| Summary: | 碩士 === 國立臺灣科技大學 === 營建工程系 === 107 === Concrete is the most widely applied material in the construction. In order to measure the performance of concrete mixtures in designing structures, compressive strength, workability and durability are used as the most common properties. Traditional method of mix design in laboratory is time and cost consuming; therefore, developing an Artificial Intelligence model to estimate concrete properties can provide the huge advantages for concrete industry. Numerous previous studies proposed different AI models for modelling concrete properties; however, most of them emphasized on two concrete properties namely compressive strength and workability, there have been no study considered the compressive strength, workability, durability properties and economic factor simultaneously. The aim of this research is to establish a model for obtaining the most economical mix design that meets those three concrete properties above using AI-Integrated approach namely SOS-LSSVM. The dataset for developing model are collected from various previous researches. The result shows that presented model produced the optimum mixture of High Performance Concrete, which has the lowest price and satisfy the desired concrete properties in term of Compressive Strength, Workability and Durability.
|
|---|
