Three-layer Back-propagation Artificial Neural Network Research Articles

Optimization of 3D Printing Parameters on Deformation by BP Neural Network Algorithm

Traditional processing technology is not suitable for the requirements of advanced manufacturing due to the disadvantages of large repeated experiments, high cost, and low economic effect. As the latest additive technology, 3D printing technology has to deal with many issues such as process parameters and nonlinear mathematical models. A three-layer backpropagation (BP) artificial neural network with a Lavenberg–Marquardt algorithm was established to train the network and predict orthogonal experimental data. Additionally, the best combination of parameters of material deformations were predicted and verified by experiments. The results show that the predicted value obtained by the BP model is in good agreement with the experimental value curve, with a small relative error and a correlation coefficient of 0.99985. Moreover, the deformation errors of the printed model are not more than 3%. The incorporation of the BP neural network algorithm into the 3D printing process can, therefore, help cope with related problems, which is a future trend.

Hot tensile deformation behaviors of TA32 titanium alloy based on back-propagation neural networks and three-dimensional thermal processing maps

The hot deformation behavior of TA32 titanium alloy in the temperature range of 650–850 °C and strain rate range of 0.0001–0.1 s−1 was studied by isothermal hot tensile tests conducted on an electronic universal testing machine equipped with an environmental chamber. The effects of temperature and strain rate on the flow stress of TA32 were investigated. Moreover, a back-propagation artificial neural network (BP-ANN) model was developed to predict the flow stress at elevated temperatures and three-dimensional (3-d) thermal processing maps. The results show that the three-layer BP-ANN model can precisely predict the flow stress as the correlation coefficient reaches 0.99365 and the mean relative error is 1.15621%. The 3-d thermal processing maps demonstrate that the lower strain rate, higher temperature, and larger strain lead to higher power dissipation efficiency and better formability. And it shows a flow instability tendency at lower temperatures and higher strain rates. Based on the 3-d thermal processing maps and microstructure analysis, the optimized temperature range and strain rate range for TA32 hot forming were 770–830 °C and 0.0001–0.00015 s−1.

Application of BP Neural Networks on the Thickness Prediction of Sherardizing Coating

Sherardizing is a surface protection method for obtaining Fe–Zn coating on the steel surface by thermal diffusion. The thickness of coating is critical for its performance. Therefore, it is necessary to propose a model to predict coating thickness based on different process parameters. The model for the relationship between sherardizing process parameters and the total thickness and each layer phase thickness of the coating is built using a three-layer backpropagation (BP) artificial neural network based on Levenberg–Marquardt algorithm. The results show that neural network model has a good predictive effect on the total thickness and each layer phase thickness of Fe–Zn coating. The prediction error of the model on the thickness of the coating is within 4.71%, and the coefficient of determination is 0.99974. It presents a new method for prediction of the thickness of sherardizing coating.

Prediction of fuel cell performance based on Bagging neural network ensemble model

The influence of the inlet and outlet heights of the flow channel on the performance of a proton exchange membrane fuel cell was studied with a three-dimensional fuel cell mathematical model. The numerical results were used as the basic training data of a three-layer back-propagation artificial neural network (ANN) model in which the inlet and outlet heights of the flow channel as well as the cell voltage were set to be the input variables while cell current density is the output variable. Then, applying the Bagging ensemble learning method to integrate the neural network model, a fuel cell performance prediction method is constructed. It was shown that as compared with the ANN model, the Bagging neural network model exhibited higher prediction accuracy and less requirements of model training data. It also can be applied for rapid prediction of fuel cell performance in a wide range of conditions beyond the training data, showing the robustness of the proposed Bagging neural network ensemble model.

Impeller fault detection under variable flow conditions based on three feature extraction methods and artificial neural networks

Nonstationary flow conditions can introduce complexities and nonlinear characteristics to pumping systems. This paper presents comparative studies of impeller fault detection techniques combined with artificial neural networks (ANNs) to propose the most appropriate diagnosis system. An experimental study, including seven impeller conditions, is performed to further explore the phenomena. Statistical parameters, frequency peaks, and wavelet packet energy present data feature sets, and a three-layer back-propagation ANN is used for fault recognition. The verification of the results proves that the detectability of the wavelet packet transform (WPT)-ANN model is considerably improved by using the energy of the decomposed vibration from WPT. This model can save computational time and provide superior diagnostic information. This study provides two key contributions. First, the feasibility and effectiveness of common monitoring techniques are compared. Second, the results demonstrate the accuracy of the proposed models for impellers operating under variable working conditions, which has not been previously addressed in the literature.

