论文部分内容阅读
针对传统发动机温度过热,检测技术无法克服存油量波动造成的温度变化信号非线性失真与故障检测效果不佳的问题,提出一种融入诊断网络算法的发动机温度过热信号检测技术,利用小波变换方法,采集动态发动机温度过热故障信号特征,以温度过热信号特征为依据,通过融合诊断网络在隐含层中对温度过热信号挖掘过程中的数据进行传输,得到过热温度变化信息,实现发动机温度过热信号的深度挖掘。实验结果表明,采用该技术能够提高发动机温度过热信号检测的准确度,有利于故障的快速修复。
Aiming at the problem that traditional engine overheating and detection technology can not overcome the non-linear distortion of temperature change signal and the fault detection effect caused by the fluctuation of oil reserves, a detection algorithm of engine overheating signal is put forward based on the diagnostic network algorithm. , Collecting the dynamic engine overheating fault signal characteristics, based on the characteristics of the overheating signal, through the fusion of diagnostic networks in the hidden layer temperature overheat signal mining process of data transmission, overheating temperature change information, the engine temperature overheating signal The depth of mining. Experimental results show that using this technology can improve the accuracy of engine overheating signal detection and facilitate the rapid repair of the fault.