Multisensor Fusion for Railway Irregularity Inspection System: Integration of RTK GNSS, MEMS IMU, Odometer, and Laser

The accurate assessment of railway irregularities plays a pivotal role in ensuring both operational safety and passenger comfort, especially in the context of high-speed railways. Traditional static methods for monitoring these irregularities, such as absolute measurement trolley systems relying on total stations, have proven to be inefficient. On the other hand, relative measurement systems like the global navigation satellite system/inertial navigation system (GNSS/INS) integrated concept offer high operational efficiency but come with substantial costs associated with navigation-grade inertial measurement unit (IMU) systems. In this research endeavor, we introduce a novel multisensor fusion system centered around a relative inspection trolley. This innovative approach serves to significantly reduce equipment expenditures while simultaneously elevating the accuracy of irregularity detection. The core components of this system comprise real-time kinematic (RTK) GNSS, a cost-effective micro-electric mechanical system (MEMS) IMU, an odometer, and two laser modules. The integration of various advanced techniques, including zero velocity update (ZUPT), Ranch-Tung-Striebel smoother (RTS), nonholonomic constraint (NHC), odometer calibration (OC), and prior knowledge (PK), collectively enhances the efficiency and precision of railway irregularity measurements. The results obtained from extensive testing on railway irregularities demonstrate the remarkable capabilities of the proposed system. It exhibits a measurement precision of approximately 0.2, 1.0, 0.7, 0.5, and 0.5 mm for rail gauge, cross level, twist, vertical irregularities, and lateral irregularities, respectively. Consequently, our suggested technique stands as a compelling choice for the precise quantification of railway irregularities, offering a compelling solution for the railway industry.