Forecasting outpatient visits using empirical mode decomposition coupled with back-propagation artificial neural networks optimized by particle swarm optimization.

Accurately predicting the trend of outpatient visits by mathematical modeling can help policy makers manage hospitals effectively, reasonably organize schedules for human resources and finances, and appropriately distribute hospital material resources. In this study, a hybrid method based on empirical mode decomposition and back-propagation artificial neural networks optimized by particle swarm optimization is developed to forecast outpatient visits on the basis of monthly numbers. The data outpatient visits are retrieved from January 2005 to December 2013 and first obtained as the original time series. Second, the original time series is decomposed into a finite and often small number of intrinsic mode functions by the empirical mode decomposition technique. Third, a three-layer back-propagation artificial neural network is constructed to forecast each intrinsic mode functions. To improve network performance and avoid falling into a local minimum, particle swarm optimization is employed to optimize the weights and thresholds of back-propagation artificial neural networks. Finally, the superposition of forecasting results of the intrinsic mode functions is regarded as the ultimate forecasting value. Simulation indicates that the proposed method attains a better performance index than the other four methods.

Multivariate analysis of hyper/multi-spectra for determining volatile compounds and visualizing cooking degree during low-temperature baking of tubers

This study was conducted to assess the potential feasibility of using hyperspectral imaging (900–1700nm) for rapid determination of the volatility of tuber compositions (VTC) and prediction of the tuber cooking degree (TCD) in low temperature baking (LTB). Tuber samples of six categories from different origins were imagedand calibrated. The partial least squares regression (PLSR) and three-layer back propagation artificial neural network (TBPANN) models were established to predict VTC and TCD using the entire spectral range and the feature wavelengths. The optimal combination of characteristic wavelengths (991, 1031, 1071, 1138, 1252, 1403, 1460 and 1641nm) were selected by first derivative and mean centering iteration algorithm (FMCIA) rather than other conventional methods. Based on the qualified eight wavelengths, the FMCIA-TBPANN approachyieldedgreater overallperformance for predicting both VTC and TCD.Furthermore, the distribution maps of VTC and TCD were generated using a resulting function to visualize each pixel of spectral image. This demonstrated the capability of spectral imaging technique for rapid and accurate evaluation of VTC and TCD during LTB.

Modeling the Effects of Cu Content and Deformation Variables on the High-Temperature Flow Behavior of Dilute Al-Fe-Si Alloys Using an Artificial Neural Network.

The hot deformation behavior of Al-0.12Fe-0.1Si alloys with varied amounts of Cu (0.002–0.31 wt %) was investigated by uniaxial compression tests conducted at different temperatures (400 °C–550 °C) and strain rates (0.01–10 s−1). The results demonstrated that flow stress decreased with increasing deformation temperature and decreasing strain rate, while flow stress increased with increasing Cu content for all deformation conditions studied due to the solute drag effect. Based on the experimental data, an artificial neural network (ANN) model was developed to study the relationship between chemical composition, deformation variables and high-temperature flow behavior. A three-layer feed-forward back-propagation artificial neural network with 20 neurons in a hidden layer was established in this study. The input parameters were Cu content, temperature, strain rate and strain, while the flow stress was the output. The performance of the proposed model was evaluated using the K-fold cross-validation method. The results showed excellent generalization capability of the developed model. Sensitivity analysis indicated that the strain rate is the most important parameter, while the Cu content exhibited a modest but significant influence on the flow stress.

