Artificial Neural Networks pruning approach for geodetic velocity field determination

The European Datum 1950 (ED50) of the Turkish national geodetic network (TNGN) and the World Geodetic System 1984 (WGS84) of the Turkish national fundamental GPS network (TNFGN) are in use as geodetic reference frames in Turkey. According to the use of two reference systems, it is necessary to transform the three-dimensional (3D) coordinate data from ED50 to WGS84 or vice versa. The seven-parameter similarity transformation method is frequently used for 3D coordinate transformation in geodesy. In this study, a back propagation artificial neural network (BPANN) that has been more widely applied in engineering among all other neural network models is evaluated as an alternative 3D coordinate transformation method. BPANN is compared with a popular seven-parameter similarity transformation (Molodensky-Badekas) method over a test area, in terms of root mean square error (RMSE). The results indicated that the employment of BPANN transformed 3D coordinates (X, Y, Z) more accurate than Molodensky-Badekas method and can be useful for 3D coordinate transformation between ED50 and WGS84. Key words: 3D coordinate transformation, back propagation artificial neural network, seven-parameter similarity transformation, BPANN, Molodensky-Badekas.

Focus in quality assessment of iron ore is the content of total iron (TFe). Laser-induced breakdown spectroscopy (LIBS) technology possesses the merits of rapid, in situ, real-time multielement analysis for iron ore, but its application to quantitative TFe content is subject to interference of the iron matrix effect and the lack of suitable data mining tools. Here, a new method of LIBS-based variable importance back propagation artificial neural network (VI-BP-ANN) for quantitative TFe content in iron ore was first proposed. After the LIBS spectra of 80 representative iron samples were obtained, random forest (RF) was optimized by out-of-bag (OOB) error and then used to measure and rank variable importance. The variable importance thresholds and the number of neurons were optimized with five-fold cross-validation (CV) with correlation coefficient (R2) and root mean square error (RMSE). With using only 1.40% of full spectral variables to construct BP-ANN model, the resulted R2, the root mean squared error of prediction (RMSEP) and the modeling time of the final VI-BP-ANN model was 0.9450, 0.3174 wt%, and 24 s, respectively. Compared with full spectrum-based model, for example, BP-ANN, RF, support vector machine (SVM), and PLS and VI-RF model, the VI-BP-ANN model reduced overfitting and obtained the highest R2 and the lowest RMSE both for calibration and prediction. Meanwhile, the characteristics of variables selected by VI were analyzed. In addition to the elemental emission lines of Ca, Al, Na, K, Mn, Si, Mg, Ti, Zr, and Li, partial spectral baselines of 540-610 nm and 820-970 nm were also selected as characteristic variables, which indicated that VI can take into full consideration the elemental interactions and the spectral baselines. Our approach shows that LIBS combined with VI-BP-ANN is able to quantify TFe content rapidly and accurately in iron ore.

Temporary storage for collection of municipal solid waste is designed with various parameters, such as size of bin, collection frequency, average filling rate, number of bins, volume of material needed for bin production, bin density, and catchment area. The environmental impact of temporary storage is evaluated with emissions generated due to the production of net material used for bin production. An artificial neural network-back propagation method is proposed for modeling the temporary storage. In this proposed artificial neural network-back propagation network, the number of bins and global warming potential (GWP) based on the quantity of generated wastes are predicted. Thus, a set of different structures of artificial neural network-back propagation are investigated and then the best model is developed for temporary storage of municipal solid waste. The artificial neural network-back propagation model is trained using various parameters, such as number of neurons in the hidden layers, training tolerance, momentum, and learning rate. This model is helpful for the decision makers in choosing the environmentally sound options in the design of temporary storage for municipal solid waste management. Artificial neural network-back propagation can be successfully used to predict number of bins, GWP from waste generation rate of the city.

The study seeks to assess the efficacy of utilizing a Back Propagation Artificial Neural Network (BPANN) as an alternative to conventional methodologies for 2D coordinate transformation. Converting the previously used reference system in Iraq to the global system is one of the main requirements that play a main role in various applications requiring the standardization of geographic referencing for different data, including geospatial applications, remote sensing, image processing, and more. This necessitates researching and finding the most suitable methods to ensure high accuracy in the results. The Karbala 1979 (Clark 1880) of the Iraqi national geodetic network and the Iraqi Geospatial Reference System (IGRS) were currently used as geodetic reference frames in Iraq. In the context of transitioning between the local system (Karbala 1979) and the global system, specifically IGRS (ITRF2000 at epoch 1997.0), a crucial procedure involves the conversion of two-dimensional (2D) coordinate data. This conversion employed the BPANN method and conventional techniques such as affine and polynomial. The outcomes derived from the BPANN methodology are juxtaposed against those from the affine and polynomial methodologies. The findings reveal that the effective implementation of each approach within this study is contingent upon the spatial arrangement of the control points leveraged to develop the transformation model. Additionally, the root mean square error (RMSE) of BPANN ranging from 3 to 5 cm closely aligns with the outcomes achieved through affine and polynomial methods.

