摘要
由于传统的发电机故障检测技术多数使用PC机进行实现,使得发电机故障检测系统缺乏较好便携性、高效性与实时性,因此,本课题利用Android为嵌入式开发平台,设计完成了一个以BP神经网络为故障判断模型的发电机声音检测与故障诊断系统。阐述了设计本系统的总体框架,并围绕框架中所包含的三大功能模块,即发电机声音信号采集模块、声音信号预处理模块和故障识别模块中所包含的关键算法进行了研究与设计,实现了对发电机工作状态下的声音信号去噪处理、特征值提取及故障识别与显示输出。实验分析结果表明,该系统运行简便可靠,具有一定的半实时性、适用性和工程实用价值。
Currently most of the generator fault detection systems rely on PC,which is lacking in portability,high efficiency and real-time,the subject takes Android as a platform,designs a generator voice signal fault detection system based on BP network fault diagnosis system.This paper introduces the overall framework of fault detection system.And around three functional modules,which are a sound signal acquisition module,a sound signal preprocessing module and a fault identification module,the key algorithm was studied and designed,sound signal De-noising,fault feature extraction and display output of the generator sound signals are achieved.The experimental analysis results show that the system operation is simple and reliable,has half real-time,certain applicability and practical value.
出处
《内蒙古工业大学学报（自然科学版）》
2015年第3期201-208,共8页
Journal of Inner Mongolia University of Technology：Natural Science Edition
