Effects of chilled storage on fish freshness using computer vision and artificial neural network modeling

The objective of this work was to develop Artificial Neural Network (ANN) based thermal conductivity (K) prediction model for Iranian flat breads. Experimental data needed for ANN models were obtained from a pilot-scale set-up. Breads were made from three different cultivars of wheat and were baked in an eclectic oven at three different baking temperatures (232°C, 249°C and 260°C). A data set of 205 conditions was used for developing ANN and empirical models. To model K using ANN, 16 different MLP (multilayer perceptron) configurations ranging from one to two hidden layers of neurons were investigated and their prediction performances were evaluated. The (4-3-5-1)-MLP network, that is a network having two hidden layers, with three neurons in its first hidden layer and five neurons in its second hidden layer, had the best results in predicting the thermal conductivity of flat bread. For this network, R2, MRE, MAE and SE were 0.988, 0.6323, 1.66×10− 3, and 8.56×10−4, respectively. Overall, ANN models (with R2 ≥ 0.95) performed superior than the empirical model (with R2 = 0. 870).

The prediction of an accurate geodetic point velocity has great importance in geosciences. The purpose of this work is to explore the predictive capacity of three artificial neural network (ANN) models in predicting geodetic point velocities. First, the multi-layer perceptron neural network (MLPNN) model was developed with two hidden layers. The generalized regression neural network (GRNN) model was then applied for the first time. Afterwards, the radial basis function neural network (RBFNN) model was trained and tested with the same data. Latitude ( $$\varphi$$ ) and longitude (λ) were utilized as inputs and the geodetic point velocities ( $${V}_{X}$$ , $${V}_{Y}$$ , $${V}_{Z}$$ ) as outputs to the MLPNN, GRNN, and RBFNN models. The performances of all ANN models were evaluated using root mean square error (RMSE), mean absolute error (MAE), and coefficient of determination ( $${\text{R}}^{2}$$ ). The first investigation demonstrated that it was possible to predict the geodetic point velocities by using all the components as output parameters simultaneously. The other result is that all ANN models were able to predict the geodetic point velocity with satisfactory accuracy; however, the GRNN model provided better accuracy than the MLPNN and RBFNN models. For example, the RMSE and MAE values were 1.77–1.88 mm and 1.44–1.51 mm, respectively, for the GRNN model.

Castellated steel beams (CSB) are an attractive option for the steel construction industry thanks to outstanding advantages, such as the ability to exceed large span, lightweight, and allowing flexible arrangement of the technical pipes through beams. In addition, the complex localized and global failures characterizing these structural members have led researchers to focus on the development of efficient design guidelines. This paper aims to propose an artificial neural network (ANN) model with optimal architecture to predict the load‐carrying capacity of CSB with a scheme of the simple beam bearing load located at the center of the beam. The ANN model is built with 9 input variables, which are essential parameters equivalent to the geometrical properties and mechanical properties of the material, such as the overall depth of the castellated beam, the vertical projection of the inclined side of the opening, the web thickness, the flange width, the flange thickness, the width of web post at middepth, the horizontal projection of inclined side of the opening, the minimum web yield stress, and the minimum flange yield stress. The output variable is the load‐carrying capacity of the CSB. With the optimal ANN architecture [9‐1‐1] containing one hidden layer, the performance of the ANN model is evaluated based on statistical criteria such as R2, RMSE, and MAE. The results show that the optimal ANN model is a highly effective predictor of the load‐carrying capacity of the CSB with the best value of R2 = 0.989, RMSE = 3.328, and MAE = 2.620 for the testing part. The ANN model seems to be the best algorithm of machine learning for predicting the CSB load‐carrying capacity.