Global Navigation Satellite System (GNSS) signals arrive at the Earth in a nonlinear and slightly curved way due to the refraction effect caused by the troposphere. The troposphere delay of the GNSS signal consists of hydrostatic and wet parts. In particular, tropospheric wet delay prediction and interpolation are more difficult than those of the dry component due to the rapid temporal and spatial variation of the water vapor content. Wet delay estimation and interpolation with a sufficient accuracy is an important issue for all parameters obtained by GNSS positioning techniques. In particular, in real-time positioning applications, errors caused by interpolation of the wet troposphere delay are reflected in the height component of about 1 to 2 cm. Furthermore, the amount of water vapor in the troposphere is very important information in weather forecast applications obtained as a function of wet delay. Therefore, real-time monitoring of the troposphere can be achieved with a higher resolution and accuracy by utilizing neural network models for interpolation of the wet tropospheric delay. In addition, in the absence of the GNSS station, wet delays can be interpolated by means of the surrounding stations to the desired location. In this study, a back propagation artificial neural network (BPNN) model based on meteorological parameters obtained from The New Austrian Meteorological Measuring Network (TAWES) was used to interpolate wet troposphere delay. Analysis was carried out at 40 reference stations of the Echtzeit Positionierung Austria (EPOSA) GNSS Network covering the whole of Austria. The interpolation of zenith wet delays based on the artificial neural network was performed by using latitude, longitude, altitude and meteorological parameters (temperature, pressure, weighted mean temperature, and water vapor pressure). These parameters were then subtracted from the artificial neural network model one by one and six different artificial neural networks were designed. In addition, the linear interpolation method (LIN) and inverse distance weighted interpolation method (IDW) were used as conventional interpolation methods. In order to investigate the effect of the network density on interpolation methods, three networks, including 40, 30, and 20 reference stations, were formed and the increased distance effect on interpolation methods was evaluated. In addition, analyses were conducted in winter, spring, and summer to evaluate the seasonal effects on interpolation methods. According to the statistical analysis, the root mean square error (RMSE) values of the IDW, LIN, and BPNN methods were found to be 12.6, 13.4, and 5.9 mm, respectively.

This paper proposes a novel prediction model for photovoltaic (PV) system output current. The proposed model is based on cascade-forward back propagation artificial neural network (CFNN) with two inputs and one output. The inputs are solar radiation and ambient temperature, while the output is output current. Two years of experimental data for a 1.4 kWp PV system are utilized in this research. The monitored performance is recorded every 2 s in order to consider the uncertainty of the system’s output current. A comparison between the proposed model and other empirical and statistical models is done in this paper as well. Moreover, the ability of the proposed model to predict performance with high uncertainty rate is validated. Three statistical values are used to evaluate the accuracy of the proposed model, namely, mean absolute percentage error (MAPE), mean bias error (MBE), and root mean square error (RMSE). These values are used to measure the deviation between the actual and the predicted data in order to judge the accuracy of the proposed model. A simple estimation of the deviation between the measured value and the predicted value with respect to the measured value is first given by MAPE. After that, the average deviation of the predicted values from measured data is estimated by MBE in order to indicate the amount of the overestimation/underestimation in the predicted values. Third, the ability of predicting future records is validated by RMSE, which represents the variation of the predicted data around the measured data. Eventually, the percentage of MBE and RMSE is calculated with respect to the average value of the output current so as to present better understating of model’s accuracy. The results show that the MAPE, MBE, and RMSE of the proposed model are 7.08%, −0.21 A (−4.98%), and 0.315 A (7.5%), respectively. In addition to that, the proposed model exceeds the other models in terms of prediction accuracy.

