A Concrete Mix Proportion Design with Manufactured Sand Based on Artificial Neural Networks and Genetic Algorithm
Due to the characteristics of manufactured sand, such as the irregular particle shape, rough surface and high stone powder content, a reasonable mix proportion of concrete with manufactured sand (MSC) is difficult to be obtained.This study proposes a new mix proportion design method of MSC based on artificial neural networks (ANN) and genetic algorithm (GA).The concepts of seven mix proportion parameters of nominal water-powder ratio, equivalent water-powder ratio, sand fraction, average paste thickness (APT), fly ash-powder ratio, slag-powder ratio and stone powder-powder ratio were introduced.The relationship between the seven parameters and the properties (compressive strength and slump) of MSC were established by ANN.Under the premise of meeting the property requirements, GA, which is an optimization technique to solve the multi-criteria problem through the simulated biological evolutionary process, was applied to obtain an optimum mix proportion of MSC with the minimum cost of MSC.The research results proved that using the proposed mix proportion design method could greatly improve design efficiency and reduce cost.
mix proportion design manufactured sand artificial neural networks (ANN) genetic algorithm (GA) average paste thickness (APT)
Tao Ji Qi-ling Zhang
College of Civil Engineering, Fuzhou University, Fuzhou, China
国际会议
福州
英文
111-123
2018-11-04（万方平台首次上网日期，不代表论文的发表时间）