错误:搜索内容不能为空,请输入英文关键词
错误:关键词超出字数限制,请精简
高级检索

Employing a support vector regression technique coupled with optimizers to estimate the compressive strength of reinforced concrete connections

  • Hongmei Yao

摘要

Keeping stability and preventing collapse in reinforced concrete structures are remarked in most real and research projects. For immunizing structures, finding the logical relationship between mechanical features, like compressive strength and geometrical aspects of structures, would be an essential task. In this regard, the present research has aimed to develop a hybrid model by coupling a machine learning technique of Support Vector Regression with optimization algorithms instead of employing single frameworks of AI-based models in another research. To investigate the capability of models, 171 RC composite samples were considered in simulating the compressive strength of composite joints. As a result, developed models appeared more accurate than a single model of SVR without an optimizer. The capability of hybrid frameworks compared to a single model of SVR has been proved with better results of correlation index of R2 that there has been an average 10 percent difference between the accuracy of hybrid and single model. Also, using optimizers could reduce the RMSE error index average by 56 percent relative to a single SVR.