基于Kalman滤波思想的接触线高度检测数据粗差修正算法

Gross Error Correction Algorithm of Contact Wire Height Detection Data Based on Kalman Filtering

接触线高度平顺性是评价接触网设备动态运行质量的重要指标。针对接触线高度检测数据中的粗差,基于Kalman滤波思想提出相应的粗差辨识与修正算法。将接触线高度视作一个连续平稳系统,利用矩形窗函数定义系统状态值,实现检测数据的分段线性化;依据系统状态值的回归预测方法和状态差异的量化方案,构建粗差辨识与修正算法,并引入适用于锚段关节处双支接触线采集特征的回溯机制;基于大量实测数据,采用显著性检验等统计方法确定算法参数,并从回溯、修正、运用和实时性等方面,对算法有效性和可行性进行分析验证,结果表明:该算法能够完成各类接触线高度检测粗差的高速实时辨识与修正。

The regularity of contact wire height is an important index to evaluate the dynamic operation quality of OCS equipment.Aiming at the gross error in contact wire height detection data,this paper proposes a corresponding identification and correction algorithm based on Kalman filtering.The contact wire height is regarded as a continuous and stable system,and the system status value is defined by rectangular window function to realize piecewise linearization of detection data.According to the regression prediction method of system status value and the quantification scheme of status difference,the gross error identification and correction algorithm is constructed,and the backtracking mechanism suitable for the acquisition characteristics of double-branch contact wires at overlap is introduced.Based on a large number of measured data,statistical methods such as significance test are used to determine the algorithm parameters,and the effectiveness and feasibility of the algorithm are analyzed and verified from backtracking,correction,application and real-time performance.The results show that this algorithm can complete high-speed real-time identification and correction of various types of gross errors in contact wire height detection.