Real-Time Compression Scheme for Urban Rail Transit Train Operation Data

Abstract

［Objective］ In view of the current challenges such as increased requirements for disk storage space, frequent disk damage and high bandwidth occupation existing in urban rail transit train operation data, research on data compression is of great significance for reducing storage space requirements, alleviating disk burden, and cutting bandwidth waste. ［Method］ By analyzing the characteristics of train operation data, the internal redundant characteristics of the data are studied from three dimensions i.e. time correlation, spatial correlation and overall redundancy of the data itself. A multi-dimensional real-time compression scheme with inter-frame compression, improved RLE (Run-Length Encoding) compression and ZIP compression is proposed to reduce the data redundancy. The scheme is proved to be effective by the real-time verification test of algorithms and the verification test of the actual train operation data. ［Result & Conclusion］ The proposed multi-dimensional real-time data compression scheme features low complexity, strong compression ability and fast compression speed. Tests and applications show that it takes only 0.16 ms for the real-time data compression with the proposed scheme, meeting the 20.00 ms data acquisition cycle requirement. The final compression ratio with the proposed scheme can reach 3.35%, and hardware of the same cost can store data for a longer time. The final amount of data written to storage disks and wirelessly transmitted is reduced to 11.83% of the original one, greatly reducing the times of disk erasures and the amount of data wirelessly transmitted, and effectively lowering the disk storage failure rate and the occupancy of wireless transmission bandwidth.

Keywords

• urban rail transit
• train operation data
• real-time data compression