车辆排气噪声品质的支持向量回归预测  

Sound Quality of Vehicle Exhaust Noise Prediction by Support Vector Regression

作　　者：赵华杰[1] ZHAO Hua-jie(Xi'an Institute of Finance and Economics, Xingzhi school, Xi'an 710038)

机构地区：[1]西安财经学院行知学院,西安710038

出　　处：《环境技术》2018年第6期99-102,共4页Environmental Technology

摘　　要：通过比较法测试得到38种车辆噪声的满意度评价,考察并选取五个参数作为描述车辆排气噪声声音品质的客观心理声学参数,应用支持向量回归机建立了客观心理声学参量与车辆噪声声音品质之间的预测模型,对排气噪声的满意度进行了预测,在相同的训练与测试样本集下,与多元线性回归模型预测结果进行了对比。结果表明,支持向量回归的预测值更接近实验值,平均绝对百分误差(3.12%)和均方根误差(0.65)都比多元线性回归(8.83%)和(1.99)的都小,相关系数高达0.99,预测精度更高,误差在8%范围以内,能更好反应客观参数与主观满意度间的非线性关系,是一种预测车辆噪声声音品质的有效方法。Through comparison test of38kinds of vehicle noise satisfaction evaluation,5parameters were checked out and chose as the objective psychoacoustic parameters of describing the sound quality of exhaust noise for vehicle.The model for predicting the relationship between the objective psychoacoustic parameters and the sound quality of vehicle exhaust noise was established through support vector regression(SVR),and the degree of satisfaction was predicted.Under the same training and test sample sets,this prediction results were compared with that obtained through multiple linear regression(MLR)prediction model.The results showed that the prediction values of SVR were close to the measured values,the mean absolute percentage error(3.12%)and root mean square(0.65)of SVR were smaller than those of MLR(8.83%and1.99),and the correlation coefficient was0.99with high prediction precision accuracy,which its error was less than8%,which suggested that SVR was an effective and powerful tool for predicting sound quality of vehicle exhaust noise.

关 键 词：公路运输 车辆 声音品质 排气噪声 支持向量回归 

分 类 号：TK411.6[动力工程及工程热物理—动力机械及工程]