In the present study three different types of neural models: multi-layer perceptron (MLP), generalized regression neural network (GRNN) and radial basis function (RBF) has been used to predict the ...exergetic efficiency of roughened solar air heater. The experiments were conducted at NIT Jamshedpur, India, using two different types of absorber plate: arc shape wire rib roughened with relative roughness height 0.0395, relative roughness pitch 10 and angle of attack 60°, and smooth absorber plates for 7 days. Total 210 data sets were collected from the experiments. Mass flow rate, relative humidity, wind speed, ambient air temperature, inlet air temperature, mean air temperature, average plate temperature and solar intensity were selected as input parameters in input layer to estimate the exergetic efficiency. In the first part of study, MLP model has been used. In this model 10–20 neurons with LM learning algorithm were used in hidden layer for optimal model selection. It has been found that LM-18 is an optimal model. In second part, GRNN model was used. The GRNN model was simulated experimentally at different spread constants and found that keeping spread constant as 1.5, optimal results have been obtained. In the third part, RBF model was used. For optimal model, 1–5 spread constant at interval of 0.5 have been used. It has been found that by taking spread constant 3.5, best results are obtained. In the last part of the study, all neural models are compared on the basis of statistical error analysis. It has been found that RBF model is better than GRNN and MLP models due to lowest value of RMSE and MAE and highest value of R2 and ME. After RBF model, GRNN model performs better results as compared to MLP model. It has been found that the values of RMSE, MAE and R2 were 0.001652, 2.86E-04 and 0.99999 respectively for RBF model.
•MLP, RBF and GRNN models were used to predict exergetic performance of SAH.•Levenberg-Marquardt (LM) learning algorithm was applied for training in MLP model.•Best results have been observed for MLP model using 18 neurons (LM-18).•GRNN and RBF models are optimal at spread constants 1.5 and 3.5 respectively.•RBF model is the best among the three models MLP, GRNN and RBF.
•The shape, size and number of fins determine the performance of microchannels.•The flow velocity and fin transverse width are the main influence factors.•Neural network can accurately predict the ...flow and heat transfer in microchannel.•The heat sink can be optimized by comprehensive performance evaluation criterion.
Conjugate fluid-solid heat transfer in a pin-fin microchannel heat sink is an effective way to dissipate heat from the heating surface with high heat flux. The introduction of fins increases the heat exchange area and enhances flow turbulence, while it increases the flow resistance at the mean time. The thermal-hydraulic performance of heat sink is affected by fin shape, density and flow parameters. In this paper, contrived numerical simulations of the flow and heat transfer process in elliptical pin-fin microchannel heat sink are carried out, including 2033 cases with different fin sizes, numbers and flow velocities. The simulation results show that the flow velocity and fin transverse width are the main factors affecting heat transfer and fluid flow. Three artificial neural networks are established to predict the average temperature, the temperature non-uniformity of heating surface and the pressure drop of microchannel. The predicted results show that the pump power and heating surface temperature are contradictory objectives. A microchannel with the optimal thermal-hydraulic performance is selected. It has numerous fins which are longer in the flow direction. The empirical correlations for Nusselt number and friction coefficient of the optimal microchannel are proposed.
The process industries play a significant role in boosting the economy of any nation. However, poor management in several industries has been posing worrisome threats to an environment that was ...previously immaculate. As a result, the untreated waste and wastewater discarded by many industries contain abundant organic matter and other toxic chemicals. It is more likely that they disrupt the proper functioning of the water bodies by perturbing the sustenance of many species of flora and fauna occupying the different trophic levels. The simultaneous threats to human health and the environment, as well as the global energy problem, have encouraged a number of nations to work on the development of renewable energy sources. Hence, bioelectrochemical systems (BESs) have attracted the attention of several stakeholders throughout the world on many counts. The bioelectricity generated from BESs has been recognized as a clean fuel. Besides, this technology has advantages such as the direct conversion of substrate to electricity, and efficient operation at ambient and even low temperatures. An overview of the BESs, its important operating parameters, bioremediation of industrial waste and wastewaters, biodegradation kinetics, and artificial neural network (ANN) modeling to describe substrate removal/elimination and energy production of the BESs are discussed. When considering the potential for use in the industrial sector, certain technical issues of BES design and the principal microorganisms/biocatalysts involved in the degradation of waste are also highlighted in this review.
Display omitted
•Different types of bioelectrochemical systems (BESs) used for waste treatment are discussed•Parameters influencing the performance of BESs are listed and examined•Artificial neural network models used to evaluate the performance of BESs are presented•Microorganisms involved in waste degradation and future trends in BESs are highlighted
Display omitted
•A novel hybrid approach is proposed for short-term wind speed forecasting.•The wavelet packet technique is used to decompose the original wind speed series.•Crisscross optimization ...algorithm is applied to train artificial neural networks.•The proposed approach attains greater performance in terms of prediction accuracy.•The prediction results are not sensitive to the vertical crossover probability.
