汽车冷却风扇噪声主动控制仿真及声品质评价
Active Noise Control Simulation and Sound Quality Evaluation of Automotive Cooling Fans
胡溧 1胡远生 1谭征宇 2王华伟 1王博 1王佳2
作者信息
- 1. 武汉科技大学 汽车与交通工程学院,武汉 430065
- 2. 东风商用车有限公司技术中心,武汉 430000
- 折叠
摘要
以汽车冷却风扇噪声为研究对象,基于采集的噪声生成具有不同音调特性的模拟噪声,利用改进的音调度模型计算音调度,使用等级评分法对生成的模拟噪声进行主观评价实验.根据音调特性十分明显的旋转噪声频率与风扇转速密切相关的特点,提出以风扇转速信号来构造次级声源参考信号进行主动噪声控制.最后建立并联型灰色神经网络模型并检验降噪效果.结果表明,冷却风扇噪声的音调特性是影响感知烦恼的重要因素之一,通过有源噪声控制能有效降低窄带噪声峰值从而降低音调度.引入并联型灰色神经网络进行预测的平均相对误差为2.86%,降噪后烦恼度有较好的改善.
Abstract
The noise of automobile cooling fans is studied.Based on the collected noise sample of the cooling fan,the modeled noise with different tone characteristics is generated.The improved tonality model is used to calculate the tone scheduling,and the subjective evaluation experiments is carried out on the generated modeled noise using the rank scoring method.According to the characteristic that the rotational noise frequency is closely related to the fan speed,the fan speed signal is used to construct the secondary sound source reference signal for active noise control.Finally,the parallel grey neural network model is established and the noise reduction effect is tested.The results show that the tone characteristics of the cooling fan noise is one of the important factors affecting the perceived annoyance,and active noise control can effectively reduce the peak of narrowband noise and thus reduce the tone scheduling.The average relative error of prediction introducing the parallel grey neural network model is only 2.86%,and the annoyance degree after the denoising is improved.
关键词
声学/汽车冷却风扇/改进的音调度模型/等级评分法/主动噪声控制/并联型灰色神经网络模型Key words
acoustics/automobile cooling fan/improved tonality model/grade scoring method/active noise control/parallel grey neural network model引用本文复制引用
基金项目
国家自然科学基金(51105283)
国家自然科学基金(51905389)
出版年
2024
国家科技期刊平台
NSTL
万方数据