PLR Model Based Forecast of Track Irregularity for Tamping Operations

摘要

With the rapid development of construction of railway in China, forecast of track irregularity becomes more significance for more efficient maintenance works. The piecewise linear prediction model is established under the conditions of known tamping operation efficiency and initial quality of railway track. The changing trend of track irregularity between two tamping operations. So in this paper, a piecewise linear representation based forecasting method is proposed to predict the track irregularity by mining the Track Quality Index (TQI) with underlying "memory" of track changing information of its special characteristic. Experiment results demonstrate that the accuracy of the proposed prediction model is over 94%, which can be used as important reference for arranging maintenance works.