摘要
文章利用巡检机器人所带拾音器采集设备运行声音样本,并利用基于负熵的FastICA算法实现设备声音和环境噪声的分离,并对分离性能指标进行了统计和分析,实验结果表明该算法成功分离了声音样本中的各独立声音分量,分离性能稳定可靠,有助于提高后续状态识别的识别率。
The microphone array of the inspection robot is used to record the sound of the electrical devices when they are in operation. The negentropy Fast ICA algorithm is adopted to achieve the blind source separation,that is,to separate the device sound from the mixture of the multiple sound sources,then the separation parameter is calculated. The experiment results show that the algorithm in this paper successfully realized blind source separation with stable and reliable performance,and it's useful for the further device operation state recognition.
出处
《信息技术》
2018年第2期82-85,共4页
Information Technology
