无人飞行器振动数据特征识别仿真

Simulation of Vibration Data Feture Identification for Unmanned Aerial Vehicles

研究无人机实时的振动状态进行监控问题,以防止出现由于异常振动造成的故障。无人飞行器在飞行过程中很容易受到外界气流、风速异常的影响,引起机体无规则振动,异常振动信号无法形成规范的可识别特征。传统的无人飞行器异常振动数据识别需要在无规范识别特征的环境下,对异常噪声数据与常规数据进行分离,信号的不确定性导致分离过程变得复杂,不能很好识别无人飞行器的振动状况。为解决上述问题,提出一种新的无人飞行器振动数据特征准确提取方法,对振动信号特征进行提纯分析,在不确定获得的振动信号中是否有振动异常信号的情况下,对疑似振动异常信号进行约束,通过计算约束环境下,正常信号与振动异常信号的关联程度,对不同种类的振动异常信号数据特征进行关联差异聚类,利用深度挖掘方法进行振动特征提取。实验表明改进方法能够准确的识别无人飞行器的振动特征,改进效果明显。

The monitoring of the uav real-time vibration state is studied to prevent the faults caused by abnormal vibration. Unmanned aerial vehicles are affected by the outside wind and wind velocity anomaly in the process of flight airflow, and the bodyls random vibration is produced. So the abnormal vibration signal cannot be formed to recognize the characteristics of the specification. Abnormal vibration data to be identifified by the traditional unmanned aerial vehicles （uavs） need to separate the abnormal noise data and nidentify characteristics, due tothe uncertainty , identi- fication, new unmanned spacecraft vibration data features accurate extraction method, the vibration signal characteris- tics , the suspected abnormal vibration signal the normal signal and abnormal vibration signals the characteristics of different kinds of abnormal vibration signal data correlation differences clustering. Experiments show that the proposed method can accurately identify the vibration characteristics of unmanned aerial vehicles, has obviousment.