ENSO)can be currently predicted reasonably well six months and longer,but large biases and uncertainties remain in its real-time prediction.Various approaches have been taken to improve understanding of ENSO processes,and different models for ENSO predictions have been developed,including linear statistical models based on principal oscillation pattern(POP)analyses,convolutional neural networks(CNNs),and so on.Here,we develop a novel hybrid model,named as POP-Net,by combining the POP analysis procedure with CNN-long short-term memory(LSTM)algorithm to predict the Niño-3.4 sea surface temperature(SST)index.ENSO predictions are compared with each other from the corresponding three models:POP model,CNN-LSTM model,and POP-Net,respectively.The POP-based pre-processing acts to enhance ENSO-related signals of interest while filtering unrelated noise.Consequently,an improved prediction is achieved in the POP-Net relative to others.The POP-Net shows a high-correlation skill for 17-month lead time prediction(correlation coefficients exceeding 0.5)during the 1994-2020 validation period.The POP-Net also alleviates the spring predictability barrier(SPB).It is concluded that value-added artificial neural networks for improved ENSO predictions are possible by including the process-oriented analyses to enhance signal representations.

Electricity price forecasting using generalized regression neural network based on principal components analysis

A combined model based on principal components analysis (PCA) and generalized regression neural network (GRNN) was adopted to forecast electricity price in day-ahead electricity market. PCA was applied to mine the main influence on day-ahead price, avoiding the strong correlation between the input factors that might influence electricity price, such as the load of the forecasting hour, other history loads and prices, weather and temperature; then GRNN was employed to forecast electricity price according to the main information extracted by PCA. To prove the efficiency of the combined model, a case from PJM (Pennsylvania-New Jersey-Maryland) day-ahead electricity market was evaluated. Compared to back-propagation (BP) neural network and standard GRNN, the combined method reduces the mean absolute percentage error about 3%.

Remaining Useful Life Prediction of Aeroengine Based on Principal Component Analysis and One-Dimensional Convolutional Neural Network

In order to directly construct the mapping between multiple state parameters and remaining useful life(RUL),and reduce the interference of random error on prediction accuracy,a RUL prediction model of aeroengine based on principal component analysis(PCA)and one-dimensional convolution neural network(1D-CNN)is proposed in this paper.Firstly,multiple state parameters corresponding to massive cycles of aeroengine are collected and brought into PCA for dimensionality reduction,and principal components are extracted for further time series prediction.Secondly,the 1D-CNN model is constructed to directly study the mapping between principal components and RUL.Multiple convolution and pooling operations are applied for deep feature extraction,and the end-to-end RUL prediction of aeroengine can be realized.Experimental results show that the most effective principal component from the multiple state parameters can be obtained by PCA,and the long time series of multiple state parameters can be directly mapped to RUL by 1D-CNN,so as to improve the efficiency and accuracy of RUL prediction.Compared with other traditional models,the proposed method also has lower prediction error and better robustness.

New artificial neural networks for true triaxial stress state analysis and demonstration of intermediate principal stress effects on intact rock strength

Simulations are conducted using five new artificial neural networks developed herein to demonstrate and investigate the behavior of rock material under polyaxial loading. The effects of the intermediate principal stress on the intact rock strength are investigated and compared with laboratory results from the literature. To normalize differences in laboratory testing conditions, the stress state is used as the objective parameter in the artificial neural network model predictions. The variations of major principal stress of rock material with intermediate principal stress, minor principal stress and stress state are investigated. The artificial neural network simulations show that for the rock types examined, none were independent of intermediate principal stress effects. In addition, the results of the artificial neural network models, in general agreement with observations made by others, show （a） a general trend of strength increasing and reaching a peak at some intermediate stress state factor, followed by a decline in strength for most rock types; （b） a post-peak strength behavior dependent on the minor principal stress, with respect to rock type; （c） sensitivity to the stress state, and to the interaction between the stress state and uniaxial compressive strength of the test data by the artificial neural networks models （two-way analysis of variance; 95% confidence interval）. Artificial neural network modeling, a self-learning approach to polyaxial stress simulation, can thus complement the commonly observed difficult task of conducting true triaxial laboratory tests, and/or other methods that attempt to improve two-dimensional （2D） failure criteria by incorporating intermediate principal stress effects.

Combining Principal Component Regression and Artificial Neural Network to Predict Chlorophyll-a Concentration of Yuqiao Reservoir’s Outflow

In order to investigate the eutrophication degree of Yuqiao Reservoir, a hybrid method, combining principal component regression (PCR) and artificial neural network (ANN), was adopted to predict chlorophyll-a concentration of Yuqiao Reservoir’s outflow. The data were obtained from two sampling sites, site 1 in the reservoir, and site 2 near the dam. Seven water variables, namely chlorophyll-a concentration of site 2 at time t and that of both sites 10 days before t, total phosphorus(TP), total nitrogen(TN),...

