•Extreme learning machine (ELM) was used to predict the compressive strength of High strength concrete.•The developed ELM model is compared with BP model.•The ELM model has good prediction accuracy ...and fast learning speed.•The results show the potential use of ELM for predicting the compressive strength.
Compressive strength is a major and significant mechanical property of concrete which is considered as one of the important parameters in many design codes and standards. Early and accurate estimation of it can save in time and cost. In this study, extreme learning machine (ELM) was used to predict the compressive strength of high-strength concrete (HSC). ELM is a relatively new method for training artificial neural networks (ANN), showing good generalization performance and fast learning speed in many regression applications. ELM model was developed using 324 data records obtained from laboratory experiments. The compressive strength was modeled as a function of five input variables: water, cement, fine aggregate, coarse aggregate, and superplasticizer. The performance of the developed ELM model was compared with that of ANN model trained by using back propagation (BP) algorithm. The simulation results show that the proposed ELM model has a strong potential for predicting the compressive strength of HSC.
•Engine load, palm oil ratio and injection advance have been chosen as input parameters.•Proposed RSM and ANN model are capable for mapping engine performance-emission paradigms with high ...accuracy.•RSM was applied to optimize the engine performance and exhaust emissions.•Trade-off study based on RSM technique showed the optimal engine operating condition for biodiesel/diesel blends.
Engine performance and emission characteristics of palm oil-diesel blends tested on single-cylinder diesel engine by several engine loads and injection advances. Exhaust emissions and smoke were recorded using MRU Delta 1600L and MRU Optrans 1600 model gas analyzer, respectively. Brake thermal efficiency (BTE), exhaust gas temperature (EGT), carbon monoxide (CO), hydrocarbon (HC), smoke and nitrogen oxides (NOx) were optimized as output factors considering engine load, injection advance and palm oil percentage as input variables using response surface methodology (RSM) and artificial neural network (ANN). The developed ANN and RSM models showed superior predictive certainty with big R2 (correlation coefficient) values. The RSM models showed better performance and have higher R2 values than ANN models. The developed RSM model has R2 values over 0.90 while the R2 values of ANN model are between 0.88 and 0.95. The values of mean relative error (MRE) and root mean square error (RMSE) for all the responses were low. Optimum responses were found by 69.11%, 196.25 ppm, 0.126%, 189.764 ppm, 155.49 ℃ and 30.75%, respectively for smoke, NOx, CO, HC, EGT and BTE with optimum operating factors as 17.88% palm oil percentage, 35 °CA injection advance and 780-watt engine load. The applied models gave good results that are beneficial for estimating and optimizing the engine performance and emission characteristics.
Particle tracking is a powerful biophysical tool that requires conversion of large video files into position time series, i.e., traces of the species of interest for data analysis. Current tracking ...methods, based on a limited set of input parameters to identify bright objects, are ill-equipped to handle the spectrum of spatiotemporal heterogeneity and poor signal-to-noise ratios typically presented by submicron species in complex biological environments. Extensive user involvement is frequently necessary to optimize and execute tracking methods, which is not only inefficient but introduces user bias. To develop a fully automated tracking method, we developed a convolutional neural network for particle localization from image data, comprising over 6,000 parameters, and used machine learning techniques to train the network on a diverse portfolio of video conditions. The neural network tracker provides unprecedented automation and accuracy, with exceptionally low false positive and false negative rates on both 2D and 3D simulated videos and 2D experimental videos of difficult-to-track species.
•A thermo-kinetic study is performed of high ash sewage sludge.•ANN, for the first time, is applied for sewage sludge pyrolysis process.•Kinetics is estimated using model-free methods as a function ...of conversion.•Results provide reference to promote pilot scale high-ash sewage sludge.
Pyrolysis of high-ash sewage sludge (HASS) is a considered as an effective method and a promising way for energy production from solid waste of wastewater treatment facilities. The main purpose of this work is to build knowledge on pyrolysis mechanisms, kinetics, thermos-gravimetric analysis of high-ash (44.6%) sewage sludge using model-free methods & results validation with artificial neural network (ANN). TG-DTG curves at 5,10 and 20 °C/min showed the pyrolysis zone was divided into three zone. In kinetics, E values of models ranges are; Friedman (10.6–306.2 kJ/mol), FWO (45.6–231.7 kJ/mol), KAS (41.4–232.1 kJ/mol) and Popescu (44.1–241.1 kJ/mol) respectively. ΔH and ΔG values predicted by OFW, KAS and Popescu method are in good agreement and ranged from (41–236 kJ/mol) and 53–304 kJ/mol, respectively. Negative value of ΔS showed the non-spontaneity of the process. An artificial neural network (ANN) model of 2 * 5 * 1 architecture was employed to predict the thermal decomposition of high-ash sewage sludge, showed a good agreement between the experimental values and predicted values (R2 ⩾ 0.999) are much closer to 1. Overall, the study reflected the significance of ANN model that could be used as an effective fit model to the thermogravimetric experimental data.
The papers in this special section focus on the technology and applications supported by deep learning. Deep learning is a growing trend in general data analysis and has been termed one of the 10 ...breakthrough technologies of 2013. Deep learning is an improvement of artificial neural networks, consisting of more layers that permit higher levels of abstraction and improved predictions from data. To date, it is emerging as the leading machine-learning tool in the general imaging and computer vision domains. In particular, convolutional neural networks (CNNs) have proven to be powerful tools for a broad range of computer vision tasks. Deep CNNs automatically learn mid-level and high-level abstractions obtained from raw data (e.g., images). Recent results indicate that the generic descriptors extracted from CNNs are extremely effective in object recognition and localization in natural images. Medical image analysis groups across the world are quickly entering the field and applying CNNs and other deep learning methodologies to a wide variety of applications.
