Juan A. Lazzús

Prediction of Flash Point Temperature of Organic Compounds Using a Hybrid Method of Group Contribution + Neural Network + Particle Swarm Optimization

Abstract The flash points of organic compounds were estimated using a hybrid method that includes a simple group contribution method (GCM) implemented in an artificial neural network (ANN) with particle swarm optimization (PSO). Different topologies of a multilayer neural network were studied and the optimum architecture was determined. Property data of 350 compounds were used for training the network. To discriminate different substances the molecular structures defined by the concept of the classical group contribution method were given as input variables. The capabilities of the network were tested with 155 substances not considered in the training step. The study shows that the proposed GCM+ANN+PSO method represent an excellent alternative for the estimation of flash points of organic compounds with acceptable accuracy (AARD1.8%; AAE6.2 K).

A GROUP CONTRIBUTION METHOD TO PREDICT ρ-T-P OF IONIC LIQUIDS

A simple and accurate group contribution method to predict the density of ionic liquids is presented. The proposed method follows the classical architecture of the group contribution concept to evaluate the molar volume at 298.15 K and 0.101 MPa. A set of 210 ionic liquids has been used to obtain the contributions for anion and cation groups. Once the molar volume is evaluated and expressed as density, values at several temperatures and pressures are calculated using a linear model that only uses the reference density value. Density data of 76 ionic liquids (3530 data points) were used to obtain the linear model for wide ranges of temperatures and pressures: 258 to 393 K and 0.09 to 207 MPa. Results show that the new group contribution method (R2 = 0.9687; AARD = 1.93%) and the linear model (R2 = 0.9989; AARD = 0.73%) are accurate enough and give lower deviations than other models available in the literature.

Neural Network Based on Quantum Chemistry for Predicting Melting Point of Organic Compounds

The melting points of organic compounds were estimated using a combined method that includes a backpropagation neural network and quantitative structure property relationship (QSPR) parameters in quantum chemistry. Eleven descriptors that reflect the intermolecular forces and molecular symmetry were used as input variables. QSPR parameters were calculated using molecular modeling and PM3 semi-empirical molecular orbital theories. A total of 260 compounds were used to train the network, which was developed using MatLab. Then, the melting points of 73 other compounds were predicted and results were compared to experimental data from the literature. The study shows that the chosen artificial neural network and the quantitative structure property relationships method present an excellent alternative for the estimation of the melting point of an organic compound, with average absolute deviation of 5%.

Optimization of activity coefficient models to describe vapor-liquid equilibrium of (alcohol + water) mixtures using a particle swarm algorithm

A method to model the vapor-liquid phase based on a particle swarm algorithm is developed in this study. Two activity coefficient models (UNIQUAC and NRTL) were optimized with particle swarm optimization (PSO), and used to describe the isobaric vapor-liquid equilibrium of fifteen binary mixtures containing alcohol + water. The results were compared with the Levenberg-Marquardt algorithm, and show that the PSO algorithm is a good method to correlate and predict the vapor-liquid equilibrium of this type of system.

Neural network-particle swarm modeling to predict thermal properties

Abstract Seven thermal properties: melting point temperature, boiling point temperature, critical temperature, autoignition temperature, flash point temperature, lower flammability limit temperature and upper flammability limit temperature, were estimated using a hybrid method that includes an artificial neural network (ANN) with particle swarm optimization (PSO). A database of 530 substances was used in the training of this hybrid algorithm. To discriminate the different substances the molecular structures were given as input parameters. Different topologies of the neural network were studied and the best architecture was determined. The optimal condition of the network was obtained adjusting the PSO parameters by trial-and-error. The results show that the proposed ANN+PSO method represent an excellent alternative for the estimation of thermophysic properties with acceptable accuracy.

