基于神经网络的柴油机声音品质评价预测  

Prediction and Evaluation of Sound Quality of Diesel Engines Based on Neural Network

作　　者：王芳[1] 丁新隆 石岩[2] 

机构地区：[1]天津交通职业学院,天津300110 [2]中国汽车技术研究中心,天津300162

出　　处：《噪声与振动控制》2014年第5期117-119,共3页Noise and Vibration Control

摘　　要：采用多元线性回归与人工神经网络系统(Artificial Neural Network,简称ANN)分别建立了柴油发动机噪声声音品质预测模型,并将两种模型的预测值与实测值进行比较。试验结果表明,该系统可以反映客观参数和主观满意度间的非线性关系,可以用来预测和描述柴油发动机噪声的声音品质。Multi- element linear regression and artificial neural network （ANN） are used respectively to the establishment of the models for prediction of sound quality of the noise of diesel engines. The prediction values of the two models are compared with the measurement values. Results of the test show that the system can reflect the nonlinear relationship between objective parameters and subjective satisfaction, and can be used to predict and describe the sound quality of the diesel engine noise.

关 键 词：声学 柴油发动机 声音品质 评价 BP神经网络 

分 类 号：O241.82[理学—计算数学] TK421.6[理学—数学]