Prediction of various functional finishing treatments using artificial neural networks

In this study, in order to produce different water-oil repellent and wrinkle resistant fabrics, 21 different kinds of blended woven fabrics were treated (padded and transfered) with both classic and nano chemicals according to 4 different levels of concentrations. Afterwards, water, oil repellency and wrinkle angle recovery properties of the fabrics were measured. The purpose of this study is to predict these aforementioned functional properties of the fabrics before manufacturing based on the fabric blend, treatment method, used chemicals and chemical concentrations with the help of multi layer perceptron, one of the most popular network architecture. As a result of the study, it can be concluded that multi layer perceptron method can also be used for the classification problems successfully.     

Effect of initial water content on saturated hydraulic conductivity in desalinization with slaking and drying

Soil slaking is an environment-friendly technique that is gaining importance in restoring saline soils. The objective of this article is to evaluate the effect of initial water content (IWC) on saturated hydraulic conductivity (K _s) in desalinization with slaking and drying. Accordingly, a slaking test was carried out during February, 2009 for evaluating the effects of slaking and drying on K _s, sodium adsorption ratio (SAR) and exchangeable sodium percentage (ESP) under various IWC. We prepared natural and air-dried soils of paddy field in Kojima Bay Polder, Japan to give different pre-drying, air-dried, and not dried (natural). The air-dried soils were resaturated. Each soil was well mixed, then dried to different initial moisture contents (60, 50, 40, 30, 20, and 10% by weight). The specimens were immersed into water in the pot for 24 h. The K _s was measured, and cations in slaked and unslaked soils were analyzed. The K _s was high under the water content below 30% in both the natural and the air-dried soils. But the effects were more pronounced in the natural soil. The air-dried soil showed far smaller K _s than the natural soil. In outer solution, the highest SAR was noted at 30% in the natural and 30 and 20% in the air-dried soils. Significant decrease in ESP of the soils (slaked + unslaked) was also observed at the same water content. Lower water content was more effective in decreasing the soil ESP after desalinization from saline soil. The natural soil showed lower ESP and higher porosity, which was considered as a reason for higher K _s of natural soil than that of air-dried soils. The results indicated that lower water content (10–30%) had no hazardous effect on K _s by slaking and drying of soil.     

Spatial variability of soil saturated hydraulic conductivity in paddy field in accordance to subsurface percolation

Insufficient puddling with inappropriate implements or imprecise time/intensity may alter saturated water flow in paddy soil spatially or temporary due to change in aggregate size distribution, dry bulk density, saturated hydraulic conductivity, and percolation rate of the soil. In this study, spatial variability of saturated hydraulic conductivity (K _s), a key parameter of the saturated water flow, in Fuchu Honmachi paddy plot (100 m × 28 m) was characterized based on dielectric or ADR dry bulk density (ρ_b-ADR) with help of non-similar media concept (NSMC) and geostatistics model to meet its correlation to subsurface percolation. A 100 cc core and an ADR data were sampled from each sub-plot (7 m × 7.5 m), and then were used for measuring and predicting ρ_b and K _s. The predicted data agreed with the measured ones, in which they fitted well the x = y line with RMSE of 0.029 cm³ cm⁻³ (R ² = 0.68), 0.027 g cm⁻³ (R ² = 0.71) (ρ_b), and 0.098 cm d⁻¹ (R ² = 0.45) for θ, ρ_b, and K _s, respectively. The predicted ρ_b and K _s had similar trend in spatial variability to the measured ones particularly within the distance of 46.3–51.9 m and 26.2–27.9 m, respectively. The spatial variability of the predicted K _s coincided to that of the subsurface percolation rate, in which they had similar distance of dependence. The results indicated that the presenting method can be reasonably accepted.     

Pedo-transfer function for saturated hydraulic conductivity of lowland paddy soils

