基于模型的执行器故障诊断

为依据执行器的动态特性分析，实现执行器设备故障的在线诊断与分离，在机理分析的基础上建立起执行器完整的非线性动态数学模型，由执行器工作过程中的实测信号和模型计算获得残差实现故障的在线诊断，通过残差的变化及其组合情况分析完成故障分离.实验结果表明，利用动态机理模型计算得到的阀位、薄膜气室压力的理论计算值和实测值进行残差分析，可以准确及时地诊断与分离出如阀体阻塞、填料函磨擦力增大、薄膜或气路接头破损、气源压力下降、弹簧老化等执行器主要的内在故障.

Model-based actuator fault diagnosis

Online actuator fault diagnosis and separation were realized based on dynamic characteristics analysis and complete nonlinear dynamic models for actuators were constructed based on mechanical analysis.The online fault diagnosis was realized by using the fault residuals obtained from the measured actuator signals and the model calculation values,then the fault separation was performed through analyzing the varieties of the residuals and its combination.The experimental results show that through analyzing the residuals of valve stem displacement and pressure value main faults can be accurately and timely diagnosed and separated,such as valve blockage,friction enlarge,diaphragm leakage,supply pressure drop and spring aging.