A cubic spline curve based track map generation method for train control

With high accuracy and storage efficiency, digital track map plays an important role in train positioning system as a special sensor, and directly affects the safety of the train operation control. In this paper, the characteristics of the track are analyzed with practical measurements, a distance discrimination is utilized to reject abnormal data, and a vertical distance criterion and cubic spline curve based algorithm is presented to reduct numerous measurement data and fit the track. By using the distance criterion and vertical distance based local optimum algorithm, a group of feature points are generated to represent the feature of rail tracks. Then the well-fitting cubic spline algorithm is employed to fit the tracks with high precision and optimize the feature point set, and then the smooth tracks could be realized by segmental cubic curves. Validation with practical measurements from Qinghai-Tibet line demonstrate that with a proper limited error of the vertical distance the proposed algorithm could realize data reduction and map generation effectively.