Adaptive Kalman filter approach for road geometry estimation

Abstract

This paper describes an adaptive Kalman filter based method for accurate estimation of forward path geometry of an automobile. The forward geometry is modeled as two contiguous clothoid segments with different geometries and continuous curvature across the transition between them. This results in a closed-form parametric expression of the same polynomial order as previous models. Instead of using a conventional Kalman filter with fixed process model parameters based on a compromise between noise and filter lag, we adaptively tune the process model parameters. This results in the better filter performance with stable estimates during constant geometry scenarios and faster response during abrupt geometry transitions. Performance evaluation of the proposed method on various simulated road geometries and comparing with previous approaches demonstrate the feasibility and higher accuracy of the proposed method. The high accuracy estimation of forward path or road geometry is directly useful in applications that rely on detecting targets in the forward path of the host vehicle, e.g., adaptive cruise control and automotive collision warning.