Sharif Digital Repository

This research presents a modified algorithm for constrained optimization of random simulation models. One output is selected as objective to be minimized, while other must satisfy the given threshold value. Moreover, the simulation inputs must be integer and satisfy linear or nonlinear constraints. The research applies a sequentialized experimental design to specify the simulation input combinations, Kriging (or spatial correlation modeling) to analyze the global simulation input/output data resulting from these designs, and nonlinear programming to estimate the optimal solution from the Kriging metamodels. In addition, a simulation model is developed for different inventory planning... 

In recent years, optimization via simulation for the systemswhose objective function has stochastic characteristic and doesn’t explicitly exist in closed form, has attracted considerable interest.Simulation of this kind of systems at times may be veryexpensive. In this research, the constraint simulation optimization problem is considered for solving problems with stochastic features based on metamodels. For this purpose, stochastic Kriging is used as a metamodel. In this method, first, a few feasible points in the solution space are identified by thek-optimal design of experiment and then the simulation runs are performed. In the next step, a metamodel is fitted to all the stochastic... 

Many simulation optimization problems are defined in random settings and their inputs have uncertainty. Therefore, in defining an optimal solution for these problems, uncertainties should be taken into account. The primary way of dealing with this , is Robust Optimization which finds solution immune to these changing settings. Aiming at finding a new approach for simulation optimization problems, this study investigates these uncertainties and robust methods. In the optimization problem, the goal and constraints are considered with separate risk measures and a related problem is defined as follows: Minimizing the weighted sum of all risks subject to the problem constraints. To solve the... 

In recent years, simulation optimization methods have been developed to solve complicated problems that cannot be solved by mathematical programming methods. In simulation optimization methods, first the problem is modeled by simulation tools and then by applying optimization tools the optimal combination of input variables that optimizes the simulation output is determined. Although simulation optimization has attracted researchers’ attention in recent years, most of the works presented do not consider uncertainty in simulation models. This becomes our motivation in this study to develop uncertain methods in metamodel-based simulation optimization based on minimax methods that are... 

Digital computer simulation and employing the concepts of the experimental design, and other analytical tools for evaluating its output have attracted many of the scientists and researchers interests in the recent decades. The importance of this topic has been increasing and more related analytical tools have been introduced to the scientific literature. One of the powerful tools for simplifying accelerating the optimization process of simulation results, are the use of the metamodels. Use of these powerful tools becomes more eminent when the simulation runs are expensive. By the use of the metamodels the needs of conducting sampling for obtaining some more new points from direct simulating... 

Worldwide coverage, online accessibility and unique accuracy of spatial and temporal satellite data such as rainfall, has encouraged researchers to apply this information in studies such as water resources engineering, hydrologic modeling, drought studies and flood forecast. This research aims to introduce a method that reduces the estimation error of rainfall data derived from a satellite product. “Satellites rainfall products” such as PERSIANN (Precipitation Estimation from Remotely Sensed Information using Artificial Neural Networks), which has been used in this research, has significant error in comparison with geo-prediction precipitation, due to stochastic and systematic errors. In... 

In this work, we propose a metamodel-based simulation optimization algorithm using a novel hybrid sequential experimental design. The algorithm starts with a metamodel construction phase in which at each stage, a sequential experimental design is used to select a new sample point from the search space using a hybrid exploration-exploitation search strategy. Based on the available design points at each stage, a metamodel is constructed using Artificial Neural Network (ANN) and Kriging interpolation techniques. The resulting metamodel is then used in the optimization process to evaluate new solutions. We use Imperialist Competitive Algorithm (ICA) which is a powerful population-based... 

simulation models are free of any restricting assumptions which may normally be considered in a stochastic system, so simulation is considered as one of the most popular tools that can be applied toward analysis of behavior of stochastic systems which are complex.
In order to analyze such problems and determine the best combination of input variables to optimize the system performance criterion, simulation optimization methods were introduced. The most important issue in these problems is that simulation models are usually considered as the black-box models in which, the output function is not usually expressed explicitly.
This work reviews different methods which developed... 

In this thesis we propose a new metamodel based simulation optimization algorithm using sequential design of experiments. The main objective is to have a new method which can be used without deep knowledge of different kinds of metamodels, optimization techniques and design of experiments. The method uses three metamodels simulataneously and gradually adapts to the best metamodel. In each iteration, some points are chosen as candidates for future simulation. These points are ranked based on the quality of metamodel prediction and their placement among simulated points, the best point will be chosen for simulation. Comparing the proposed algorithm with some of the popular simulation... 

Control Charts are the most popular monitoring tools for profiles. The time that a control chart gives an out-of-control signal is not the real time of change. The actual time of change is called the change point. This study suggests two new algorithms to find the change (shift) in parameters of nonlinear profiles. First, using ordinary Kriging method, new points are estimated. Then, with the help of Bernoulli hypothesis test, the probability of detecting the change for new points is tested. Nonlinear profiles in this study follow the exponential family of distributions; in particular, Exponential, Poison and Gaussian distribution structures are used as nonlinear profiles. The proposed... 

Additive manufacturing (AM), also known as Rapid prototyping or D printing, is a new technology for the manufacturing of the physical parts through an additive manner. In the AM process, the orientation pattern of the part is one of the most important factors that significantly affects the product properties such as the build time, the surface roughness, the mechanical strength, the wrinkling, and the amount of support material. The build time and the surface roughness are the more imperative criteria than others that can be considered to find the optimum orientation of parts. In this research, Two method is used to optimize part build orientation (PBO). In the first method a new combined...