A localization algorithm for railway vehicles

Abstract

Why l&M? Measurement is a part of almost all engineering, for "to measure is to know," and we must understand the universe in order to engineer solutions in it. A professional development program based on improving your knowledge of measurement and your use of instrumentation is a key part of engineering from biomedical to aerospace. Power electronics needs it and so does high speed computing. The Institute for Electrical and Electronic Engineers is the largest professional organization in the world. Networking is recognized as a secret of success so why not start at the top and network with the most diverse group of professionals in your field. And if Instrumentation and Measurement isn't your first Society of interest within IEEE, it should be your second. Join us today and take advantage of resources and connections that the l&M Society can provide to boost your expertise in any field of electrical engineering.

Experimental Comparison of Visual-Aided Odometry Methods for Rail Vehicles

“…To decrease the dependency on track-side infrastructure, new sensor modality research is highly focused on the use of an inertial measurement unit (IMU) together with a tachometer [18] or GNSS. Otegui et al [19] summarizes many works fusing IMU and GPS signals employing an extended Kalman filter (EKF) or a particle filter (PF).…”

Robust Odometry using Sensor Consensus Analysis

“…Slipping wheels were detected by comparing the acceleration of the wheel with the measured acceleration from the IMU, with the affected measurements subsequently discarded. The performance of this approach was verified in [10] using a 3D multi-body simulation and hardware-in-the-loop test rig that was capable of reproducing the motion of a railway vehicle. While the above approaches produce good results, they lack any measure quantifying the confidence or uncertainty of the odometry estimate due to their heuristic nature.…”

Train distance and speed estimation using multi sensor data fusion

“…This error is reduced by the real‐time calibration of the wheel diameter using the distance information from balises [7] or radio frequency identification [3, 8]. Here, the wheel sensor is not the stand‐alone solution, the combination of alternate sensors such as radar [2, 3, 7, 9, 10], gyroscope [11–14], and accelerometer [10, 12] is used. Furthermore, several studies have reported on the autonomous train navigation solutions based on the global navigation satellite system (GNSS) receiver along with a track map [15–18], the accelerometers, the gyroscopes, the magnetometer, the wheel sensor [16, 19, 20] and the camera [21, 22].…”

“…Several multi‐sensor data fusion architectures were established based on the federated Kalman filter [2, 3, 6, 8, 10] and distributed Kalman filter [13, 14, 16] with the base as the extended Kalman filter (EKF) [15, 17, 20] for train state estimation over the years. In the case of the federated Kalman filter architecture [2, 3, 6, 8, 10], the master filter combines the estimates from the local filters.…”

“…Often, the designing and tuning of the central filter become difficult due to the sensor data measured at a different rate and non‐uniform noise filtering capability of local filters. In [13, 14, 16], the distributed Kalman filter was designed to use the complete link with all‐to‐all identical measurement noise characteristic sensors, which will increase the computational load when the number of sensors increases. These fusion algorithms are computationally expensive to implement in most real‐time applications that outweigh their advantages.…”