Real-time Bounded-error State Estimation for Vehicle Tracking

Abstract

Estimating the configuration of a vehicle is crucial for its navigation. Most approaches are based on (extended) Kalman filtering or particle filtering. An attractive alternative is considered here, which relies on interval analysis. Contrary to classical extended Kalman filtering it allows global localization, and contrary to particle filtering it provides guaranteed results in the sense that a set is computed that contains all of the configurations that are consistent with the data and hypotheses. This paper presents a real-time implementation of the process including a description of the platform and its modeling, the integration of the errors on the model and the localization method itself.

A Low-Cost Consistent Vehicle Localization Based on Interval Constraint Propagation

“…EKF use a probabilistic motion model to fuse sensor data [30]. We make an extension of this model to obtain an interval based model defined by Seignez et al [18]:…”

“…Followed-up researchers focused on real-time implementation of such localization technique. Kieffer [17] and Seignez [18] presented results for a mobile robot navigating in an indoor environment, with odometers mounted on each rear wheel to measure the motion and sonars located in different directions to perceive the environment. The ability of interval method to cope with erroneous data and to obtain accurate estimations of the robot pose was demonstrated.…”

“…We improve on it and present an extension of others' works dealing with the localization problem by using ICP. Instead of using ultrasonic sensors [16], sonar [18,25], or GPS [19,26] as exteroceptive sensors, we propose to use a monocular camera. The aim is to achieve both consistent and low imprecision localization result.…”

Real time simultaneous localization and mapping: towards low-cost multiprocessor embedded systems

“…x k−1|k−1 represents the state vector at time k-1, x k|k−1 represented the state vector after the prediction step, x k|k represents the state vector after the estimation step. The classical evolution model, described in [18], is considered:…”

A multihypothesis set approach for mobile robot localization using heterogeneous measurements provided by the Internet of Things

“…An alternative to PF could be the bounded-error state estimation. Experiments have demonstrated that the approach is operational on vehicle navigating in an indoor environment or in an outdoor environment [11]. In this approach, known as set-membership or set-theoretic estimation, all model and measurement errors are assumed to be bounded, with known bounds.…”