Predict typhoon-induced storm surge deviation in a principal component back-propagation neural network model

To reduce typhoon-caused damages, numerical and empirical methods are often used to forecast typhoon storm surge. However, typhoon surge is a complex nonlinear process that is difficult to forecast accurately. We applied a principal component back-propagation neural network (PCBPNN) to predict the deviation in typhoon storm surge, in which data of the typhoon, upstream flood, and historical case studies were involved. With principal component analysis, 15 input factors were reduced to five principal components, and the application of the model was improved. Observation data from Huangpu Park in Shanghai, China were used to test the feasibility of the model. The results indicate that the model is capable of predicting a 12-hour warning before a typhoon surge.

Coupled Cross-correlation Neural Network Algorithm for Principal Singular Triplet Extraction of a Cross-covariance Matrix

This paper proposes a novel coupled neural network learning algorithm to extract the principal singular triplet (PST) of a cross-correlation matrix between two high-dimensional data streams. We firstly introduce a novel information criterion (NIC), in which the stationary points are singular triplet of the crosscorrelation matrix. Then, based on Newton's method, we obtain a coupled system of ordinary differential equations (ODEs) from the NIC. The ODEs have the same equilibria as the gradient of NIC, however, only the first PST of the system is stable (which is also the desired solution), and all others are (unstable) saddle points. Based on the system, we finally obtain a fast and stable algorithm for PST extraction. The proposed algorithm can solve the speed-stability problem that plagues most noncoupled learning rules. Moreover, the proposed algorithm can also be used to extract multiple PSTs effectively by using sequential method. © 2014 Chinese Association of Automation.

Identification of Mine Water Inrush Source Based on PCA-BP Neural Network

It is of great significance to analyze the chemical indexes of mine water and develop a rapid identification system of water source, which can quickly and accurately distinguish the causes of water inrush and identify the source of water inrush, so as to reduce casualties and economic losses and prevent and control water inrush disasters. Taking Ca2+, Mg2+, Na+ + K+, , , Cl-, pH value and TDS as discriminant indexes, the principal component analysis method was used to reduce the dimension of data, and the identification model of mine water inrush source based on PCA-BP neural network was established. 96 sets of data of different aquifers in Panxie mining area were selected for prediction analysis, and 20 sets of randomly selected data were tested, with an accuracy rate of 95%. The model can effectively reduce data redundancy, has a high recognition rate, and can accurately and quickly identify the water source of mine water inrush.

PC-based artif icial neural network inversion for airborne time-domain electromagnetic data

Traditionally, airborne time-domain electromagnetic （ATEM） data are inverted to derive the earth model by iteration. However, the data are often highly correlated among channels and consequently cause ill-posed and over-determined problems in the inversion. The correlation complicates the mapping relation between the ATEM data and the earth parameters and thus increases the inversion complexity. To obviate this, we adopt principal component analysis to transform ATEM data into orthogonal principal components （PCs） to reduce the correlations and the data dimensionality and simultaneously suppress the unrelated noise. In this paper, we use an artificial neural network （ANN） to approach the PCs mapping relation with the earth model parameters, avoiding the calculation of Jacobian derivatives. The PC-based ANN algorithm is applied to synthetic data for layered models compared with data-based ANN for airborne time-domain electromagnetic inversion. The results demonstrate the PC-based ANN advantages of simpler network structure, less training steps, and better inversion results over data-based ANN, especially for contaminated data. Furthermore, the PC-based ANN algorithm effectiveness is examined by the inversion of the pseudo 2D model and comparison with data-based ANN and Zhody＇s methods. The results indicate that PC-based ANN inversion can achieve a better agreement with the true model and also proved that PC-based ANN is feasible to invert large ATEM datasets.

Feature evaluation and extraction based on neural network in analog circuit fault diagnosis

Choosing the right characteristic parameter is the key to fault diagnosis in analog circuit. The feature evaluation and extraction methods based on neural network are presented. Parameter evaluation of circuit features is realized by training results from neural network; the superior nonlinear mapping capability is competent for extracting fault features which are normalized and compressed subsequently. The complex classification problem on fault pattern recognition in analog circuit is transferred into feature processing stage by feature extraction based on neural network effectively, which improves the diagnosis efficiency. A fault diagnosis illustration validated this method.

Artificial neural network model for identifying taxi gross emitter from remote sensing data of vehicle emission

Vehicle emission has been the major source of air pollution in urban areas in the past two decades. This article proposes an artificial neural network model for identifying the taxi gross emitters based on the remote sensing data. After carrying out the field test in Guangzhou and analyzing various factors from the emission data, the artificial neural network modeling was proved to be an advisable method of identifying the gross emitters. On the basis of the principal component analysis and the selection of algorithm and architecture, the Back-Propagation neural network model with 8-17-1 architecture was established as the optimal approach for this purpose. It gave a percentage of hits of 93%. Our previous research result and the result from aggression analysis were compared, and they provided respectively the percentage of hits of 81.63% and 75%. This comparison demonstrates the potentiality and validity of the proposed method in the identification of taxi gross emitters.

