The demand for energy and cost efficient buildings has made architects and contractors more aware of the resources consumed by the built environment. While the actual economic and environmental costs of future construction can never be completely predicted, energy simulations and cost modeling have become accepted ways to guide the design and construction process by comparing possible outcomes. These tools are now commonplace in the construction industry, and researchers are continuing to develop new and innovative strategies to optimize building design and construction. Previous research has proven that genetic algorithms are effective methods to evaluate and optimize building design in situations that contain a large number of possible solutions. The technique makes a computationally difficult multi-optimization process possible but is still a reactive and time consuming process that focuses on evaluation rather than solution generation. This research presented in this paper builds upon established multi-objective optimization techniques that use an energy simulator to estimate a conceptual building’s energy use as well as construction cost. The study compares simulations of a simplified model of a 3-story inpatient hospital located in Atlanta, Georgia using a defined set of variables. A combined global minimum of annual energy consumption and total construction is sought after using a method that utilizes a genetic algorithm. The second phase of this research uses a modified approach that combines the traditional genetic algorithm with a seeding method that utilizes previous results. A new set of simulations were established that duplicates the initial trials using a slightly modified set of design variables. The simulation was altered, and the phase one trials were utilized as the first generation of simulated solutions. The objective of this thesis is to explore one method of making energy use and cost estimating more accessible to the construction industry by combining simulation optimization and indexing. The results indicate that this study’s proposed augmented approach has potential benefits to building design optimization, although more research is required to validate this hypothesis in its entirety. This study concludes that the proposed approach can potentially reduce the time needed for individual optimization exercises by creating a cumulative, robust catalog of previous computations that will inform and seed future analyses. The research was conducted in five general stages. The first part defines the research problem and scope of research to be conducted. In the second part, the concepts of genetic algorithms and energy simulation are explored in a comprehensive literature review. The remaining parts explain the trial simulations performed in this study. Part three explains the experiment’s methodology, and part four describes the simulation results. The fifth and final part looks at what the possible conclusions that can be made from analyzing the study’s results.

This research presents an investigation to actively improve the rollover stability of articulated heavy vehicles (AHVs) during high speed manoeuvres using anti-roll control systems. A 3-dimensional (3-D) linear yaw/roll model with 5 degrees of freedom is developed. Based on this model a linear quadratic regulator (LQR) controller is designed to improve the rollover stability of a tractor/semi-trailer combination. A design optimization method for AHVs using genetic algorithms (GAs) and multibody vehicle system models is also presented. AHVs have poor manoeuvrability when travelling at low speeds on local roads and city streets. On the other hand, these vehicles exhibit unstable motion modes at high speeds, including jack-knifing, trailer sway and rollover. From the design point of view, the low-speed manoeuvrability and high-speed stability have conflicting requirements on some design variables. The design method based on a GA and a multibody vehicle dynamic package, TruckSim, is proposed to coordinate this trade-off relationship. To test the effectiveness of the design method, a tractor/semi-trailer combination is optimized using the proposed method. It is demonstrated that the proposed design method can be used for identifying desired design variables and predict performance envelopes in the early design stages of AHVs. / UOIT

As a commonly used technique in data preprocessing for machine learning, feature selection identifies important features and removes irrelevant, redundant or noise features to reduce the dimensionality of feature space. It improves efficiency, accuracy and comprehensibility of the models built by learning algorithms. Feature selection techniques have been widely employed in a variety of applications, such as genomic analysis, information retrieval, and text categorization. Researchers have introduced many feature selection algorithms with different selection criteria. However, it has been discovered that no single criterion is best for all applications. We proposed a hybrid feature selection framework called based on genetic algorithms (GAs) that employs a target learning algorithm to evaluate features, a wrapper method. We call it hybrid genetic feature selection (HGFS) framework. The advantages of this approach include the ability to accommodate multiple feature selection criteria and find small subsets of features that perform well for the target algorithm. The experiments on genomic data demonstrate that ours is a robust and effective approach that can find subsets of features with higher classification accuracy and/or smaller size compared to each individual feature selection algorithm. A common characteristic of text categorization tasks is multi-label classification with a great number of features, which makes wrapper methods time-consuming and impractical. We proposed a simple filter (non-wrapper) approach called Relation Strength and Frequency Variance (RSFV) measure. The basic idea is that informative features are those that are highly correlated with the class and distribute most differently among all classes. The approach is compared with two well-known feature selection methods in the experiments on two standard text corpora. The experiments show that RSFV generate equal or better performance than the others in many cases.

