Sharif Digital Repository

Application of advanced process control methods in a chemical plant requires a model which represents the transient behaviour of the plant. However, only a few plant-wide identification methods have been proposed for chemical processes. In this paper, an objectoriented process identifier (OPI) framework has been introduced for plant-wide identification to show how using object-oriented design can provide a general plant-wide modelling framework for various systems including chemical processes. The interactions between units are considered in plant-wide identification methods. As a result, the obtained models can predict the system behaviour in new operating conditions better than those... 

Heat exchanger network synthesis (HENS) is one of the most efficient process integration tools to save energy in chemical plants. In this work, a new optimization framework is proposed for the synthesis of HENS, based on a genetic algorithm (GA) coupled with a commercial process simulator through the ActiveX capability of the simulator. The use of GA provides a robust search in complex and nonconvex spaces of mathematical problems, while the use of a simulator facilitates the formulation of rigorous models for different alternatives. To include the most common heat exchanger structures in the model, a promising superstructure has been used. Allowing nonisothermal mixing of streams in the new... 

Background and objective: The human lungs exchange air with the external environment via the conducting airways. The application of an anatomically accurate model of the conducting airways can be helpful for simulating gas exchange and fluid distribution throughout the bronchial tree in the lung. Methods: In the current study, Lindenmayer system (L-system) has been formulated to generate the bronchial tree structure in a human lung. It has been considered that the structure of the bronchial tree is divided into two main segments: 1) The central airways (from the trachea to segmental bronchi) and 2) the dichotomous structure (from segmental bronchi to terminal bronchioles). Two sets of... 

The curve of exhaled inert gas concentration against exhaled volume is called gas washout curve. The slope at the end part of gas washout curve (Sn) is a measure of structural abnormalities. Sn depends on the spatial concentration distribution and dynamic of gas washout, which depends on several mechanisms including asymmetry of airways, nonhomogeneous ventilation, sequential emptying, and gas exchange with blood. Due to a large number of airways in human lungs, using simplified models is inevitable. On the other hand, these simplified models cannot capture some of the mentioned mechanisms and subsequently were not able to predict experimental trend of change in Sn with breath number in... 

The curve of exhaled inert gas concentration against exhaled volume is called gas washout curve. The slope at the end part of gas washout curve (Sn) is a measure of structural abnormalities. Sn depends on the spatial concentration distribution and dynamic of gas washout, which depends on several mechanisms including asymmetry of airways, nonhomogeneous ventilation, sequential emptying, and gas exchange with blood. Due to a large number of airways in human lungs, using simplified models is inevitable. On the other hand, these simplified models cannot capture some of the mentioned mechanisms and subsequently were not able to predict experimental trend of change in Sn with breath number in... 

Ensemble Kalman filter, EnKF, as a Monte Carlo sequential data assimilation method has emerged promisingly for subsurface media characterization during past decade. Due to high computational cost of large ensemble size, EnKF is limited to small ensemble set in practice. This results in appearance of spurious correlation in covariance structure leading to incorrect or probable divergence of updated realizations. In this paper, a universal/adaptive thresholding method is presented to remove and/or mitigate spurious correlation problem in the forecast covariance matrix. This method is, then, extended to regularize Kalman gain directly. Four different thresholding functions have been considered... 

In this paper sustainability analysis (exergy, environmental and economic) and multi-objective optimization for an aromatic plant are provided and interactions between decision variables are discussed. Environmental evaluation shows that the cancer human toxicity and global warming are the most important environmental concerns and the weight of EIs (environmental impacts) are mainly due to process wastes. The optimizations results demonstrate parameters like reactor temperature have a wide range in the optimizations while some variables, such as extraction unit variables have the same value. Utility EI reduction occurred in economic and exergy optimizations rather than environmental... 

In many cases, multiple-fault diagnosis of plant-wide systems based on steady-state data is impossible. To solve this problem, a new diagnosis strategy based on neural networks has been proposed. In the suggested framework, the neural network is used as the diagnoser trained by a hybrid set of steady and dynamic characteristic data of the system. The dynamic characteristic data include overshoot and undershoot values of measured variables and their corresponding occurrence times. To evaluate its performance, the proposed scheme was used in the diagnosis of the concurrent faults of the Tennessee Eastman (TE) process. Various combinations of concurrent faults were considered in this... 

