基于时间序列的音频采集通道故障检测研究

Research of Audio Data Collection Channel Fault Detection Based on Time Series

音频数据采集系统是航空电子系统的重要组成部分,其采集通道故障与否关系到舱音采集功能能否有效工作。针对传统对音频采集通道的故障检测方法存在检测效率不高、开发和维护成本较高等问题。提出了一种音频采集通道的故障自动检测方法,基于音频数据的弧度时间序列,利用参考音频信号与音频采集通道输出的数字音频信号的相似度对比结果,判断音频采集电路能否保持音频的一致性,进而判断音频采集通道是否存在故障。方法不仅检测成本低、精度高,并且实现了音频采集通道的快速、自动化的故障检测,同时针对不同领域的音频采集系统具有一定通用性。

Audio data collection system plays an important role in avionic systems.The fault detection of audio data collection channel makes the audio data collection system work well.Traditional fault detection methods for audio data collection channel mainly consists of the manual and the electrical detection methods,which have problems of inefficiency detection,higher developing and maintaining costs.This paper introduces a new method based on radian time series for fault detection of audio data collection channel.This method uses the result of similarity matching by the default input audio data and the output of detection needed channel to judge whether it can make the audio data match well,thereby to conclude the fault detection result.The introduced fault detection method not only has lower costs but also with higher accuracy.Also,the designed way can make fault detection of audio data channel fast and automatic.Meanwhile it can be used for other audio data collection systems in different domains.