论文部分内容阅读
以某型液体火箭发动机为研究对象,根据其传感器的故障特性,提出了基于BP神经网络的传感器故障检测与数据恢复算法。通过定义的传感器置信度来判断传感器是否发生故障,以及确定故障传感器,利用已训练好的神经网络结构对故障传感器进行数据恢复。研究内容能够实现传感器的故障检测、定位与补偿,能够有效提高发动机故障检测方法的可靠性和鲁棒性。
Taking a liquid rocket engine as a research object, a sensor fault detection and data recovery algorithm based on BP neural network is proposed based on the fault characteristics of the sensor. Through the definition of the sensor confidence to determine whether the sensor fails, as well as to determine the fault sensor, the use of trained neural network structure of the fault sensor for data recovery. The research content can realize the sensor fault detection, positioning and compensation, which can effectively improve the reliability and robustness of the engine fault detection method.