摘要
对匀速工况下车内噪声信号分别进行主观评价与客观参量计算,并对主、客观评价结果进行了相关分析。在此基础上,基于Adaboost算法并结合BP神经网络、极限学习机(ELM)和支持向量机(SVM)建立了声品质预测模型,并将其预测结果与经过遗传算法(GA)参数优化后的GA-BP,GA-ELM和GA-SVM预测模型进行了对比。结果表明:基于Adaboost算法的车内噪声声品质预测模型效果最优,提升了声品质预测的准确度。
Both subjective evaluation and objective parameter calculation are conducted on the interior noise signals in constant speed driving,with a correlation analysis performed between the results of subjective and objective evaluations. On this basis,a sound quality prediction model is set up based on Adaboost algorithm and combined with BP neural network,extreme learning machine （ELM ） and support vector machine （ SVM ）,and the prediction results are compared with that of GA - BP, GA - ELM and GA - SVM prediction models,whose parameters have been optimized by genetic algorithm. The results show that the sound quality prediction model for interior noise based on Adaboost algorithm achieves the best effects, enhancing the accuracy of sound quality prediction.
出处
《汽车工程》
EI
CSCD
北大核心
2016年第9期1120-1125,共6页
Automotive Engineering
基金
国家自然科学基金(51475387)
四川省教育厅自然科学重点项目(16ZA0010)资助