Wind speed forecasting is of great significance for wind farm management and safe integration into electric power grid. As wind speed is characterized by high autocorrelation and inherent volatility, it is difficult to predict with a single model. The aim of this study is to develop a new hybrid model for predicting the short wind speed at 1h intervals up to 5h based on wavelet packet decomposition, crisscross optimization algorithm and artificial neural networks. In the data pre-processing phase, the wavelet packet technique is used to decompose the original wind speed series into subseries. For each transformed components with different frequency sub-bands, the back-propagation neural network optimized by crisscross optimization algorithm is employed to predict the multi-step ahead wind speed. The eventual predicted results are obtained through aggregate calculation. To validate the effectiveness of the proposed approach, two wind speed series collected from a wind observation station located in the Netherlands are used to do the multi-step wind speed forecasting. To reduce the statistical errors, all forecasting methods are executed 50 times independently. The results of this study show that: (1) the proposed crisscross optimization algorithm has significant advantage over the back-propagation algorithm and particle swarm optimization in addressing the prematurity problems when applied to train the neural network. (2) Compared with the previous hybrid models used in this study, the proposed hybrid model consistently has the minimum mean absolute percentage error regardless of one-step, three-step or five-step prediction.
Early recognition and timely intervention for urosepsis are key to reducing morbidity and mortality. Blood culture has low sensitivity, and a long turnaround time makes meeting the needs of clinical ...diagnosis difficult. This study aimed to use biomarkers to build a machine learning model for early prediction of urosepsis.
Through retrospective analysis, we screened 157 patients with urosepsis and 417 patients with urinary tract infection. Laboratory data of the study participants were collected, including data on biomarkers, such as procalcitonin, D-dimer, and C-reactive protein. We split the data into training (80%) and validation datasets (20%) and determined the average model prediction accuracy through cross-validation.
In total, 26 variables were initially screened and 18 were statistically significant. The influence of the 18 variables was sorted using three ranking methods to further determine the best combination of variables. The Gini importance ranking method was found to be suitable for variable filtering. The accuracy rates of the six machine learning models in predicting urosepsis were all higher than 80%, and the performance of the artificial neural network (ANN) was the best among all. When the ANN included the eight biomarkers with the highest influence ranking, its model had the best prediction performance, with an accuracy rate of 92.9% and an area under the receiver operating characteristic curve of 0.946.
Urosepsis can be predicted using only the top eight biomarkers determined by the ranking method. This data-driven predictive model will enable clinicians to make quick and accurate diagnoses.
End-to-end learning machines enable a direct mapping from the raw input data to the desired outputs, eliminating the need for hand-crafted features. Despite less engineering effort than the ...hand-crafted counterparts, these learning machines achieve extremely good results for many computer vision and medical image analysis tasks. Two dominant classes of end-to-end learning machines are massive-training artificial neural networks (MTANNs) and convolutional neural networks (CNNs). Although MTANNs have been actively used for a number of medical image analysis tasks over the past two decades, CNNs have recently gained popularity in the field of medical imaging. In this study, we have compared these two successful learning machines both experimentally and theoretically. For that purpose, we considered two well-studied topics in the field of medical image analysis: detection of lung nodules and distinction between benign and malignant lung nodules in computed tomography (CT). For a thorough analysis, we used 2 optimized MTANN architectures and 4 distinct CNN architectures that have different depths. Our experiments demonstrated that the performance of MTANNs was substantially higher than that of CNN when using only limited training data. With a larger training dataset, the performance gap became less evident even though the margin was still significant. Specifically, for nodule detection, MTANNs generated 2.7 false positives per patient at 100% sensitivity, which was significantly (p<0.05) lower than the best performing CNN model with 22.7 false positives per patient at the same level of sensitivity. For nodule classification, MTANNs yielded an area under the receiver-operating-characteristic curve (AUC) of 0.8806 (95% CI: 0.8389–0.9223), which was significantly (p<0.05) greater than the best performing CNN model with an AUC of 0.7755 (95% CI: 0.7120–0.8270). Thus, with limited training data, MTANNs would be a suitable end-to-end machine-learning model for detection and classification of focal lesions that do not require high-level semantic features.
•MTANNs yielded higher performance than CNNs for nodule detection and classification.•Deep CNN architectures achieved higher performance than shallow architectures for nodule detection.•CNN architectures with varying depths performed comparably for nodule classification.•MTANNs can achieve desired performance with a smaller training dataset than do the CNNs.•MTANNs tend to learn the appearance of lesion parts, whereas CNNs attempt to learn the lesion appearance as a whole.
