Evolution of tar ball aggregates in Caspian Sea: Implications of connectionist tools linked with image analysis

Abstract

Intensive petroleum activities in the Caspian Sea lead to a considerable oil pollution in the various parts of this region. Tar balls stranded on the south coast of the Caspian Sea evidently cause (and accelerate) this serious environmental threat. In this study, a poloxamine copolymeric surfactant is used to agglomerate tar balls in a system containing the solvent, antisolvent, and Caspian Sea tar balls. This research work is conducted to determine the cumulative size distribution of tar balls in the presence of a poloxamine copolymer. The evolution of tar ball aggregates is experimentally and theoretically investigated and the steady‐state size distribution of the tar balls is estimated by Back Propagation Artificial Neural Network (ANN‐BP) and a new evolutionary algorithm, called Imperialist Competitive Algorithm (ICA). The average size of tar ball aggregates increases with time and reaches a maximum value. It then declines and eventually approaches a steady‐state condition so that an almost constant size for the particles is attained. The statistical analysis shows a good agreement between the predicted values (obtained by the neural network combined with ICA, known as ANN‐ICA) and the real data for the steady‐state size distribution of the aggregates. This research study can considerably help to further understand evolution of tar balls (that hit the beaches) and consequently to find effective and economical ways for tar balls removal from the beaches. © 2018 American Institute of Chemical Engineers Environ Prog, 37: 1901–1907, 2018

Hybrid smart model to determine concentration of acidic gases in absorption tower of sweetening process

“…Each country is made up of a matrix having a 1 × N array of optimization parameters ( p i ), which is shown by Equation (3). In this method, each optimization problem has N variables, as presented below [ 25,41 ] : The countries are made up of the colonies and imperialists. Through this algorithm, some of the best countries are picked as the imperialists ( N imp ), featuring low cost, while the others are selected as the colonies ( N col ).…”

“…The last countries are linked to one of the imperialists. The cost function ( f ) applied to each variable is defined as follows [ 25,41 ] : The entire principal colonies are assigned to the corresponding imperialists on the basis of their power. The power of each country is inversely related to its cost.…”

“…Given the dimensionless cost for all the imperialists, the dimensionless power of each imperialist is defined by the following relationship [ 25,41 ] : The number of colonies of the n th imperialist is then obtained by the following expression [ 41 ] : where N.C n introduces the initial number of the colonies that are randomly selected and are assigned to the n th imperialist country of the n th imperialism. The stronger imperialisms own a greater number of colonies, and the weaker ones have a lower number of colonies.…”

“…[ 25 ] For instance, Sayyad Amin et al implemented this methodology to model the data collected from an experimental work on tar ball aggregation. [ 25 ] This info/data contains vital tar ball aggregate characteristics, including cumulative size distribution of tar ball aggregates under steady‐state condition. As the experimental data were collected in a long time period, an ANN‐ICA model and a conventional ANN were employed to estimate the objective function (e.g., size distribution).…”

“…It was also found that the ANN‐ICA model is more effective and precise, compared to the stand‐alone ANN model. [ 25 ]…”