A robot navigating in an unstructured outdoor environment must determine its own location in spite of problems due to environmental conditions, sensor limitations and map inaccuracies, exact measurements are seldom known, and the combination of approximate measures can lead to large errors in self-localization. The conventional approach to this problem has been to deal with the errors either during processing or after they occur. The authors maintain that it is possible to limit the errors before they occur. The authors analyze how measurement errors affect errors in localization and propose that a simple algorithm can be used to exploit the geometric properties of landmarks in the environment in order to decrease errors in localization. The authors' goal is to choose landmarks that will provide the best localization regardless of measurement error, determine the best areas in which to identify new landmarks to be used for further localization and choose paths that will provide the least chance of "straying". The authors show the result of implementing this concept in experiments run in simulation with USGS 30 m DEM data for a robot statically locating, following a path and identifying new landmarks. >

/pdf/localizing-in-unstructured-environments-dealing-with-the-2l90z1oaik.pdf

Localizing in unstructured environments: dealing with the errors

The assured-integrity monitored-extrapolation (AIME) navigation apparatus selectively utilizes measurements provided by ancillary sources at periodic intervals in determining the state of the platform on which the apparatus is mounted. The measurements have attributes which are measures of quality, quality being a measure of the usefulness of the measurement in accurately estimating the state of a platform. The AIME apparatus makes its selection of measurements for state determination on the basis of estimates of the values of these quality attributes. The determination of the quality of a time sequence of measured values of a particular quantity requires an evaluation time for its accomplishment. The AIME apparatus therefore determines the platform's state in two phases. It obtains highly-accurate determinations of the states of the platform at times prior to present time minus the evaluation time by using the quality measures available at these times and using only those measurements that are determined to be of high quality in the determination of state at these times. The platform state at present time is then obtained by extrapolation of the highly-accurate state at time minus the evaluation time using measurements whose quality is more uncertain.

Assured-integrity monitored-extrapolation navigation apparatus

To achieve improved performance over a cascaded Kalman Global Positioning System (GPS)/Inertial Navigational System (INS) when biases are present, a fuzzy state estimator can be used. A Kalman filter assumes there is no bias in the measurement source, e.g., the GPS receiver. However, standard positioning system (SPS) GPS has selective availability (SA) and other satellite errors that translate into a bias associated with each satellite. This causes the GPS receiver position output to have a bias that is dependent upon which satellites are being used to determine the position. As the vehicle maneuvers, the GPS receiver uses different satellites, causing the bias to change. Another changing bias in a cascaded filter is the GPS receiver time lag. This time lag creates a bias that is dependent upon the vehicle's velocity and changes as the vehicle maneuvers. A standard Kalman filter will introduce errors into the filter's state velocities and positions as it tries to follow the GPS data with these changing biases. By using a fuzzy state estimator, which can incorporate heuristics into the filter that compensate for the changing biases, the errors in the state position and velocity due to these changing biases can be significantly reduced, providing a better performing state estimator for a SPS GPS/IRU system.

A fuzzy navigational state estimator for GPS/INS integration

Geometric inference is widely used in computer vision, but very little attention has been given to the question of how geometric properties affect the resulting errors in the inferences made. This thesis addresses the problem of the stability of geometric inference in determining locations with a goal of being able to predict type and magnitude of the errors which occur and to determine on what basis to make geometric inferences which will minimize error. It is shown that the amount of the error occurring in a localization process using angular measurements to features depends heavily on which features are used, that the amount of the error occurring in such a localization process is not a function of the number of features used, that it is possible to develop simple heuristic functions for choosing features for localization which will significantly decrease error in that localization, that it is possible to decrease localization error in a particular direction, and that, if features have been identified but knowledge of left to right order in the view is unknown, simple steps can be taken to aid in determining that ordering. This knowledge is applied in the domain of robot navigation in outdoor, unstructured environments.

/pdf/the-stability-of-geometric-inference-in-location-4oisbw706a.pdf

The stability of geometric inference in location determination

An optimized Linear-Ellipsoidal-Bounded (LEB) filter has been developed and applied to data obtained from a ground test using a combined INSIGPS configuration. In this cascaded configuration, the filter receives eight outputs from the INS (accelerations, velocity, angles, altitude) and six outputs from the GPS (velocities and positions). The GPS measurements have included the effect of SA -of varying or unknown spectrum which, although likely to be estimated and compensated with some modelling techniques -at the expense of including extra state variables-, could also be dealt with the approach indicated in this article a t much less effort. An optimized formulation for the LEB filter is presented, in which the volume of the ellipsoid containing the estimation errors is minimized at every step or a t selected intervals. The SA effect is modelled as an unknownbut-bounded (UBB) noise process. Comparisons with an Extended Kalman filter (KF) show that KF innovations are not white and the LEB filter estimates are one order of magnitude smaller that those produced by the KF.

Development of an optimized LEB filter and its application to INS/GPS test data

The authors describe research to compensate for selected availability (SA) corruption by modifying the Kalman filter used in a typical Global Positioning System/inertial navigation system (GPS/INS) integration. Two modified filters are developed: one filter that models SA using individual error-states and another filter that compensates for SA by increasing the measurement noise. Neither of these filters totally eliminates the corruption due to SA, but both provide a significant increase in the accuracy of the navigation solution. >

Compensation of selected availability using a GPS/INS extended Kalman filter

To quantify the performance of navigation systems, the US Air Force has used a baseline navigation system, the Completely Integrated Reference Instrumentation System (CIRIS), which obtains an accurate navigation solution by combining information from three major subsystems. Although the navigation solution produced by CIRIS is accurate, it will soon be inadequate as the standard against which future navigation systems can be tested. The authors propose an alternative to CIRIS, a hybrid Enhanced Navigation Reference System (ENRS) which takes advantage of a newer ring laser gyro strapdown inertial navigation system (INS), certain features of the current CIRIS, and certain features of differential Global Positioning System (DGPS) measurements. Analysis conducted using a Kalman filter development package known as the Multimode Simulation for Optimal Filter Evaluation (MSOFE) is presented. Both a large order baseline truth model for the ENRS and full- and reduced-order Kalman filters are developed. >

/pdf/augmentation-of-a-navigation-reference-system-with-1c3xqk4nhm.pdf

R.N. Paschall

Papers

Compensation of selected availability using a GPS/INS extended Kalman filter

Augmentation of a navigation reference system with differential global positioning system pseudorange measurements