大数据实时流计算的高铁转向架数据过滤算法研究

Data Filtering Algorithm for High Speed Bogie Based on Real-time Data Stream Computation

为解决高铁转向架数据过滤在大数据流计算中受多工况影响的计算效率低下和精准度不高的问题。在高铁大数据实时流计算中使用多判据因子方差斜率算法进行特征提取多工况数据,并结合交路线上相应GPS坐标点上的权重参考值进行数据过滤。通过高铁实际项目运行验证:该方法能有效降低数据干扰,提升数据过滤准确率到95%以上,实现准确监控和预测高铁转向架故障,大幅降低了高铁转向架的检修工作量,提高了检修效率;同时能满足实时流计算每秒上百万的计算效率。

In order to solve the problem of low efficiency and low accuracy of data filtering for high-speed railway bogie,which is affected by many working conditions in large data stream calculation.Multi-criterion factor variance slope algorithm is used to extract multi-condition data in the real-time stream calculation of high-speed railway large data,and the data is filtered by combining the weight reference values of the corresponding GPS coordinate points on the routing.Through the actual operation of high-speed rail projects,the method can effectively reduce data interference,improve the accuracy of data filtering to more than 95%,achieve accurate monitoring and prediction of high-speed rail bogie faults,greatly reduce the maintenance workload of high-speed rail bogies,and improve the maintenance efficiency.At the same time,it can satisfy the computation efficiency of real-time computation of millions of streams per second.