Communications: directive radio wave systems and devices (e.g. – Directive – Including a satellite
Reexamination Certificate
2000-03-08
2001-07-31
Issing, Gregory C. (Department: 3662)
Communications: directive radio wave systems and devices (e.g.,
Directive
Including a satellite
Reexamination Certificate
active
06268823
ABSTRACT:
BACKGROUND
1. Field of the Invention
The present invention relates generally to the field of range navigation systems.
2. Description of the Background Art
In the available art related to range navigation systems, GPS and other navigation systems commonly measure the ranges to several transmitters by observing the times at which signals transmitted at earlier known times are received, wherein the distance is generally proportional to the measured time difference. The measurements are commonly related to an assumed position by subtracting from each range measurement a prediction of the range from that assumed position to the satellite or other navigation transmitter. The measurements are also intrinsically related to the time assumed in making the measurements, since any error in that time appears as an offset in the measured ranges. Such systems then commonly obtain a position fix by updating the assumed position and time on the basis of the discrepancies between the range measurements and predictions.
Although the relationship between the measurements and the physical coordinates of the vehicle is generally nonlinear, navigation systems commonly linearize the relationship about the assumed position and time, using the easily-computed first derivatives of the measurements with respect to the physical coordinates. This assumption of linearity allows the measurement-prediction discrepancies to be expressed easily as a function of the unknown error in the assumed position and time, specifically, as a matrix equation:
z=hx;
(1)
where z is the measurement-discrepancy vector, h is the matrix of partial derivatives of, commonly, three spatial and one temporal physical coordinates, and x is the update vector. The relationship (1) is then easily inverted by well-known matrix methods to express x as a function of z:
x=h
−1
z=nz;
(2)
where n is the navigation matrix. The resulting vector x in physical coordinates is then commonly applied as an update to the assumed position and time. Furthermore, there are well-known techniques to optimize that update when an excess of measurements over unknowns over-determines the solution, as by the least-squares or weighted-least-squares criteria.
Errors in the elements of z cause errors in the update x. If the z errors are assumed to be uncorrelated and equal in variance, as is generally true of GPS measurements, then the variance of the resulting position-update error is the product of the variance of each element of z with the sum of the squares of the elements of the rows of n that correspond to the three Cartesian position coordinates, by virtue of the statistical orthogonality of the measurements and the geometric orthogonality of the position coordinates. The square-root of that sum of squares is therefore the ratio of the standard deviation of position error to the standard deviation of measurement error and is commonly termed as the “position dilution of precision” or PDOP.
Although this treatment of an essentially nonlinear relationship as linear causes inevitable error in each update, that error is ordinarily negligible, because of the great disparity between the magnitudes of the update on one hand and the distances to the satellites on the other. Moreover, this same consideration makes it unnecessary in many applications to recompute the linearized relationship for each update, since it changes so little. Finally, if the linearized relationship causes update errors to become unacceptable though small compared to the update's size, iteration of the update process with the same set of measurements can reduce the need for frequent recomputation of the update matrix.
Once the measurement-prediction discrepancies z are available, the computation of the update x is a straightforward and relatively inexpensive matrix multiplication. However, the calculation of the predictions of the ranges from the assumed position to the satellites is more difficult, in that it generally requires a square-root operation for each satellite. This follows from the Pythagorean relationship that the square of the range is the sum of the squares of the differences in the three Cartesian coordinates commonly used to express the assumed and satellite positions. Thus, the predicted range to be subtracted from the measured range is the square-root of that sum.
What is needed is an unconventional range navigation system with efficient update process that avoids the frequent recalculation of square-roots in range predictions.
SUMMARY OF THE INVENTION
To address the shortcomings of the available art, the present invention provides a method and a system for determining a position of a platform using a set of range measurements between the platform and a set of known positions that avoids the frequent recalculation of square-roots in range predictions.
In one embodiment, the method comprises the following steps: (a) relating an estimate of the error in an assumed state of the platform to a set of measurement residuals; and (b) updating the platform state estimate by substituting the difference between the square of the range measurement and the square of the distance from the assumed position of the platform to the known position to which the range measurement corresponds, wherein each difference is divided by twice a distance between the predicted position of the platform and the known position, for each of the measurement residuals in the step (a). In another embodiment, the method of the present invention further includes the step (c) of repeating the step (b) until the platform state converges on a limit to a desired accuracy.
The assumed state includes an assumed position of the platform and an assumed bias which is the same in each element of the set of range measurements. Each element of the set of measurement residuals is the difference between an actual range measurement and a predicted range measurement. One actual measurement determines the distance between an actual position of the platform and one known position. The actual measurement includes an actual bias which is the same for each element of the set of range measurements. The predicted range measurement is a distance between the predicted position of the platform and one known position.
In one embodiment, the actual range measurement is performed by measuring the time for a signal to travel between one known position and the platform. The actual bias associated with this actual range measurement is due to a discrepancy between a transmission clock used to measure the time instance when the signal was transmitted and between a reception clock used to measure the time instance when the signal was received. In one embodiment, a radio signal generated using a transmitter located at one known position is used to perform an actual range measurement. In the preferred embodiment, the radio signal is generated using a GPS satellite. In an alternative embodiment, the radio signal is generated using a GLONASS satellite.
REFERENCES:
patent: 5957982 (1999-09-01), Hughes et al.
Issing Gregory C.
Tankhilevich Boris G.
Trimble Navigation LTD
LandOfFree
Unconventional range navigation system with efficient update... does not yet have a rating. At this time, there are no reviews or comments for this patent.
If you have personal experience with Unconventional range navigation system with efficient update..., we encourage you to share that experience with our LandOfFree.com community. Your opinion is very important and Unconventional range navigation system with efficient update... will most certainly appreciate the feedback.
Profile ID: LFUS-PAI-O-2468714