基于MHMM-SVM混合模型的雷达发射机性能退化状态监测

摘要

雷达发射机结构复杂,早期性能退化状态呈现多样化.为此,给出了一种MHMM-SVM故障诊断模型.该模型首先通过小波分析来提取观测信号的特征向量,然后从提取的特征向量中提取部分作为样本完成模型训练并生成发射机各种退化状态的分类器,最后给出诊断结果.仿真验证结果表明:MHMM-SVM故障诊断模型对电子装备早期故障的识别率明显优于MHMM、SVM单一模型.%As the structures of radar transmitter are complicated, there are various incipient faults. Thus, a new method for incipient fault diagnosis of radar transmitter was proposed based on MHMM-SVM. First, the wavelet transform technology was used to extract characteristic vectors from the observed signals. Then some of the characteristic vectors were used as samples to establish the MHMM-SVM model. The experimental results by comparing the MHMM-based and SVM-based diagnostician methods show that the method has more advantages in the diagnosis of incipient faults.