Much research has been conducted on the retrieval and classification of web-based information. A big challenge is the performance issue, especially for a classification algorithm returning results for a large set of data that is typical when accessing the Web. This thesis describes a grid-enabled approach for automatic web page classification. The basic approach is first described that uses a vector space model (VSM). An enhancement of the approach through the use of a genetic algorithm (GA) is then described. The enhanced approach can efficiently process candidate web pages from a number of web sites and classify them. A prototype is implemented and empirical studies are conducted. The contributions of this thesis are: 1) Application of grid computing to improve performance of both VSM and GA using VSM based web page classification; 2) Improvement of the VSM classification algorithm by applying GA that uniquely discovers a set of training web pages while also generating a near optimal parameter values set for VSM.

The purpose of this thesis was to determine how shaft stiffness affects clubhead speed and how it alters clubhead orientation at impact. For the first time, a 3D, six-segment forward dynamics model of a golfer and club was developed and optimized to answer these questions. A range of shaft stiffness levels from flexible to stiff were evaluated at three levels of swing speed (38, 45 and 53 m/s). At any level of swing speed, the difference in clubhead speed did not exceed 0.1 m/s across levels of shaft stiffness. Therefore, it was concluded that customizing the stiffness of a golf club shaft to perfectly suit a particular swing will not increase clubhead speed sufficiently to have any meaningful effect on performance. The magnitude of lead deflection at impact increased as shaft stiffness decreased. The magnitude of lead deflection at impact also increased as swing speed increased. For an optimized swing that generated a clubhead speed of 45 m/s, with a shaft of regular stiffness, lead deflection of the shaft at impact was 6.25 cm. The same simulation resulted in a toe-down shaft deflection of 2.27 cm at impact. Using the model, it was estimated that for each centimeter of lead deflection of the shaft, dynamic loft increased by approximately 0.8 degrees. Toe-down shaft deflection had relatively no influence on dynamic loft. For every centimeter increase in lead deflection of the shaft, dynamic closing of the clubface increased by approximately 0.7 degrees. For every centimeter increase in toe-down shaft deflection, dynamic closing of the clubface decreased by approximately 0.5 degrees. The results from this thesis indicate that improvements in driving distance brought about by altering shaft stiffness are the result of altered clubhead orientation at impact and not increased clubhead speed.

In this thesis, a mixed integer linear program is used to formulate the optimization process of a wind farm. As a start point, a grid was superimposed into the wind farm, in which grid points represent possible wind turbine locations. During the optimization process, proximity and wind interference between wind turbines were considered in order to found the power loss of the wind farm. Power loss was analyzed by using wind interference coefficient, which is a function of wind intensity interference factor (WIIF), weibull distribution and power of the wind turbines. Two different programs / Genetic Algorithm and Lingo, were used to solve the MILP optimization formula and results were compared for different cases in the conclusion part.

No description available.

No description available.

A compliant mechanism gains its motion from the deflection of flexible members or the deformation of one portion of materials with respect to other portions. Design and operation of compliant mechanisms are very important, as most of the natural objects are made of compliant materials mixed with rigid materials, such as the bird wings. The most serious problem with compliant mechanisms is their fatigue problem due to repeating deformation of materials in compliant mechanisms. This thesis presents a study on the computational framework for designing a compliant mechanism under fatigue strength control. The framework is based on the topology optimization technique especially ground structure approach (GSA) together with the Genetic Algorithm (GA) technique. The study presented in this thesis has led to the following conclusions: (1) It is feasible to incorporate fatigue strength control especially the stress-life method in the computational framework based on the GSA for designing compliant mechanisms and (2) The computer program can well implement the computational framework along with the general optimization model and the GA to solve the model. There are two main contributions resulting from this thesis: First one is provision of a computational model to design compliant mechanisms under fatigue strength control. This model also results in a minimum number of elements of the compliant mechanism in design, which means the least weight of mechanisms and least amount of materials. Second one is an experiment for the feasibility of implementing the model in the MATLAB environment which is widely used for engineering computation, which implies a wide applicability of the design system developed in this thesis.

No description available.