The purpose of this study was to develop a method of measuring the amylose content of rice using image processing techniques and an Artificial Neural Network (ANN) model. The rice samples came from six varieties, namely Way Apo Buru, Mapan P05, IR-64, Cibogo, Inpari IR Nutri Zinc, and Inpari 33. The amylose content was measured by laboratory tests and the color intensity was measured based on the RGB (Red, Green, Blue). The ANN model will correlate the RGB color intensity as input with the amylose content as the output. The ANN model used is backpropagation type with 3 input layer nodes and 2 hidden layers with 3-5-5-1 architecture. Variations in the training model used are 27 variations of the activation function. The amount of data used for model training of 30 data while for validation of 12 data. The best ANN model is determined from the high value of accuracy (100%-MAPE) and the value of coefficient of determination (R2). The results showed the best network architecture on the activation function purelin-logsig-tansig. The R2 value on the best training and validation results of 0.98 and 0.66 while the accuracy values for the best training and validation results of 98.15 and 66.82. The validation results show that the developed non-destructive method can be used to quickly and accurately measure the amylose value of rice based on RGB color value. The test results show that the non-destructive method developed cannot be used to measure the amylose content of rice quickly and accurately based on the RGB color intensity, so it needs further development. Keywords: Amylose, Artificial neural networks, Image processing, Rice

In the chemical industry is important to control the process in order to guarantee the quality and repeatability of the final product. Using sensors in the industrial plant allows a large volume of data to be captured regarding the process. These data can be used for modelling to better understanding and predict the properties of the product in the process. In this work, two types of Artificial Neural Networks (ANN) and the hybrid model Adaptive Neuro Fuzzy Inference System (ANFIS) were used to predict the density of polystyrene along the styrene polymerization process. The dataset used was extracted from the batch of polymerization reactions performed in open-loop, manual control and closed-loop and monitored in each 5 seconds. The Feedforward and Elman ANN has coefficient of correlation (R) equal 94.2%. However, the best topology obtained to Feedforward ANN presents 2 hidden layers and error index RMSE (Root Mean Squared Error) equal to 2.69x10-2. The Elman ANN presents only 1 hidden layer and RMSE of 3.39x10-2. The ANFIS model, in turn, presented R equal to 91% and RMSE of 0.2123. Therefore, ANFIS model did not prove to be the most adequate for the prediction of the polystyrene density in the studied process.Polymerization process pose significant challenges to the industrial community as it is difficult to control with high nonlinearity behaviour and fast dynamic response. The monitoring and control of polymer processes guarantee to the final product the qualities required by the market. Muhammad and Aziz (2017), for instance, studied the production of low density polyethylene (LDPE) and presented a review of the control strategies developed for the LDPE process. The strategies presented were developed in tubular and autoclave reactors and highlights the importance of nonlinear control in polymerization process.The use of sensors to monitor production allows a large volume of process data to be collected. Therefore, it is necessary to construct mathematical models of prediction to interpret and correlate significant patterns, indispensable to assist in the management of decisions and risk analysis.The development of phenomenological models for polymerization is complex and requires deep knowledge about the processes involved in each step. Modelling using artificial intelligence is a strategy that can provide valuable information about the process and allows the construction of an intelligent model capable of predicting process response based on parameters provided. ANN and ANFIS are artificial intelligence tools that can be used to build predictive models. In supervised learning, the data are presented to the network and its main objective is to provide a model that correctly correlates the pairs inputs - outputs of the problems. The use of ANN and ANFIS to predict the density of the polystyrene produced in the process becomes attractive since it is a type of non-linear modelling.Jumari and Mohd-Yusof (2017) presents models to measure melt flow index (MFI) in industrial polypropylene loop reactors using first principle (FP) model and ANN model. The authors state that the prediction of the ANN model is more accurate compare to the MFI calculated by the FP model. Furthermore, the CPU time recorded that ANN model is much faster than FP model.This work aims to develop direct and recursive ANN and ANFIS models from a set of experimental data from a controlled styrene polymerization plant capable of predicting the density of the product satisfactorily.

Comparison of Two Different Artificial Neural Networks for Prostate Biopsy Indication in Two Different Patient Populations

