Document Type : Original Article
Logistics Department, Taizhou Vocational &amp;amp;amp; Technical College, Taizhou, Zhejiang, 318000, ChinaD
Logistics Department, Taizhou Vocational & Technical College, Taizhou, Zhejiang, 318000, China
Self-compacting concrete (SCC), as a liquid aggregate, is suitable for use in reinforced constructions with no need for vibration. SCC utilization has been found in a wide range of projects. Nevertheless, those applications are often limited due to lacking the knowledge about such mixed materials, especially from experimental testing. The factor of Compressive Strength (CS), which is one of the vital mechanical variables in structure immunization, can be computed either through costly tests or predictive models. Intelligent systems can appraise CS based on ingredients’ data fed to the models. This research aims to model the CS of SCC via a machine learning technique of Support Vector Regression (SVR). The Particle Swarm Optimization (PSO) and Henry’s Gas Solubility Optimization (HGSO) have been utilized to optimize the SVR in finding some internal parameters. Different metrics were chosen to evaluate the performance of models. Consequently, the R2 in the testing stage for SVR-HGSO was computed at 0.90 and for SVR-PSO, 0.93. In the calibration phase, the correlation rate was computed at 0.93 for SVR-HGSO with a 3% difference from the SVR-PSO with 0.90.