Estimation of global solar radiation over the city of La Serena (Chile) using a neural network

An artificial neural network for the estimation of hourly global solar radiation in La Serena (Chile), was developed using data measured from a meteorological station. La Serena city (29°54′ S, 71°15′ W) is located in the bay area at south of the hyper-arid Atacama Desert. In this study, 25123 data points of global solar radiation of 5 years (2001–2005) were used to train the network and then 7618 data points of global solar radiation not used in the training process were predicted (years 2006 and 2007). The meteorological data used in the model were: wind speed, relative humidity, air temperature, and soil temperature. The results were compared with the real data and other models available in the literature, and shows that the neural network obtained can be properly trained and can estimate the hourly global radiation with acceptable accuracy.

HYBRID NEURAL NETWORK{PARTICLE SWARM ALGORITHM TO DESCRIBE CHAOTIC TIME SERIES

An artificial neural network (ANN) based on particle swarm optimiza- tion (PSO) was developed for the time series prediction. This hybrid ANN+PSO algorithm was applied on Mackey-Glass series in the short-term prediction x(t + 6) and the long-term prediction x(t + 84), from the current value x(t) and the past values: x(t 6), x(t 12), x(t 18). Four cases were studied, alternating the time- delay parameter as 17 or 30. Also, the first four largest Lyapunov exponents were obtained for different time-delay. Simulation shows that this ANN+PSO method is a very powerful tool for making prediction of chaotic time series.

Forecasting the Dst index using a swarm‐optimized neural network

A hybrid technique that combines an artificial neural network with a particle swarm optimization (ANN+PSO) was used to forecast the disturbance storm time (Dst) index from 1 to 6 hours ahead. Our ANN was optimized by PSO to update ANN weights and to predict the short-term Dst-index using past values as input parameters. The database used contains 233,760 hourly data from 01 January 1990 to 31 August 2016, considering storms and quiet period, grouped into three data sets: learning set (with 116,880 hourly data points), validation set (with 58,440 data points), and testing set (with 58,440 data points). Several ANN topologies were studied, and the best architecture was determined by systematically additing neurons and evaluating the root mean square error (RMSE) and the correlation coefficient (R) during the training process. These results show that the hybrid algorithm is a powerful technique for forecasting the Dst index a short time in advance like t + 1 to t + 3, with RMSE from 3.5nT to 7.5nT, and R from 0.98 to 0.90. However, t + 4 to t + 6 predictions become slightly more uncertain, with RMSE from 8.8nT to 10.9nT, and R from 0.86 to 0.79. Additionally, an exhaustive analysis according to geomagnetic storm magnitude was conduced. In general, the results show that our hybrid algorithm can be correctly trained to forecast the Dst index with appropriate precision and that Dst past behavior significantly affects adequate training and predicting capabilities of the implemented ANN.

New group contribution method for the prediction of normal melting points

The melting point of organic compounds was estimated using a simple group contribution method. The optimum parameters of this new method were obtained using particle swarm optimization in a multivariate linear regression. The melting temperatures of 250 pure compounds were predicted, and the results were compared with experimental data and other models available in the literature. Compared to the currently used group contribution methods, the new method makes significant improvements in accuracy and applicability of this important property. The study shows that the proposed method presents an excellent alternative for the estimation of the melting temperature of organic compounds (AARD of 10%) from the knowledge of the molecular structure.

Impact of noise on a dynamical system: prediction and uncertainties from a swarm-optimized neural network

An artificial neural network (ANN) based on particle swarm optimization (PSO) was developed for the time series prediction. The hybrid ANN+PSO algorithm was applied on Mackey-Glass chaotic time series in the short-term x(t + 6). The performance prediction was evaluated and compared with other studies available in the literature. Also, we presented properties of the dynamical system via the study of chaotic behaviour obtained from the predicted time series. Next, the hybrid ANN+PSO algorithm was complemented with a Gaussian stochastic procedure (called stochastic hybrid ANN+PSO) in order to obtain a new estimator of the predictions, which also allowed us to compute the uncertainties of predictions for noisy Mackey-Glass chaotic time series. Thus, we studied the impact of noise for several cases with a white noise level (σ N) from 0.01 to 0.1.