In paddy field, soil saturated hydraulic conductivity (K _s) plays as an important component in the calculation of irrigation requirement of the water balance equation and also for irrigation efficiency. Several laboratory and field methods can be used to determine K _s. Laboratory and field determinations are usually time consuming, expensive and labour intensive. Pedo-transfer functions (PTF) serve to translate the basic information found in the soil survey into a form useful for broader applications through empirical regression of functional relationships, such as simulation modelling. Since PTFs have not been applied to paddy soils in the study area, a lot of field measurements will require high labour input to determine K _s hence high cost. This study attempts to seek a simplified method for determining K _s values based on common existing soil properties through PTF technique. Soil samples (n = 408 samples) were collected randomly depending on the soil series within the 2,300 ha Sawah Sempadan rice cultivation area. Both field work and laboratory work were carried out. The samples were then analysed for the following properties: dry bulk density (D _b), soil particle percentage (Sand-S, Silt-Si and Clay-C), organic matter (OM) and geometric mean diameter (GMD). The measured K _s values were obtained by using the falling head method. The parameters were then used as inputs for developing a K _s model by regression analysis using Statistical Analysis System (SAS) package. Stepwise regression analysis was applied to determine the best fit model based on R ² and significant level. The results of the study showed that there is a high spatial variability of the saturated hydraulic conductivity in the paddy area. The best regression model for estimating K _s was based on C, D _b, OM and GMD with the dependent variable (K _s) in a form of natural logarithm. The model inputs introduced by stepwise regression are commonly available therefore, this model is useful to replace the conventional method.     

Prediction of milled rice grades using Fourier transform near-infrared spectroscopy and artificial neural networks

This paper describes a method to rapidly and objectively predict the grades of milled rice according to the surface lipid content (SLC), which was determined by using near-infrared (NIR) spectroscopy. Sixty-six rice varieties were milled to different degrees. Then each sample was graded by a three-member panel. After the NIR spectra for each sample were collected over the wavenumber range of 11,000–4000 cm⁻¹, the SLC of each sample was measured according to the official method. The calibration equations relating the Fourier Transform Near-infrared (FT-NIR) spectra to the measured SLC were developed based on the partial least square (PLS) regression. The best model gave the root mean square error of the prediction (RMSEP) of 0.0248% and the determination coefficients of 0.9905. If the relationships between the grades and the SLC predicted by the developed NIR model were described with the linear and the logarithmic regression equations, the correct prediction percents (CCP) were 75.76% and 83.33%, respectively. When the back propagation artificial neural network (BP-ANN) model was developed to estimate the grades according to SLC, the resultant CCP was 95.45%, indicating that the milled rice grades could be predicted by the proposed BP-ANN model with satisfactory accuracy.     

Predicting bending rigidity of woven fabrics using artificial neural networks

This paper reports an investigation on the predictability of bending property of woven fabrics from their constructional parameters using artificial neural network (ANN) approach. Number of cotton grey fabrics made of plain and satin weave designs were desized, scoured, and relaxed. The fabrics were then conditioned and tested for bending properties. Thread density in fabric, yarn linear density, twist in yarn, and weave design were accounted as input parameters for the model whereas bending rigidity in warp and weft directions of fabric formed the outputs. Gradient descent with momentum and an adaptive learning rate back-propagation was employed as learning algorithm to train the network. A sensitivity analysis was carried out to study the robustness of the model.     

Prediction of baking results from farinograph measurements by using stepwise linear regression and artificial neuronal networks

In an effort to extract additional data from farinograph experiments a model was developed to simulate the measurements and correlate the parameters of the model with results from baking tests. This additional information can be used in bakeries to predict the baking properties of the flours and adjust the recipes to maintain a constant product quality. For this eight different flours were characterized with a farinograph and 13 different results from baking experiments. An approach with five nonlinear differential equations was able to model the farinograph measurements very well (average R² = 0.995 ± 0.005). While a stepwise multilinear regression only showed weak correlations in cross validation between a single parameter of the model and the baking volume (R² = 0.745) and the volume yield (R² = 0.796) respectively, the artificial neuronal network was more successful. For the baking weight (R² = 0.926), the dough yield gross (R² = 0.909) and net (R² = 0.913) strong correlations were found. A good correlation for the baking volume (R² = 0.853) was also determined, while the volume yield showed comparable results to the linear regression (R² = 0.792).     

Estimating soil hydraulic conductivity using different data-driven models of ANN,GMDH and GMDH-HS

The saturated hydraulic conductivity (\(K_{\text{s}}\)) is one of the important soil hydraulic properties which plays a significant role in developing flow transport models and irrigation and drainage practices. In this research, artificial neural networks approaches, group method of data handling (GMDH) model and a hybrid intelligent model based on combination of GMDH and harmony search (HS) model (GMDH-HS) were developed to estimate \(K_{\text{s}}\) based on 151 field samples collected from the northeast of Iran. Eleven topsoil properties were used as input parameters to estimate \(K_{\text{s}}\). The five quantitative standard statistical performance evaluation measures, i.e., coefficient of efficiency, root-mean-square error, mean square relative error, mean absolute percentage error and relative bias, were employed to evaluate the performance of various developed models. Statistical results indicated that the best performance can be obtained by GMDH-HS in terms of different evaluated criteria during the training and testing datasets for \(K_{\text{s}}\) estimation.     