Reduction of passenger cars fuel consumption and associated emissions are two major goals of sustainable transport over the last years. Passenger car fuel consumption is directly related to a number of technological aspects of a given car, driver behaviour, road and weather conditions and, especially at urban level, road structure and traffic flow and conditions. In this paper, passenger car fuel consumption was assumed to be a function of three input variables, i.e. day of week, hour of day and city zone. Over the period of 6 months (during 2015) a car was driven in the randomly chosen routes in the city of Niš (Serbia) in the period from 8 to 23 h. The fuel consumption data recorded through on-board diagnostics equipment were used for the development of Artificial Neural Network (ANN) models. In order to efficiently deal with a number of ANN design issues, to avoid usual trial and error procedure and develop robust, high performance ANN models, the Taguchi method was applied. For experimentation with ANN design parameters (transfer function, the number of neurons in the first hidden layer, the number of neurons in the second hidden layer, training algorithm), the standard L18 orthogonal array with two replications was selected. Statistical results indicate the dominant influence of the training algorithm, followed by the ANN topology, i.e. interaction of the number of neurons in hidden layers, on the ANN models performance. It has been observed that 3-8-8-1 ANN model represents an optimal model for prediction of passenger car fuel consumption. This model has logistic sigmoid transfer functions in hidden layers trained with scaled conjugate gradient algorithm. By using the Taguchi optimized ANN models, analysis of passenger car fuel consumption has been discussed based on traffic conditions, i.e. different days of the week and hours of the day, for each city zone and separately for summer and winter periods.

The present study investigated the possible use of artificial neural networks (ANN) to classify five chestnut (Castanea sativa Mill.) varieties. For chestnut classification, back-propagation neural networks were framed on the basis of physical and mechanical parameters. Seven physical and mechanical characteristics (geometric mean diameter, sphericity, volume of nut, surface area, shell thickness, shearing force and strength) of chestnut were determined. It was found that these characteristics were statistically different and could be used in the classification of species. In the developed ANN model, the design of the network is 7-(5-6)-1 and it consists of 7 input, 2 hidden and 1 output layers. Tansig transfer functions were used in both hidden layers, while linear transfer functions were used in the output layer. In ANN model, R2 value was obtained as 0.99999 and RMSE value was obtained as 0.000083 for training. For testing, R2 value was found as 0.99999 and RMSE value was found as 0.00031. In the approximation of values obtained with ANN model to the values measured, average error was found as 0.011%. It was found that the results found with ANN model were very compatible with the measured data. It was found that the ANN model obtained can classify chestnut varieties in a fast and reliable way.

This article tries to develop an integrated artificial neural network (ANN) model for spatial and temporal forecasting of daily suspended sediment discharge at multiple gauging stations in Eel River watershed in northwest California. Complexity of runoff–sediment process and its variability in space and time and also lack of historical sediment data cause difficulties in spatiotemporal modeling of this process. Initially, and for comparison purpose, six single-station ANN models, which are customary in modeling sediment yield, were developed. Then an integrated ANN model for modeling multiple stations was proposed and its spatiotemporal modeling ability was examined through a cross-validation technique for a station. In this way, different multilayer perceptron neural networks were trained using Levenberg–Marquardt algorithm to estimate daily values of suspended sediment discharge. Various combinations of input and hidden layers' neurons were applied and the optimum architectures of the models were selected according to the obtained evaluation criteria in the terms of Nash–Sutcliffe efficiency coefficient, root mean-squared error, and ratio of absolute error of peak flow. To improve the model, input data were classified into two clusters by k-means clustering scheme. Afterwards, clustered data were used as inputs for two integrated models and their performances were evaluated. The proposed integrated ANN model shows reasonable performance in spatiotemporal modeling both before and after clustering. Nevertheless, clustering decreases the complexity of the model.

This study intends to predict the influence of injection pressure and injection timing on performance, emission and combustion characteristics of a diesel engine fuelled with waste cooking palm oil based biodiesel using the artificial neural network (ANN) model. To acquire data for training and testing in the proposed ANN, experiments were carried out in a single cylinder, four stroke direct injection diesel engine at a constant speed of 1500 rpm and at full load (100%) condition. From the experimental results, it was observed that waste cooking palm oil methyl ester provided better engine performance and improved emission and combustion characteristics at injection pressure of 280 bar and timing of 25.5° bTDC. An ANN model was developed using the data acquired from the experiments. Training of ANN was performed based on back propagation learning algorithm. Multilayer perceptron (MLP) network was used for non-linear mapping of the input and output parameters. Among the various networks tested the network with two hidden layers and 11 neurons gave better correlation coefficient for the prediction of engine performance, emission and combustion characteristics. The ANN model was validated with the test data which was not used for training and was found to be very well correlated.

