Method for navigating vehicles, particularly land vehicles
First Claim
1. A method for the navigation of a vehicle, the vehicle including:
- course reference means for furnishing a course angle signal θ
M which represents the direction of the vehicle with reference to an earth-bound coordinate system;
longitudinal movement sensor means for detecting longitudinal movement of the vehicle and generating a longitudinal movement signal VM corresponding to the longitudinal movement of the vehicle;
position computer means for calculating vehicle position data, segragated into north and east position values, from signals generated by the course reference means and the longitudinal movement sensor means;
display means connected to the position computer for displaying vehicle position data calculated by the position computer; and
input means including at least one of manual input means and signal receiving means for providing navigation support data including at least one of additional position, course, velocity and path data;
said method comprising;
checking the longitudinal movement signal VM and the course angle signal θ
M for plausibility;
adding a known, empirically derived, deterministic velocity error component value DF(V) to the VM signal to produce, a corrected longitudinal movement signal CV;
optimally estimating, with the use of a Kalman filter, the stochastic position and direction errors contained in the VM and θ
M signals and using the estimate of such errors to calculate direction and change-in-direction correction values C(θ
) and C(ε
), respectively, and north and east position correction values C(RN) and C(RE), respectively;
adding the direction correction value C(θ
) and a known, empirically derived, deterministic course angle error component value DF(θ
) to the θ
M signal to produce a corrected course angle signal Cθ
;
feeding the position correction values C(RN) and C(RE) to the position computer means for use in correcting the position data;
forming corrected north and east component signals CVN and CVE, respectively, from the corrected longitudinal vehicle movement signal CV and from the corrected course angle signal Cθ and
feeding the CVN and CVE signals to the position computer means;
calculating, with the use of the position computer means, corrected north and east position coordinate values CRN and CRE, respectively, in dependence of the C(RN) and C(RE) correction values and the CVN and the CVE corrected north and east component signals;
obtaining north and east position bearing data RNS.sup.(jP) and RES.sup.(jP), respectively, from the input means;
comparing the corrected north and east position coordinate values CRN and CRE with the position bearing data RNS.sup.(jP) and RES.sup.(jP), respectively, to form north and east position bearing signals CZN.sup.(jP) and CZE.sup.(jP), respectively; and
feeding the CZNjP) and CZE.sup.(jP) signals to the Kalman filter, with the Kalman filter developing the following error model of the vehicle course angle error;
space="preserve" listing-type="equation">Δ
θ
(t)=Δ
θ
.sub.1 (t)+Δ
θ
.sub.2 (t)+Δ
θ
.sub.3 (t), wherein Δ
θ
1 (t) comprises a component of exponentially, time correlated, colored noise;
Δ
θ
2 (t) comprises a time linearly variable component representing drift angle with an unknown starting value Δ
θ
2 (0) and an unknown pitch ε
(t) representing a random ramp process; and
Δ
θ
3 (t) comprises a component of Gaussian white, time uncorrelated, noise; and
wherein the component Δ
θ
1 (t) is described by a form filter excited with white noise in a Gauss-Markov process of the first order; and
error which is contained in the position bearing data, RNS.sup.(jP) and RES.sup.(jP) is developed solely by stationary Gaussian white, time uncorrelated, noise.
1 Assignment
0 Petitions
Accused Products
Abstract
A navigation method for the navigation of a vehicle employing direction and longitudinal movement sensors. Signals from the sensors are fed to a position computer which determines the vehicle position relative to a coordinate grid system. When the vehicle reaches a position of known coordinates the position values which are determined from the sensor signals are corrected with reference to the known coordinates, and a Kalman filter is provided which effects an error estimate and subsequently a correction value determination from the error estimate which results in a significant increase in accuracy of the displayed navigation data. The particular advantage of the method is the use of uncomplicated sensors which are able, in connection with the Kalman filter, to obtain highly accurate vehicle position data.
37 Citations
3 Claims
-
1. A method for the navigation of a vehicle, the vehicle including:
- course reference means for furnishing a course angle signal θ
M which represents the direction of the vehicle with reference to an earth-bound coordinate system;
longitudinal movement sensor means for detecting longitudinal movement of the vehicle and generating a longitudinal movement signal VM corresponding to the longitudinal movement of the vehicle;
position computer means for calculating vehicle position data, segragated into north and east position values, from signals generated by the course reference means and the longitudinal movement sensor means;
display means connected to the position computer for displaying vehicle position data calculated by the position computer; and
input means including at least one of manual input means and signal receiving means for providing navigation support data including at least one of additional position, course, velocity and path data;
said method comprising;checking the longitudinal movement signal VM and the course angle signal θ
M for plausibility;adding a known, empirically derived, deterministic velocity error component value DF(V) to the VM signal to produce, a corrected longitudinal movement signal CV; optimally estimating, with the use of a Kalman filter, the stochastic position and direction errors contained in the VM and θ
M signals and using the estimate of such errors to calculate direction and change-in-direction correction values C(θ
) and C(ε
), respectively, and north and east position correction values C(RN) and C(RE), respectively;adding the direction correction value C(θ
) and a known, empirically derived, deterministic course angle error component value DF(θ
) to the θ
M signal to produce a corrected course angle signal Cθ
;feeding the position correction values C(RN) and C(RE) to the position computer means for use in correcting the position data; forming corrected north and east component signals CVN and CVE, respectively, from the corrected longitudinal vehicle movement signal CV and from the corrected course angle signal Cθ and
feeding the CVN and CVE signals to the position computer means;calculating, with the use of the position computer means, corrected north and east position coordinate values CRN and CRE, respectively, in dependence of the C(RN) and C(RE) correction values and the CVN and the CVE corrected north and east component signals; obtaining north and east position bearing data RNS.sup.(jP) and RES.sup.(jP), respectively, from the input means; comparing the corrected north and east position coordinate values CRN and CRE with the position bearing data RNS.sup.(jP) and RES.sup.(jP), respectively, to form north and east position bearing signals CZN.sup.(jP) and CZE.sup.(jP), respectively; and feeding the CZNjP) and CZE.sup.(jP) signals to the Kalman filter, with the Kalman filter developing the following error model of the vehicle course angle error;
space="preserve" listing-type="equation">Δ
θ
(t)=Δ
θ
.sub.1 (t)+Δ
θ
.sub.2 (t)+Δ
θ
.sub.3 (t),wherein Δ
θ
1 (t) comprises a component of exponentially, time correlated, colored noise;
Δ
θ
2 (t) comprises a time linearly variable component representing drift angle with an unknown starting value Δ
θ
2 (0) and an unknown pitch ε
(t) representing a random ramp process; and
Δ
θ
3 (t) comprises a component of Gaussian white, time uncorrelated, noise; and
wherein the component Δ
θ
1 (t) is described by a form filter excited with white noise in a Gauss-Markov process of the first order; and
error which is contained in the position bearing data, RNS.sup.(jP) and RES.sup.(jP) is developed solely by stationary Gaussian white, time uncorrelated, noise. - View Dependent Claims (2, 3)
- course reference means for furnishing a course angle signal θ
Specification