Predictive modeling of swell-strength of expansive soils using artificial intelligence approaches: ANN, ANFIS and GEP

Solubility is a significant physical and chemical property. The solubility of carbon dioxide(CO 2 ) in polymers is an important application of green chemistry. Aimed at the problem of insufficient precision of the existing prediction model, a solubility prediction model based on the adaptive particle swarm optimization algorithm and the least‐squares support vector machine(APSO‐LSSVM) is proposed. Different from the traditional particle swarm algorithm, APSO algorithm improves the problem of easily falling into local optimal solution. The regularization parameters and kernel function tuning parameters of the LSSVM were optimized by APSO and then use this model to predict the solubility of CO 2 in eight polymers within a wide range of temperatures and pressures. APSO‐LSSVM model has great prediction ability, with a good correlation between prediction data and experimental data, high prediction accuracy, short calculation time and strong stability. Compared with back propagation artificial neural network(BP ANN), back propagation‐particle swarm optimization algorithm artificial neural network(BP‐PSO ANN) and LSSVM, APSO‐LSSVM has better comprehensive performance. The average absolute relative deviation(AARD), root mean square error(RMSE), determination coefficient(R 2 ) were respectively 0.2130, 0.0120, 0.9853. In addition, the model has good expansibility and can be applied to other fields such as chemistry and medicine.

Temperature is a predominant environmental factor affecting grass germination and distribution. Various thermal-germination models for prediction of grass seed germination have been reported, in which the relationship between temperature and germination were defined with kernel functions, such as quadratic or quintic function. However, their prediction accuracies warrant further improvements. The purpose of this study is to evaluate the relative prediction accuracies of genetic algorithm (GA) models, which are automatically parameterized with observed germination data. The seeds of five P. pratensis (Kentucky bluegrass, KB) cultivars were germinated under 36 day/night temperature regimes ranging from 5/5 to 40/40°C with 5°C increments. Results showed that optimal germination percentages of all five tested KB cultivars were observed under a fluctuating temperature regime of 20/25°C. Meanwhile, the constant temperature regimes (e.g., 5/5, 10/10, 15/15°C, etc.) suppressed the germination of all five cultivars. Furthermore, the back propagation artificial neural network (BP-ANN) algorithm was integrated to optimize temperature-germination response models from these observed germination data. It was found that integrations of GA-BP-ANN (back propagation aided genetic algorithm artificial neural network) significantly reduced the Root Mean Square Error (RMSE) values from 0.21~0.23 to 0.02~0.09. In an effort to provide a more reliable prediction of optimum sowing time for the tested KB cultivars in various regions in the country, the optimized GA-BP-ANN models were applied to map spatial and temporal germination percentages of blue grass cultivars in China. Our results demonstrate that the GA-BP-ANN model is a convenient and reliable option for constructing thermal-germination response models since it automates model parameterization and has excellent prediction accuracy.

In this study, the artificial neural networks (ANNs) and regression models were used to downscale the simulated outputs of the general circulation models (GCMs). The simulated precipitation for 25.18 º N to 34.51 º N and 45 º E to 60 º E, geopotential height at 850 mb and zonal wind at 200 mb for 12.56° N to 43.25° N and 19.68° E to 61.87° E data sets as the predictors were extracted from ECHAM5 GCM for the period 1960-2005. The observed monthly precipitation data of Abadan, Abadeh, Ahwaz, Bandar Abbas, Bushehr, Shiraz and Fasa stations as the predictand were extracted for the period 1960-2005. The principal components (PCs) of the simulated data sets were extracted and then six PCs were considered as the input file of the ANN and multiple regression models. Also the combinations of the simulated data sets were used as the input file of these models. The periods 1960-2000 and 2001-2005 were considered as the train and test data in the ANN, respectively. The Pearson correlation coefficient and normalized root mean square error results indicated that ANN predicts precipitation more accurate than multiple regression. For the monthly time scale, the geopotential height is the best predictor and for the seasonal time scale (winter) the simulated precipitation is the best predictor in ANN based standardized precipitation principal components.

The protein content is an important parameter for the estimation of the quality of cheese products. Nonlinear multivariate methods including kernel partial least squares (KPLS), support vector machine (SVM) and back propagation artificial neural network (BP-ANN) combing with the visible and near-infrared spectroscopy (Vis/NIRS) techniques are introduced for predicting the protein content of cheese. Method of augmented partial residual plot (APaRP) is performed to diagnose the nonlinearity of the raw spectral data. Multiplicative signal correction (MSC), standard normal variate (SNV) with detrending, Savitzky-Golay second-derivative and direct orthogonal signal correction (DOSC) are used to preprocess the spectral data and compared. DOSC algorithm obtains the optimal prediction performance and was chosen as the preprocess method for the following prediction models. Multivariate models including PLS, KPLS, SVM and BP-ANN are trained on the preprocessed cheese data for predicting the protein content. The square of correlation coefficient (R2), root mean squares error (RMSE) and ratio performance deviation (RPD) are used to evaluate the performance of models. DOSC-KPLS model obtains the highest scores of R2 and RPD of 0.974 and 5.587, respectively, which indicates better prediction performance than other models. Therefore, DOSC-KPLS model combined with the Vis/NIRS technique is the most promising method for accurately detecting the protein content of cheese.

