测控技术2025，Vol.44Issue(10)：48-53,66,7.DOI:10.19708/j.ckjs.2025.09.251
一种基于风洞横向连续扫描系统的异常数据检测方法
Detection Method of Abnormal Data Based on Wind Tunnel Lateral Continuous Scanning System
黄心跃 1刘赟1
作者信息
- 1. 中国空气动力研究与发展中心低速空气动力研究所,四川绵阳 621000
- 折叠
摘要
Abstract
Incremental encoder data in industrial control system often presents the theoretical linear incremental characteristics,but the actual pulse signal in the continuous scanning mode of wind tunnel test is often affected by mechanical vibration,wind load disturbance and other factors.However,due to the large amount of data,long duration and complex field conditions in the acquisition process,it is difficult for testers to accurately judge the actual operation state of encoder,the accuracy of transmission signals and whether there are intermit-tent faults.Firstly,a compound anomaly detection method based on polynomial regression is proposed,which combines the difference constraint of adjacent points and the threshold analysis of nonlinearity to build an a-nomaly detection system for encoder pulse data.Then,the performance of the detection system under wind tun-nel test conditions is analyzed,and a double model library with wind load and without wind load is established to adjust the difference threshold flexibly,which can prevent false alarm caused by aerodynamic disturbance and compression phenomenon.Finally,regarding the nonlinear situation of the system under the influence of wind load,nonlinear terms and interactive terms are introduced,and quadratic and cubic polynomial regression modeling is adopted to effectively improve the nonlinear index and residual distribution.The proposed method matches the characteristics of the system itself and has the auxiliary monitoring capability of pulse data in the lateral continuous scanning mode.关键词
编码器脉冲数据异常检测/多项式回归/风洞试验/连续扫描Key words
encoder pulse data anomaly detection/polynomial regression/wind tunnel test/continuous scanning分类
信息技术与安全科学引用本文复制引用
黄心跃,刘赟..一种基于风洞横向连续扫描系统的异常数据检测方法[J].测控技术,2025,44(10):48-53,66,7.
