论文部分内容阅读
为了实现对某涡扇发动机传感器故障的在线诊断,提出并设计了1种基于在线贯序极端学习机的故障诊断算法。其核心思想是在定位某传感器故障后,在线建立针对该故障传感器“预学习”的信号重构算法,解决多故障混叠问题。在线信号重构算法以泛化能力指标为判定条件,利用选择策略对算法网络权值进行选择性更新,提高了故障诊断系统的实时性。以某型涡扇发动机为对象开展了传感器故障诊断与重构仿真,结果表明:该算法能够对发动机单、双传感器故障进行准确地诊断与信号重构,且具有良好的实时性。
In order to realize the on-line diagnosis of turbofan engine sensor fault, a fault diagnosis algorithm based on online sequential extreme learning machine is proposed and designed. The core idea is to establish a signal reconstruction algorithm for the fault sensor “pre-learning ” online after locating a sensor fault to solve the multi-fault alias problem. The online signal reconstruction algorithm takes the generalization ability index as the judgment condition, and uses the selection strategy to selectively update the algorithm network weights, which improves the real-time performance of the fault diagnosis system. The fault diagnosis and reconstruction simulation of a turbofan engine is carried out. The results show that the algorithm can accurately diagnose and reconstruct engine single and dual sensor faults and has good real - time performance.