In this work, optimization of metalorganic chemical vapor deposition process for uniform layer thickness with especial attention to reactor geometric parameters as decision variables is presented. A numerical solution to a steady thermal flow associated with multi-species and chemical reactions in atmospheric pressure axisymmetrical rotating disk reactor by the CFD technique is obtained. Such a simulation is conducted on the assumption that the low Mach number flow is laminar. Then the validation of the numerical results with the benchmark solutions is conducted. Finally, integrating the CFD simulator with an optimization program, based on the Nelder-Mead algorithm, as a new approach is... 

The Ensemble Kalman Filter (EnKF) is a Monte Carlo based method to assimilate the measurement data sequentially in time. Although, EnKF has some advantages over the other Kalman based methods to deal with non-linear and/or high dimensional reservoir models, it also suffers from deficiency in estimation of non-Gaussian parameters. In this work, we propose a re-parameterization method to handle non-Gaussian parameters via Ensemble Kalman Filter framework. For this purpose, concept of cumulative distribution function transformation has been used. In addition, the statistics of prior information have been aggregated in the state vector in order to capture the prior uncertainties of non-Gaussian... 

An ensemble-based, sequential assimilation procedure is developed and successfully applied to estimate absolute and relative permeabilities jointly under multi-phase (oil, gas and water) flow condition in the porous media. Two-phase oil-water and gas-oil relative permeabilities are represented by power-law models, and Stone's Model II is used to calculate three-phase oil relative permeability. Prior absolute permeability field is also generated with isotropic Gaussian covariance. The proposed method is validated by a twin numerical setup with three different case studies, in which a synthetic 2D reservoir under multi-phase flow condition is considered. In the first case, there is... 

Ensemble-based assimilation methods are the most promising tools for dynamic characterization of reservoir models. However, because of inherent assumption of Gaussianity, these methods are not directly applicable to channelized reservoirs wherein the distribution of petrophysical properties is multimodal. Transformation of facies field to level set functions have been proposed to alleviate the problem of multimodality. Level set representation ensures that the estimated fields are facies realizations as well as no modification of the assimilation method is required. Moreover, due to the complexity of the history matching problem in the channelized reservoirs, most researchers conventionally... 

Neural networks have been used for process fault diagnosis. In this work, the cluster analysis is used to design a structurally optimized Probabilistic Neural Network. This network is called the Clustered-Based Design Probabilistic Neural Network (CBDPNN). The CBDPNN is capable of diagnosing the faults of nonlinear processes operating over several regions. The performance and training status of the proposed CBDPNN is compared to a conventional Multi-Layer Perceptron (MLP) that is trained on the whole operating region. Simulation results indicate that both schemes have the same performance, but, the training of CBDPNN is much easier than the conventional MLP, although it has about 50% more... 

On-line optimization for the base case of the Tennessee Eastman (TE) challenge problem is presented; furthermore, an interesting operating condition near base case has been obtained, which results in a lower cost function. The proposed method is based on the estimation of the internal states and the time varying parameters of the process model based on an Extended Kalman filter. The sequential quadratic programming method has been used to accomplish the non-linear programming (NLP) task. The objective function is the operational cost while the constraints are the reactor mass balance, safe operation of the process equipment, and the conditions that satisfy the product quality and flow. The... 

A dynamical hybrid observer is proposed for online reconstruction of the active mode and continuous states of Pressure Swing Adsorption (PSA) processes as an integral part of a hybrid control system. A mode observer is designed for estimation of the active mode, and the continuous spatial profiles are estimated by a Distributed and Decentralized Switching Kalman Filter. The proposed hybrid observer has been applied, in silico, for a two-bed, six-step PSA process. The active mode of the process along with the continuous spatial profiles of its adsorption beds have been estimated quite accurately based on very limited number of noise corrupted temperature and pressure measurements. © 2017  

In this study, coupled Generative Adversarial and Auto-Encoder neural networks have been used to reconstruct realizations of three-dimensional porous media. The gradient-descent-based optimization method is used for training and stabilizing the neural networks. The multi-scale reconstruction has been conducted for both sandstone and carbonate samples from an Iranian oilfield. The sandstone contains inter and intra-grain porosity. The generative adversarial network predicts the inter-grain pores and the auto-encoder provides the generative adversarial network result with intra-grain pores (micro-porosity). Different matching criteria, including porosity, permeability, auto-correlation...