In this study, the analytical calculation and analysis for the data corresponding to these gears have been made to obtain alternative test and training data sets to be used at the artificial neural networks(ANNs) with the constraints and requirements for the design of chain gears of power and motion transmission mechanisms are determined. In the input layer, the constraints and requirement values(input power, number of revolution of the pinion, number of revolution of the gear and center distance) of chain gears are used while at the output layer chain code(i.e., the chain gear type and chain sequence number), specifying the functional and physical properties, is used. Then the network is tested with the test data. The analytical calculation results and ANN predictions are compared by using statistical error analyzing methods such as absolute fraction of variance(R2), root mean square error(RMSE), and mean error percentage(MEP) for the training and test data. The chain code has been determined by the ANN with an acceptable accuracy. It is concluded that ANNs can be used as an alternative method in chain gear design. © 2009 Wiley Periodicals, Inc. Comput Appl Eng Educ 20: 38–44, 2012

Variants of feedforward backpropagation (BP) artificial neural network (ANNs) such as pseudo Mac Laurin power series order 1 (McL1BP), order 2 (McL2BP), order 3 (McL3BP) and order 4 (McL4BP) models are used to fill in mean annual streamflows. The baseline of modeling process is the standard feedforward backpropagation ANN (StandBP). The sigmoid function is used as activation function. Performance comparisons of data infilling models (ANNs) are conducted using the Root Mean Square Error of Predictions (RMSEp) and graphical plots. To test the performance of the data infilling techniques, selected streamflow gauges (i.e. the Diepkloof (control) gauge on the Wonderboomspruit River and the Molteno (target) gauge on Stormbergspruit River) of the Orange drainage river systems of South Africa have been used. The results demonstrated that relatively higher order ANN models; i.e. McL3BP, McL4BP can still outperform the rest of techniques for 7 % (except StandBP), 20 % and 30 % missing data proportions at Molteno gauge. At the same time they have been outperformed by the rest of ANN models at 13 %. Generally higher and low order Pseudo MacLaurin ANNs are acceptable to fill in missing mean annual flows at Molteno gauge. The accuracy of estimated mean annual values at Molteno is substantially negatively affected beyond 20 % gap size for all ANNs. Hence gap size beyond 20 % yields to relatively higher value of RMSEp for both high and low MacLaurin feedforward ANN techniques. A linear relationship could describe accuracy of estimated values and gap size at Molteno. Further work should include other data regimes as well as other South African catchment areas. Other activation functions should also be tested.

Preparation and artificial neural networks analysis of ultrafine β-Sialon powders by microwave-assisted carbothermal reduction nitridation of sol–gel derived powder precursors

As it is well-known, orange peel is used for making jam and oil. For this purpose, orange samples with high peel thickness are best. In order to predict peel thickness in orange fruit, we present a system based in image features, comprising: area, eccentricity, perimeter, length/area, blue value, green value, red value, wide, contrast, texture, wide/area, wide/length, roughness, and length. A novel identification solution based on the hybrid of particle swarm optimization (PSO), genetic algorithm (GA) and artificial neural network (ANN) is proposed. In addition, principal component analysis (PCA) has been applied to reduce the number of dimensions, without much loss of information. Taguchi's robust optimization technique has been applied to determine the optimal setting for parameters of PSO, GA, and ANN. The optimal level of factors were: Number of Neuron in first layer=7, Number of Neuron in second layer=2, Maximum Iteration=400, Crossover probability=0.7, Mutation probability=0.1, and Swarm (Population) Size=200. Results for prediction of orange peel thickness based on levels that are achieved by Taguchi method were evaluated by five performance measures: the coefficient of determination (R2), mean squared error (MSE), mean absolute error (MAE), sum square error (SSE), and root mean square error (RMSE), reaching values of 0.8571, 0.0123, 0.0924, 1.392, and 0.1109, respectively.