With the development of industry, air pollution has become a serious problem. It is very important to create an air quality prediction model with high accuracy and good performance. Therefore, a new ...method of CT-LSTM is proposed in this paper, in which the prediction model is established by combining chi-square test (CT) and long short-term memory (LSTM) network model. CT is used to determine the influencing factors of air quality. The hourly air quality data and meteorological data from Jan. 1, 2017 to Dec. 31, 2018 are used to train the LSTM network model. The data from Jan. 1, 2019 to Dec. 31, 2019 are used to evaluate the LSTM network model. The AQI level of Shijiazhuang of Hebei Province of China from Jan. 1, 2019 to Dec. 31, 2019 is predicted with five methods (SVR, MLP, BP neural network, Simple RNN and this paper's new method). Then, a contrastive analysis of the five prediction results is made. The experimental results show that the accuracy of this new method reaches 93.7%, which is the highest in the five methods and the maximum error of this new method is 1. The correct number of days predicted by this new method is also the highest among the five methods, which is 342 days. The new method also shows good characteristics in MAE, MSE and RMSE, which makes it more accurate for people to predict the AQI level.
Artificial neural networks have great potential for learning and stability in the face of tiny input data changes. As a result, artificial intelligence techniques and modeling tools have a growing ...variety of applications. To estimate a solution for fractal‐fractional differential equations (FFDEs) of high‐order linear (HOL) with variable coefficients, an iterative methodology based on a mix of a power series method and a neural network approach was applied in this study. In the algorithm's equation, an appropriate truncated series of the solution functions was replaced. To tackle the issue, this study uses a series expansion of an unidentified function, where this function is approximated using a neural architecture. Some examples were presented to illustrate the efficiency and usefulness of this technique to prove the concept's applicability. The proposed methodology was found to be very accurate when compared to other available traditional procedures. To determine the approximate solution to FFDEs‐HOL, the suggested technique is simple, highly efficient, and resilient.
The vacuum impregnation (VI) process parameters (vacuum pressure = 20–60 kPa; VI temperature = 35–55°C; concentration of the sucrose solution = 40–60 °Brix; and vacuum process time = 8–24 min) for ...pineapple rings were optimized based on the moisture content (MC), water loss (WL), solids gain (SG), yellowness index (YI), and total soluble solids (TSS) content of pineapple rings using response surface methodology (RSM). A relationship was developed between the process and response variables using RSM and artificial neural network (ANN) techniques. The effectiveness of VI was evaluated by comparing it with the osmotic dehydration (OD) technique. The optimum condition was found to be 31.782 kPa vacuum pressure, 50.441°C solution temperature, and 60 °Brix sucrose concentration for 20.068 min to attain maximum TSS, YI, SG, and WL, and minimum MC of pineapple rings. The R2 values of RSM models for all variables varied between 0.70 and 0.91, whereas mean square error values varied between 0.76 and 71.58 and for ANN models varied between 0.87–0.93 and 0.53–193.78, respectively.
Scanning electron micrographs (SEM) revealed that parenchymal cell rupture was less in VI than in OD. The VI pineapple rings exhibited more pores and high SG, as compared to OD, due to the pressure impregnation. Spectroscopic analysis affirmed that the stretching vibrations of intermolecular and intramolecular interactions were significant in VI as against OD. The VI reduced the drying time by 35% compared to OD, with the highest overall acceptability score and lower microbial load during storage.
Practical Application
Pineapple is a perishable fruit, which necessitates processing for extended shelf life. This study highlights the potential of the vacuum impregnation process as a promising alternative to conventional preservation methods such as osmotic dehydration for pineapples.
This investigation aimed to produce a new composited catalyst from a waste chalk powder, a waste generated by the construction industry, to produce biodiesel from sunflower oil. The waste chalk was ...modified by CoFe2O4 nanoparticles and K2CO3. The surface tests showed that the obtained catalyst has been successfully synthesized with desired surface properties. The surface areas of waste chalk, waste chalk/CoFe2O4, and waste chalk/CoFe2O4/K2CO3 were determined 20.8, 77.8, and 5.8 m2/g, respectively. This indicates that the waste chalk/CoFe2O4/K2CO3 catalyst has a lower surface area due to K2CO3 being placed on the catalyst. Results showed the efficiency of RSM-CCD (R2 = 0.992) compared to ANN (R2 = 0.974). It was shown that a contact time of 180 min, a temperature of 65 °C, a waste chalk/CoFe2O4/K2CO3 mass of 2 wt%, and methanol to oil mole ratio of 15:1 gave the highest efficiency (98.87%) of biodiesel production at the laboratory conditions. The kinetic results of the process showed the energy of activation and frequency factor of 11.8 kJ/mol and 0.78 min−1, respectively. Also, the values of ΔH°, ΔS°, and ΔG° at 65 °C was calculated to be 9010.7 J/mol, −256.3 J/mol and 95.7 kJ/mol, respectively, indicating that the biodiesel production process is endothermic requiring high energy for proceeding. The generated catalyst has an efficiency of over 90% up to 6 steps of reuse. The generated biodiesel was met most of the international standard levels.
Display omitted
•A catalyst was prepared from waste chalk waste and its structure was modified.•Waste chalk/CoFe2O4/K2CO3 was used to produce biodiesel from oil sunflower.•The BET of waste chalk/CoFe2O4/K2CO3 were 5.8 m2/g.•The catalyst had a good efficiency for biodiesel production up to 6 stages.•Ea (11.8 kJ/mol) and frequency factor (0.78 min−1) of the process was obtained.