基于特征提取的飞机结冰严重程度识别

Aircraft In-flight Icing Severity Recognition Based on Feature Extraction

飞机结冰会影响飞机飞行的各项性能指标,严重威胁飞行安全,但是飞机的结冰信息难以全面获取,因此提出一种基于特征和支持向量机的飞机结冰严重状态分类的算法。获取多个结冰传感器的信息构成测量向量并提取其特征,采用主要飞行参数构造飞机结冰严重状态特征空间,并建立多个支持向量机进行飞机结冰状况分类。该方法建立测量向量与飞机结冰严重状态之间的联系,能对飞机空中结冰严重程度作出较好的识别。

Aircraft in-flight icing affects the properties of airplane and causes serious threats to flight safety.Aiming at the difficulty of the aircraft icing information acquisition,the icing severityrecognition algorithm model was built which mixed the icing feature extraction and method of icing condition classification based on SVM（support vector machine）.The feature vectors are extracted from the measurement space consisting of the key parameter by the sensors.A number of SVMs are built to tackle the icing severityrecognition problem.This method connects the measurement vectors and the icing condition vectors,achieving the aircraft icing severityrecognition.