Development of prediction system using artificial neural networks for the optimization of spinning process

This article correlates draw frame settings with quality characteristics of sliver and ring spun yarn using artificial neural networks. Considering the importance of draw frame as the last quality improvement machine in the spinning process, the quality influencing parameters of the draw frame were used as input for artificial neural networks. The neural networks were trained using a combination of Levenberg-Marquardt algorithm and Bayesian regularization for better generalization of the networks. Cross validation was performed for each trained network to test the performance of networks. The promising results achieved by this research work emphasize the ability of neural networks to predict the quality characteristics of sliver and yarn using the artificial neural networks. Therefore, draw frame parameters can be adjusted on the basis of required sliver and yarn quality. Furthermore, machines can be involved in the decision making process in spinning mills.     

Evaluation of effective electrospinning parameters controlling gelatin nanofibers diameter via modelling artificial neural networks

The aim of this work was to evaluate the effective parameters for prediction of the electrospun gelatin nanofibers diameter using artificial neural network (ANN) technique. The various sets of electrospinning process including temperature, applied voltage and polymer and solvent concentrations were designed to produce pure gelatin nanofibers. The obtained results by analyzing Scanning Electron Microscopy (SEM) images indicated that the produced nanofibers diameter was in the range of 85 to 750 nm. Due to the volume of the data, k fold cross-validation method was used for data setting. Data were divided into the five categories and trained and tested using ANN technique. The results indicated that the network including 4 input variables, 3 hidden layers with 10, 18 and 9 nodes in each layers, respectively, and one output layer had the best performance in the testing sets. The mean squared error (MSE) and linear regression (R) between observed and predicted nanofibers diameter were 0.1531 and 0.9424, respectively. The obtained results demonstrated that the selected neural network model had acceptable performance for evaluating involved parameters and prediction of nanofibers diameter.     

Simulation for response of crop yield to soil moisture and salinity with artificial neural network

In saline fields, irrigation management often requires understanding crop responses to soil moisture and salt content. Developing models for evaluating the effects of soil moisture and salinity on crop yield is important to the application of irrigation practices in saline soil. Artificial neural network (ANN) and multi-linear regression (MLR) models respectively with 10 (ANN-10, MLR-10) and 6 (ANN-6, MLR-6) input variables, including soil moisture and salinity at crop different growth stages, were developed to simulate the response of sunflower yield to soil moisture and salinity. A connection weight method is used to understand crop sensitivity to soil moisture and salt stress of different growth stages. Compared with MLRs, both ANN models have higher precision with RMSEs of 1.1 and 1.6 t ha⁻¹, REs of 12.0% and 17.3%, and R² of 0.84 and 0.80, for ANN-10 and ANN-6, respectively. The sunflower sensitivity to soil salinity varied with the different soil salinity ranges. For low and medium saline soils, sunflower yield was more sensitive at crop squaring stage, but for high saline soil at seedling stage. High soil moisture content could compensate the yield decrease resulting from salt stress regardless of salt levels at the crop sowing stage. The response of sunflower yield to soil moisture at different stages in saline soils can be understood through the simulated results of ANN-6. Overall, the ANN models are useful for investigating and understanding the relationship between crop yield and soil moisture and salinity at different crop growth stages.     

Investigation on air permeability of finished stretch plain knitted fabrics. I. Predicting air permeability using artificial neural networks

Air permeability is one of the most important utility properties of textile materials as it influences air flow through textile material. Air permeability plays a significant role in well-being due to its influence on physiological comfort. The air permeability of textile materials depends on their porosity. There are a lot of structural properties of textile materials also operating parameters (knitting+finishing) influencing air permeability and there are also statistically significant interactions between the main factors influencing the air permeability of knitted fabrics made from pure yarn cotton (cellulose) and viscose (regenerated cellulose) fibers and plated knitted with elasthane (Lycra) fibers. Two types of artificial neural networks (ANNs) model have been set up before modeling procedure by utilizing multilayer feed forward neural networks, which take into account the generality and the specificity of the product families respectively. A virtual leave one out approach dealing with over fitting phenomenon and allowing the selection of the optimal neural network architecture was used. Moreover this study exhibited that air permeability could be predicted with high accuracy for stretch plain knitted fabrics treated with different finishing processes. Within the framework of the work presented, ANNs were applied to help industry to adjust the operating parameter before the actual manufacturing to reach the desired air permeability and satisfy their consumers.     