FAST RECURSIVE LEAST SQUARES LEARNING ALGORITHM FOR PRINCIPAL COMPONENT ANALYSIS

Based on the least-square minimization a computationally efficient learning algorithm for the Principal Component Analysis(PCA) is derived. The dual learning rate parameters are adaptively introduced to make the proposed algorithm providing the capability of the fast convergence and high accuracy for extracting all the principal components. It is shown that all the information needed for PCA can be completely represented by the unnormalized weight vector which is updated based only on the corresponding neuron input-output product. The convergence performance of the proposed algorithm is briefly analyzed.The relation between Oja’s rule and the least squares learning rule is also established. Finally, a simulation example is given to illustrate the effectiveness of this algorithm for PCA.

Dimensionality Reduction with Input Training Neural Network and Its Application in Chemical Process Modelling

Many applications of principal component analysis (PCA) can be found in dimensionality reduction. But linear PCA method is not well suitable for nonlinear chemical processes. A new PCA method based on im-proved input training neural network (IT-NN) is proposed for the nonlinear system modelling in this paper. Mo-mentum factor and adaptive learning rate are introduced into learning algorithm to improve the training speed of IT-NN. Contrasting to the auto-associative neural network (ANN), IT-NN has less hidden layers and higher training speed. The effectiveness is illustrated through a comparison of IT-NN with linear PCA and ANN with experiments. Moreover, the IT-NN is combined with RBF neural network (RBF-NN) to model the yields of ethylene and propyl-ene in the naphtha pyrolysis system. From the illustrative example and practical application, IT-NN combined with RBF-NN is an effective method of nonlinear chemical process modelling.

Fault prediction method for nuclear power machinery based on Bayesian PPCA recurrent neural network model

Early fault warning for nuclear power machinery is conducive to timely troubleshooting and reductions in safety risks and unnecessary costs. This paper presents a novel intelligent fault prediction method, integrated probabilistic principal component analysis(PPCA), multi-resolution wavelet analysis, Bayesian inference, and RNN model for nuclear power machinery that consider data uncertainty and chaotic time series. After denoising the source data, the Bayesian PPCA method is employed for dimensional reduction to obtain a refined data group. A recurrent neural network(RNN) prediction model is constructed, and a Bayesian statistical inference approach is developed to quantitatively assess the prediction reliability of the model. By modeling and analyzing the data collected on the steam turbine and components of a nuclear power plant, the results of the goodness of fit, mean square error distribution, and Bayesian confidence indicate that the proposed RNN model can implement early warning in the fault creep period. The accuracy and reliability of the proposed model are quantitatively verified.

Wet Gas Meter Development Based on Slotted Orifice Couple and Neural Network Techniques

A slotted orifice has many superiorities over a standard orifice. For single-phase flow measurement, its flow coefficient is insensitive to the upstream velocity profile. For two phase flow measurement, various characteristics of its differential pressure （DP） are stable and closely correlated with the mass flow rate of gas and liquid. The complex relationships between the signal features and the two-phase flow rate are established through the use of a back propagation （BP） neural network. Experiments were carried out in the horizontal tubes with 50ram inner diameter, ooerated with water flow rate in the range of 0.2m^3·h^-1 to 4m3·h^-1, gas flow rate in the range of 100m^3·h^-1 to 1000m^3·h^-1, and pressure at 400kPa and 850kPa respectively, where the temperature is ambient temperature. This article includes the principle of wet gas meter development, the experimental matrix, the signal processing techniques and the achieved results. On the basis of the results it is suggested that the slotted orifice couple with a trained neural network may provide a simple but efficient solution to the wet gas meter development.

Application of artificial neural network for calculating anisotropic friction angle of sands and effect on slope stability

The anisotropy effect is one of the most prominent phenomena in soil mechanics. Although many experimental programs have investigated anisotropy in sand, a computational procedure for determining anisotropy is lacking. Thus, this work aims to develop a procedure for connecting the sand friction angle and the loading orientation. All principal stress rotation tests in the literatures were processed via an artificial neural network. Then, with sensitivity analysis, the effect of intrinsic soil properties,consolidation history, and test sample characteristics on enhancing anisotropy was examined. The results imply that decreasing the grain size of the soil increases the effect of anisotropy on soil shear strength. In addition, increasing the angularity of grains increases the anisotropy effect in the sample. The stability of a sandy slope was also examined by considering the anisotropy in shear strength parameters. If the anisotropy effect is neglected, slope safety is overestimated by 5%-25%. This deviation is more apparent in flatter slopes than in steeper ones. However, the critical slip surface in the most slopes is the same in isotropic and anisotropic conditions.

Reconstruction based approach to sensor fault diagnosis using auto-associative neural networks

Fault diagnostics is an important research area including different techniques.Principal component analysis(PCA)is a linear technique which has been widely used.For nonlinear processes,however,the nonlinear principal component analysis(NLPCA)should be applied.In this work,NLPCA based on auto-associative neural network(AANN)was applied to model a chemical process using historical data.First,the residuals generated by the AANN were used for fault detection and then a reconstruction based approach called enhanced AANN(E-AANN)was presented to isolate and reconstruct the faulty sensor simultaneously.The proposed method was implemented on a continuous stirred tank heater(CSTH)and used to detect and isolate two types of faults(drift and offset)for a sensor.The results show that the proposed method can detect,isolate and reconstruct the occurred fault properly.