Nonoxidative Conversion of Methane in a Dielectric Barrier Discharge Reactor: Prediction of Reaction Performance Based on Neural Network Model

The nonoxidative conversion of CH4 into H2 and higher hydrocarbons has been performed in a coaxial dielectric barrier discharge reactor at atmospheric pressure and low temperatures. The effect of discharge power, gas flow rate, and excitation frequency on the reaction performance of the plasma methane conversion is investigated. A three-layer back-propagation artificial neural network (ANN) model has been developed and trained to simulate and predict the complex plasma chemical reaction in terms of the conversion of CH4, the selectivity and yield of gas products, and the energy efficiency of the plasma process. A good agreement between the experimental and simulated results is achieved. The ANN model shows that the maximum CH4 conversion of 36% can be obtained at a discharge power of 75 W with a high selectivity of C2H6 (42.4%). In this study, the discharge power is found to be the most influential parameter with a relative weight of 45–52% for the plasma nonoxidative coupling of methane, while the excitation frequency of the plasma system is the least important parameter affecting the plasma process. The results successfully demonstrate that the well-trained ANN model can accurately simulate and predict a complex plasma chemical reaction.

一种基于光纤布喇格光栅振动传感器的光纤围栏入侵监测系统及其模式识别

A fiber-optical perimeter intrusion detection system based on high-sensitivity FBG vibrating sensors with intelligent identification function was proposed.Stasitical features with an adaptive dynamic threshold were extracted in the time domain first.By inputting the extracted features into a three-layer Back-Propagation Artificial Neural Network classifier,the type of the event can be descerned and determined.The simulated signals and acquired data in the test field were both tested to validate the effectiveness of the proposed system and its identification method.The results show that the average identification rate of the system for the simulation signals can be up to 100%,and that for the actually acquired signals,the average identification rate of the system achieves 96.83%.

A hybrid artificial neural network—mechanistic model for centrifugal compressor

A mathematical model is an important tool for design and optimization of centrifugal compressor. However, owing to the varying compressor speeds and the complexity of the flow dynamics inside the impeller and diffuser, the currently available mechanistic models may yield inaccurate results. The purpose of this paper is to present a hybrid modeling approach for developing a quantitatively accurate model for centrifugal compressor. Two novel hybrid models, that is, additive and multiplicative hybrid models each of which consists of a three-layer back-propagation artificial neural network (ANN) component and a mechanistic component suitably modified to describe the performances of multistage centrifugal compressor, were constructed and compared with the well-developed ANN model. The results from the hybrid models showed better performance compared to the ANN model. Besides, the hybrid models demonstrated much better performance than the pure mechanistic model, and the multiplicative hybrid model, in general, showed better accuracy than that of the additive hybrid model in our case.

Colorimetric Grading Scale Can Promote the Standardization of Experiential and Sensory Evaluation in Quality Control of Traditional Chinese Medicines

Experiential and sensory evaluation is an ancient method that remains important in the current quality control system of Traditional Chinese Medicines (TCMs). The process is rapid and convenient when evaluating the quality of crude materials in TCM markets. However, sensory evaluation has been met with skepticism because it is mainly based on experience and lacks a scientific basis. In this study, rhubarb was selected to demonstrate how color-based sensory evaluation could differentiate the quality of herbal medicines objectively. The colors of the rhubarb samples, expressed as RGB values, were obtained from different parts and forms of the plant, including the plant’s surface, fracture surface color, and a powdered form with or without treatment with a color-developing reagent. We first divided the rhubarb samples into three grades based on the total content of five hydroxyanthraquinone derivatives, the major pharmacological components in rhubarb. Then, a three-layer back-propagation artificial neural network (BP-ANN), calibrated with selected training samples, was used to correlate the quality of the rhubarb with its color. The color of the rhubarb powder after coloration attained the highest accuracy (92.3%) in predicting the quality grade of the test samples with the established artificial neural networks. Finally, a standardized colorimetric grading scale was created based on the spatial distribution of the rhubarb samples in a two-dimensional chromaticity diagram according to the colors of the powdered rhubarb after color enhancement. By comparing the color between the scale and the tested samples, similar to performing a pH test with indicator paper, subjects without sensory evaluation experience could quickly determine the quality grade of rhubarb. This work illustrates the technical feasibility of the color-based grading of rhubarb quality and offers references for quantifying and standardizing the sensory evaluation of TCMs, foods and other products.