The accurate estimations of the failure probability with low-computational burden play a vital role in structural reliability analyses. Due to high-calculation cost and time-consuming Monte Carlo ...simulation (MCS), this paper focused on developing a novel enhanced MCS approach with an advanced machine learning method, namely hybrid enhanced MCS (HEMCS), for achieving accurate approximation of failure probability with high-efficiency computations. The failure probability is approximated by a probabilistic model using a scalar factor less than one which is multiplied on the capacity term of performance function. An adaptive input for scalar factor is proposed by the coefficient of variation of failure probability for active region in the training process of artificial neural network (ANN) with multilayer back-propagation algorithm. Four analytical optimization training approaches, active region and number of hidden nodes are discussed for accurate approximation of ANN models. The results of the HEMCS are compared with several analytical reliability methods for numerous engineering problems. Laminated composite plate and turbine bladed disk are selected to illustrate the capability of HEMCS for approximation of failure probability. The proposed method provides higher flexibility for the prediction of failure probability when compared to the enhanced MCS which also offers significantly more accurate and computationally efficient results for high-nonlinear problems.
•Enhanced MCS with Machine Learning method was proposed for estimating accurate failure probability.•An active artificial intelligent model was introduced for efficient structural reliability method.•Random sampling and active artificial neural network is proposed for complex reliability problems.•Engineering cases verified accuracy, simplicity, efficiency and capability using hybrid method.
In the present study, the performance and emission parameters of a single cylinder diesel engine powered by biodiesel-diesel fuel blends were predicted by Artificial Neural Network (ANN) and ...optimized by Response Surface Methodology (RSM). The data to be used for ANN and RSM applications were obtained by using biodiesel/diesel fuel blends at different engine loads and various injection pressures. ANN model has been developed to predict the outputs such as brake thermal efficiency (BTE), brake specific fuel consumption (BSFC), exhaust gas temperature (EGT), nitrogen oxides (NOx), hydrocarbons (HC), carbon monoxide (CO) and smoke regarding engine load, biodiesel ratio and injection pressure. A feed-forward multi-layer perceptron network is used to show the correlation among the input factors and the output factors. The RSM is applied to find the optimum engine operating parameters with the purpose of simultaneous reduction of emissions, EGT, BSFC and increase BTE. The obtained results reveal that the ANN can correctly model the exhaust emission and performance parameters with the regression coefficients (R2) between 0.8663 and 0.9858. It is seen that the maximum mean relative error (MRE) is less than 10%, compared with the experimental results. The RSM study demonstrated that, biodiesel ratio of 32% with 816-W engine load and 470 bar injection pressure are the optimum engine operating parameters. It is found that the ANN with RSM support is a good tool for predict and optimize of diesel engine parameters powered with diesel/biodiesel mixtures.
An artificial Neural Network (NNW) is designed to serve as a surrogate model of micro-scale simulations in the context of multi-scale analyses in solid mechanics.
The design and training ...methodologies of the NNW are developed in order to allow accounting for history-dependent material behaviors. On the one hand, a Recurrent Neural Network (RNN) using a Gated Recurrent Unit (GRU) is constructed, which allows mimicking the internal variables required to account for history-dependent behaviors since the RNN is self-equipped with hidden variables that have the ability of tracking loading history. On the other hand, in order to achieve accuracy under multi-dimensional non-proportional loading conditions, training of the RNN is achieved using sequential data. In particular the sequential training data are collected from finite element simulations on an elasto-plastic composite RVE subjected to random loading paths. The random loading paths are generated in a way similar to a random walking in stochastic process and allow generating data for a wide range of strain-stress states and state evolution.
The accuracy and efficiency of the RNN-based surrogate model is tested on the structural analysis of an open-hole sample subjected to several loading/unloading cycles. It is shown that a similar accuracy as with a FE2 multi-scale simulation can be reached with the RNN-based surrogate model as long as the local strain state remains in the training range, while the computational time is reduced by four orders of magnitude.
•An artificial Neural Network (NNW) is designed as surrogate in multiscale simulation.•The NNW design and training are developed for history-dependent material behaviors.•Recurrent Neural Network (RNN) with hidden variables can track loading history.•A random walk is used to generate complex multi-dimensional training data.•The cyclic loading of a composite open-hole sample is modeled using the surrogate.
In this study, a new hybrid model, namely the Electromagnetism-based Firefly Algorithm - Artificial Neural Network (EFA-ANN), is proposed to forecast the energy consumption in buildings. The model is ...applied to evaluate the heating load (HL) and cooling load (CL) using two given datasets. Each dataset was obtained by monitoring the effect of the façade system and dimensions of the building, respectively, on energy consumption. The performance of EFA-ANN is validated by comparing the obtained results with other methods. It is shown that EFA-ANN provides a faster and more accurate prediction of HL and CL. A sensitivity analysis is performed to identify the impact of each input on the energy performance of the building. From the results of this study, it is evident that EFA-ANN can assist civil engineers and construction managers in the early designs of energy-efficient buildings.
•Early prediction of HL and CL in buildings can help to design energy-efficient buildings.•A hybrid model is proposed to predict the HL and CL in buildings.•The model is based on ANN model and EFA optimization method.•Two different datasets were used to validate the model.•The proposed approach can assist in the early stage of building design.