CAJ | 학술논문

混凝土抗压强度预测是一个动态的系统工程，其精度受到多种高维非线性、随机性因素的影响。为有效提高混凝土抗压强度的预测精度，在分析支持向量机的基础上，构建了基于灰色关联支持向量机的混凝土抗压强度预测模型。该模型基于灰色关联分析确定混凝土抗压强度的主导因素，然后通过支持向量机建立其与变量之间的非线性映射关系，同时利用网格搜索算法对支持向量机进行参数寻优。仿真结果表明：与单纯支持向量机和BP神经网络模型预测结果相比，基于灰色关联支持向量机的预测模型更为有效可靠，为提高混凝土抗压强度预测精度提供了新的途径。
혼응토항압강도예측시일개동태적계통공정，기정도수도다충고유비선성、수궤성인소적영향。위유효제고혼응토항압강도적예측정도，재분석지지향량궤적기출상，구건료기우회색관련지지향량궤적혼응토항압강도예측모형。해모형기우회색관련분석학정혼응토항압강도적주도인소，연후통과지지향량궤건립기여변량지간적비선성영사관계，동시이용망격수색산법대지지향량궤진행삼수심우。방진결과표명：여단순지지향량궤화BP신경망락모형예측결과상비，기우회색관련지지향량궤적예측모형경위유효가고，위제고혼응토항압강도예측정도제공료신적도경。
The prediction of concrete compressive strength was dynamic system engineering, and its accu-racy was affected by a variety of high dimensional nonlinear, random factors. To effectively improve the prediction accuracy of concrete compressive strength, a prediction model of concrete compressive strength based on grey relational-support vector machine ( GR-SVM) was constructed on the basis of the analysis of support vector machine ( SVM) . The model based on grey relational analysis identified the main factors affecting the compressive strength of concrete, and established the nonlinear mapping relationship be-tween compressive strength and variables through the SVM. The grid search algorithm was used to opti-mize the parameters of SVM. Simulation results showed that compared with single SVM and BP ANN, the prediction results based on GR-SVM forecasting model was more effective and reliable, and a new way would be introduced to improve the prediction accuracy of concrete compressive strength.