Prediction of semolina technological quality by FT-NIR spectroscopy

The feasibility of applying FT-NIR spectroscopy (a rapid and non-destructive method) to evaluate and predict semolina characteristics by means of spectra collected directly from the kernels was investigated. More than 500 samples of durum wheat grains and of the corresponding semolina, representative of the Italian production of 4 different crops (from 2002/2003 to 2005/2006) were analyzed. Pasta-making capability of each semolina sample was assessed by the reference methods, for protein, gluten content, gluten index and alveographic indices. The kernels were also evaluated by a FT-NIR spectrometer, fitted with an integration sphere working in diffuse reflectance. The processed spectra collected on the kernels were correlated with the chemical and rheological parameters obtained by the reference tests performed on semolina. The PLS algorithm was used to develop calibration models from the original spectra datasets. Protein content proved to be well correlated to kernel spectral data: high values for the RPD indicate efficient NIR reflectance predictions for protein content. The models obtained for gluten content, gluten index and alveographic W and P/L parameters were less successful. The results of this work highlighted the feasibility of applying FT-NIR spectroscopy to evaluate and predict the technological properties of semolina, in particular that of the protein content, by collecting the spectra directly from the kernels, without performing further grinding or milling operations.     

Prediction of CIELab data and wash fastness of nylon 6,6 using artificial neural network and linear regression model

We tried to predict the CIELab data and wash fastness values of scoured nylon 6.6 knitted fabric dyed with 1:2 metal-complex acid dyes and aftertreated using three different methods named as syntan, syntan/cation and full backtan by artificial neural network (ANN) with Levenberg-Marquardt algorithm and regression models. Afterward, the predicting performance of these models was tested and compared with each other using unseen data sets. We were able to achieve to predict the all colorimetric data satisfactorily such as L*, a*, b*, C, h ^o and wash fastness performance using both models. The statistical findings indicated that the regression models provide more accurate prediction for all colour data with an average error of 1% contrast to previous study. In terms of prediction of fastness, artificial neural network is a bit more useful than regression models for prediction of staining value on the nylon part of adjacent multifiber.     

Prediction of the changes on the CIELab values of fabric after chemical finishing using artificial neural network and linear regression models

Changes on the CIELab values of the dyed materials after the different chemical finishing treatments using artificial neural network (ANN) and linear regression (LR) models have been predicted. The whole structural properties of fabrics and some process data which were from fiber to the finishing parameters were accepted as inputs in these models. The networks having different structures were established, and it was also focus on the parameters which could affect the performance of the established networks. It was determined that we could successfully predict the color differences values occurring on the material after the finishing applications. In addition, we realized that some ANN parameters affected the prediction performance while establishing the models. After training ANN models, the prediction of the color difference values was also tried by linear regression models. Then, extra ANN models were established for all outputs using the parameters as inputs in the LR equations, and the prediction performances of both established models were compared. According to the results, the neural network model gives a more accurate prediction performance than the LR models.     

不同有机质含量的菜园土保水效果研究

2002～2004年,在东安采用田间小区试验,研究了不同有机质含量的黄泥菜园土种植辣椒时对其土壤含水量、保水能力、产量和水分利用效率的影响.结果表明,有机质含量高的菜园土保水能力强,土壤含水量和产量也相应较高,水分利用效率更高,并且土壤含水量随土层加深呈上升趋势.     

Predicting the tensile strength of polyester/cotton blended woven fabrics using feed forward back propagation artificial neural networks

Tensile strength plays a vital role in determining the mechanical behavior of woven fabrics. In this study, two artificial neural networks have been designed to predict the warp and weft wise tensile strength of polyester cotton blended fabrics. Various process and material related parameters have been considered for selection of vital few input parameters that significantly affect fabric tensile strength. A total of 270 fabric samples are woven with varying constructions. Application of nonlinear modeling technique and appreciable volume of data sets for training, testing and validating both prediction models resulted in best fitting of data and minimization of prediction error. Sensitivity analysis has been carried out for both models to determine the contribution percentage of input parameters and evaluating the most impacting variable on fabric strength.     

基于神经网络的甘蔗产量预测系统

首次将人工神经网络应用到甘蔗产量的预测中,介绍了ＢＰ网络模型及其算法,讨论了系统的开发方法,并对系统进行验证。对广西忻城糖厂蔗区８８／８９－９７／９８榨季的甘蔗单产和相应的气象条件运行结果表明模型具有较高的精度,复测误差在－５．３－１０．２％。     

土壤特性对大麻纤维产量和品质的影响

通过对多处麻田的定点观测和取样测定,分析了栽培大麻的主要土壤类型和土壤条件,明确了影响大麻纤维产量和品质的主要土壤理化因素及相关程度.