论文部分内容阅读
基于特征参数趋势进化的故障预测是一种有效的方法,引入了一种考虑特征参数概率分布的新型判据进行多故障模式诊断与预测.基于过程神经网络建立了高精度预测模型,根据模型和部件使用记录进行趋势预测.基于方法对机载电子设备进行案例研究,结果表明,方法的判定结果更加符合多故障模式并存、故障严重程度不同的实际情况,而具有较高拟和、泛化预测精度的PNN模型是一种有效的趋势预测方法.
Fault prediction based on trend evolution of characteristic parameters is an effective method, which introduces a new criterion to consider the probability distribution of characteristic parameters to diagnose and predict multi-fault modes.A high-precision prediction model is established based on process neural network, and according to the model and Parts usage records are used to forecast the trend.According to the case study on airborne electronic equipment, the results show that the method’s decision results are more in line with the fact that the multi-fault modes coexist and the fault severity is different, but have higher fitting and generalized prediction Accurate PNN model is an effective trend prediction method.