Separation of lead ions from wastewater using electrodialysis: Comparing mathematical and neural network modeling

Prediction of void fraction for gas–liquid flow in horizontal, upward and downward inclined pipes using artificial neural network

The impact of climate on crop production has vital importance. Climate variables affect the different crops during different stages of the growth and the development. This research aims to study the environmental factors affecting the growth and production of barley (Hordeum Sp., Gramineae) in a hydroponic system, to provide information to farmers and decision makers by using Artificial Neural Network (ANN) Model for production prediction. Multilayer feed-forward ANN (fully connected) was used in supervised manner and the training method was the back-propagation algorithm by using MATLAB program. The inputs in the ANN model of barley were: seeds density (kg/m2), lighting duration (h/day), light intensity (Lux), temperature (co), relative humidity (%) and growing period (days). The outputs were: plant length (cm), yield (kg/m2), protein (%), dry matter (%), and conversion factor. Results revealed that the optimal configuration for the ANN model consisted of four layers (6-25-30-5). The hidden layers had 25 and 30 nodes in the first and second hidden layers respectively for the ANN model. Hyperbolic tangent transfer function was employed in hidden and output layers of the ANN model. The learning rate and the momentum parameter were 0.005 and 0.9 respectively for the ANN model. Iterations were 10000 epochs during training process for the ANN model. The results showed that the variation between target and predicted outputs was small while the correlation coefficient (R) was 0.99. Also, the results revealed that the major parameters affecting on all the outputs were seeds density and the duration of the lighting followed by the other factors i.e. temperature (co), relative humidity (%), growing period (days) and light intensity (Lux). Seeds density has a higher percent relative importance, on yield, plant length, protein (%), DM (%) and conversion factor equal to 22.8%, 24%, 25%, 24% and 22.8% respectively. The developed ANN model was beneficial tool for barley production prediction. The barley yield prediction could be helpful for farmers, decision makers and planning to manage their crop better by providing a series of recommendations about crops planting and clarifying its impact on changes to these factors under the study in order to avoid losses and reach the best benefit (maximization of yield).

Aim of study: Wheat appropriate harvest date (WAHD) is an important factor in farm monitoring and harvest campaign schedule. Satellite remote sensing provides the possibility of continuous monitoring of large areas. In this study, we aimed to investigate the strength of vegetation indices (VIs) derived from Landsat-8 for generating the harvest schedule regional (HSR) map using Artificial Neural Network (ANN), a robust prediction tool in the agriculture sector.Area of study: Qorveh plain, Iran.Material and methods: During 2015 and 2016, a total of 100 plots was selected. WAHD was determined by sampling of plots and specifying wheat maximum yield for each plot. The strength of eight Landsat-8 derived spectral VIs (NDVI, SAVI, GreenNDVI, NDWI, EVI, EVI2, CVI and CIgreen) was investigated during wheat growth stages using correlation coefficients between these VIs and observed WAHD. The derived VIs from the required images were used as inputs of ANNs and WAHD was considered as output. Several ANN models were designed by combining various VIs data.Main results: The temporal stage in agreement with dough development stage had the highest correlation with WAHD. The optimum model for predicting WAHD was a Multi-Layer Perceptron model including one hidden layer with ten neurons in it when the inputs were NDVI, NDWI, and EVI2. To evaluate the difference between measured and predicted values of ANNs, MAE, RMSE, and R2 were calculated. For the 3-10-1 topology, the value of R2 was estimated 0.925. A HSR map was generated with RMSE of 0.86 days.Research highlights: Integrated satellite-derived VIs and ANNs is a novel and remarkable methodology to predict WAHD, optimize harvest campaign scheduling and farm management.

The artificial neural network (ANN) method was used in comparison with the adaptive neuro-fuzzy inference system (ANFIS) to describe polygalacturonase (PG) production by Bacillus subtilis in submerged fermentation. ANN was evaluated with five neurons in the input layer, one hidden layer with 7 neurons, and one neuron in the output layer. Five fermentation variables (pH, temperature, time, yeast extract concentration, and K2HPO4 concentration) served as the input of the ANN and ANFIS models, and the polygalacturonase activity was the output. Coefficient of determination (R-2) and root mean square values (RMSE) were calculated as 0.978 and 0.060, respectively for the best ANFIS structure obtained in this study. The R-2 and RMSE values were computed as 1.00 and 0.030, respectively for the best ANN model. The results showed that the ANN and ANFIS models performed similarly in terms of prediction accuracy.