Artificial neural network modeling of dissolved oxygen in the Heihe River, Northwestern China

Identification and quantification of dissolved oxygen (DO) profiles of river is one of the primary concerns for water resources managers. In this research, an artificial neural network (ANN) was developed to simulate the DO concentrations in the Heihe River, Northwestern China. A three-layer back-propagation ANN was used with the Bayesian regularization training algorithm. The input variables of the neural network were pH, electrical conductivity, chloride (Cl(-)), calcium (Ca(2+)), total alkalinity, total hardness, nitrate nitrogen (NO3-N), and ammonical nitrogen (NH4-N). The ANN structure with 14 hidden neurons obtained the best selection. By making comparison between the results of the ANN model and the measured data on the basis of correlation coefficient (r) and root mean square error (RMSE), a good model-fitting DO values indicated the effectiveness of neural network model. It is found that the coefficient of correlation (r) values for the training, validation, and test sets were 0.9654, 0.9841, and 0.9680, respectively, and the respective values of RMSE for the training, validation, and test sets were 0.4272, 0.3667, and 0.4570, respectively. Sensitivity analysis was used to determine the influence of input variables on the dependent variable. The most effective inputs were determined as pH, NO3-N, NH4-N, and Ca(2+). Cl(-) was found to be least effective variables on the proposed model. The identified ANN model can be used to simulate the water quality parameters.

Simulation of phytoplankton biomass in Quanzhou Bay using a back propagation network model and sensitivity analysis for environmental variables

Prediction and sensitivity models, to elucidate the response of phytoplankton biomass to environmental factors in Quanzhou Bay, Fujian, China, were developed using a back propagation (BP) network. The environmental indicators of coastal phytoplankton biomass were determined and monitoring data for the bay from 2008 was used to train, test and build a three-layer BP artificial neural network with multi-input and single-output. Ten water quality parameters were used to forecast phytoplankton biomass (measured as chlorophyll-a concentration). Correlation coefficient between biomass values predicted by the model and those observed was 0.964, whilst the average relative error of the network was −3.46% and average absolute error was 10.53%. The model thus has high level of accuracy and is suitable for analysis of the influence of aquatic environmental factors on phytoplankton biomass. A global sensitivity analysis was performed to determine the influence of different environmental indicators on phytoplankton biomass. Indicators were classified according to the sensitivity of response and its risk degree. The results indicate that the parameters most relevant to phytoplankton biomass are estuary-related and include pH, sea surface temperature, sea surface salinity, chemical oxygen demand and ammonium.

Study on nondestructive discrimination of genuine and counterfeit wild ginsengs using NIRS

A new approach for the nondestructive discrimination between genuine wild ginsengs and the counterfeit ones by near infrared spectroscopy (NIRS) was developed. Both discriminant analysis and back propagation artificial neural network (BP-ANN) were applied to the model establishment for discrimination. Optimal modeling wavelengths were determined based on the anomalous spectral information of counterfeit samples. Through principal component analysis (PCA) of various wild ginseng samples, genuine and counterfeit, the cumulative percentages of variance of the principal components were obtained, serving as a reference for principal component (PC) factor determination. Discriminant analysis achieved an identification ratio of 88.46%. With samples' truth values as its outputs, a three-layer BP-ANN model was built, which yielded a higher discrimination accuracy of 100%. The overall results sufficiently demonstrate that NIRS combined with BP-ANN classification algorithm performs better on ginseng discrimination than discriminant analysis, and can be used as a rapid and nondestructive method for the detection of counterfeit wild ginsengs in food and pharmaceutical industry.

An Artificial Neural Network Approach to Predict the Relationship between the Processing Parameters and Properties of TC21 Titanium Alloy

This paper develops a three-layer back-propagation artificial neural network model to analyze and predict the correlation between processing parameters and properties of the damage tolerance type titanium alloy TC21. The inputs of the ANN are working temperatures, deformation extent, deformation rate and heat treatment conditions. And the outputs are mechanical properties namely ultimate strength, yield strength, elongation, reduction of area, plane strain fracture toughness and microstructure concerned parameters such as β phase fraction, βphase grain size, substructure length and thickness. The ANN is trained with experimental data and achieves a very good performance, which has already been applied to the optimization of processing for forging of aero-parts.

