摘要
针对电液伺服作动器早期故障信号的微弱性和不易检测性,提出了一种基于故障特征概率分布相对熵的健康评估算法。采用离线训练对正常状态特征概率密度进行参数估计,在线监测作动器特征概率密度分布相对于正常状态的相对熵,通过判断相对熵的值对伺服作动器的进行健康评估。仿真结果证明了该方法的可行性和有效性。
Aiming at that the electro-hydraulic servo actuator of early fault signal is weak and not easy to de- tect, a new arithmetic based on fault feature probability distribution relative entropy is put forward for flight health evaluation. The parameter of normal state feature probability density is estimated by off-line training, and the actuator feature probability density relative entropy between current status and normal state is monitored on- line. The health status of servo actuator is evaluated by relative entropy. The simulation results prove the feasi- bility and validity of the method.
出处
《测控技术》
CSCD
北大核心
2012年第10期140-142,共3页
Measurement & Control Technology
关键词
概率密度
相对熵
伺服作动器
健康评估
probability density
relative entropy
servo actuator
health evaluation
