基于信息融合的控制系统中模拟电路故障诊断

针对控制系统中的模拟电路具有多种工作状态的特殊情况和故障诊断过程中的不确定性问题，提出了模糊一神经网络信息融合算法．给出了该算法的基本模型和应用于控制系统中模拟电路故障诊断的一般规律，通过将电路不同工作状态下的有效信息进行融合，结合模糊信息处理技术和神经网络推理技术，最终实现故障定位．将其应用到某型船舶主机遥控系统中延时电路板的故障诊断系统中，结果表明，该方法能够有效地提高故障模式的识别能力，将故障准确地定位到元器件．

Fault diagnosis of analog circuits in control system based on information fusion

In order to effectively utilize information from various working status of control system and handle uncertainties in fault diagnosis of analog circuits,an information fusion method based on fuzzy theory and neural network is proposed.The basic model is given and the general rule to use this method for fault diagnosis of analog circuits in control system is described in detail.Combined with fuzzy information processing technology and neural network inference technology,fault component is located by fusing the effective information from various working status.The proposed method was applied in the fault diagnosis system of time delay circuit board used in the control system of marine main engine.The results show that this method can improve the ability to recognize fault patterns and accurately locate the fault component.