Ecosystem health assessment of Honghu Lake Wetland of China using artificial neural network approach

Honghu Lake, located in the southeast of Hubei Province, China, has suffered a severe disturbance during the past few decades. To restore the ecosystem, the Honghu Lake Wetland Protection and Restoration Demonstration Project (HLWPRDP) has been implemented since 2004. A back propagation (BP) artificial neural network (ANN) approach was applied to evaluatinig the ecosystem health of the Honghu Lake wetland. And the effectiveness of the HLWPRDP was also assessed by comparing the ecosystem health before and after the project. Particularly, 12 ecosystem health indices were used as evaluation parameters to establish a set of three-layer BP ANNs. The output is one layer of ecosystem health index. After training and testing the BP ANNs, an optimal model of BP ANNs was selected to assess the ecosystem health of the Honghu Lake wetland. The result indicates that four stages can be identified based on the change of the ecosystem health from 1990 to 2008 and the ecosystem health index ranges from morbidity before the implementation of HLWPRDP (in 2002) to middle health after the implementation of the HLWPRDP (in 2005). It demonstrates that the HLWPRDP is effective and the BP ANN could be used as a tool for the assessment of ecosystem health.

Optimization for Loading Paths of Tube Hydroforming Using a Hybrid Method

Tube hydroforming (THF) is an advanced technology with the advantages of lightweight and integrity, which can be used to manufacture hollow structural components. The process of THF is influenced by many factors, among which the matching relation between the internal pressure and axial feed, i.e., loading paths, is particularly important. In this article, a hybrid method is proposed to optimize loading paths of THF. Firstly, a three-layer back-propagation artificial neural network (BP-ANN) is built, and 200 samples from finite element (FE) simulations are applied to train and test the artificial neural network (ANN). Then genetic algorithm (GA) is adopted to search the optimal loading paths in the specified bounds of the design variables by using the trained ANN as the solver of the objective function and constraint functions. After 59 iterations, the optimal loading paths are obtained. Finally, the verified experiments are performed on the special hydroforming press. The results show that the proposed method can effectively search the optimal loading paths of THF and remarkably improve the quality of the final formed parts.

Dewpoint Temperature Prediction Using Artificial Neural Networks

AbstractDewpoint temperature, the temperature at which water vapor in the air will condense into liquid, can be useful in estimating frost, fog, snow, dew, evapotranspiration, and other meteorological variables. The goal of this study was to use artificial neural networks (ANNs) to predict dewpoint temperature from 1 to 12 h ahead using prior weather data as inputs. This study explores using three-layer backpropagation ANNs and weather data combined for three years from 20 locations in Georgia, United States, to develop general models for dewpoint temperature prediction anywhere within Georgia. Specific objectives included the selection of the important weather-related inputs, the setting of ANN parameters, and the selection of the duration of prior input data. An iterative search found that, in addition to dewpoint temperature, important weather-related ANN inputs included relative humidity, solar radiation, air temperature, wind speed, and vapor pressure. Experiments also showed that the best models included 60 nodes in the ANN hidden layer, a ±0.15 initial range for the ANN weights, a 0.35 ANN learning rate, and a duration of prior weather-related data used as inputs ranging from 6 to 30 h based on the lead time. The evaluation of the final models with weather data from 20 separate locations and for a different year showed that the 1-, 4-, 8-, and 12-h predictions had mean absolute errors (MAEs) of 0.550°, 1.234°, 1.799°, and 2.280°C, respectively. These final models predicted dewpoint temperature adequately using previously unseen weather data, including difficult freeze and heat stress extremes. These predictions are useful for decisions in agriculture because dewpoint temperature along with air temperature affects the intensity of freezes and heat waves, which can damage crops, equipment, and structures and can cause injury or death to animals and humans.

Fast Discrimination of Apple Varieties Using Vis/NIR Spectroscopy

We evaluated the potential of visible/near-infrared (Vis/NIR) spectroscopy for its ability to nondestructively differentiate apple varieties. The apple varieties used in this research included, Fuji apples, Red Delicious apples, and Copefrut Royal Gala apples. The chemometrics procedures applied to the Vis/NIR data were principal component analysis (PCA), wavelet transform (WT), and artificial neural network (ANN). The apple varieties could be qualitatively discriminated in the PC1-PC2 space resulted from PCA. Wavelet transform was used as a tool for dimension reduction and noise removal, reducing spectral to wavelet components. Wavelet components were utilized as input for three-layer back propagation ANN model. WT-ANN model gave the highest level of correct classification (100%) of the apple varieties.