Method for determining the kinematic state of an object, by evaluating sensor measured values

• US 7,928,898 B2
• Filed: 01/01/2008
• Issued: 04/19/2011
• Est. Priority Date: 02/14/2007
• Status: Active Grant

• Current Assignee: Airbus DS Electronics And Border Security GmbH
• Original Assignee: EADS Deutschland GmbH (Airbus SE)
• Inventors: Fraenken, Dietrich
• Primary Examiner(s): Tarcza; Thomas H
• Assistant Examiner(s): Brainard; Timothy A
• Application Number:
  US12/527,146
• Publication Number:
  US 20100026556A1
• Time in Patent Office: 1,204 Days
• Field of Search: 342106-107
• US Class Current: 342/107
• CPC Class Codes: G01S 13/726 Multiple target tracking

Abstract

In a method for determining the kinematic state of an object by evaluating a sequence of discrete polar measured values of a sensor, the polar measurements r_m, α_m are converted to Cartesian coordinates and subsequently scaled to Cartesian pseudo-measurements using a scaling factor β calculated as a function of measured range r_m. Associated pseudo-measurement error variance matrices are determined, each comprising nominal measurement error variances in the range direction R²_m and transversely thereto C²_m as a function of the measured range r_m. The state of the object is estimated, with an estimated variance {circumflex over (σ)}²_cross being determined transversely with respect to the range direction in an estimation device, based on the Cartesian pseudo-measurements and the pseudo-measurement error variance matrices. The scaling factor β is chosen such that no systematic error results for a position estimate comprising n>1 measurements; and the nominal pseudo-measurement error variance R²_m in the range direction is calculated as a function of the nominal pseudo-measurement error variance C²_m transversely thereto or conversely such that the variance {circumflex over (σ)}²_cross which is estimated after processing of n>1 measurements, transversely with respect to the range direction, on average matches the actual variance {circumflex over (σ)}²_cross of the estimated error transversely with respect to the range direction.

Claims

1. A method for determining a kinematic state of an object, including at least one of position, velocity and acceleration of the object, said method comprising:

a sensor making a sequence of discrete polar measurements with respect to said object, said polar measurements including as polar measurement variables, range and azimuth measurements from said sensor to said object;

converting the polar measurements to Cartesian coordinates by conversion of the polar measurement variables range and azimuth;

scaling the Cartesian coordinates to Cartesian pseudo-measurements using a scaling factor β

which is determined as a function of measured range;

determining associated nominal pseudo-measurement error variance matrices, each comprising specific nominal pseudo-measurement error variances (R²_m) in a range direction and transversely thereto (C²_m), as a function of the measured range; and

based on the Cartesian pseudo-measurements and the associated nominal pseudo-measurement error variance matrices, an estimation unit estimating the kinematic state of the object, with an estimated variance {circumflex over (σ

)}²_cross being determined transversely with respect to the range direction;

wherein,

the scaling factor β

is determined in the estimation device such that no systematic error results for a position estimate comprising n>

1 measurements; and

the nominal pseudo-measurement error variance R²_m in the range direction is determined in the estimation device as a function of the nominal pseudo-measurement error variance C²_m transversely to the range direction, or such that the variance {circumflex over (σ

)}²_cross which is estimated after the processing of n>

1 measurements, transversely with respect to the range direction, on average matches the actual variance {circumflex over (σ

)}²_cross of the estimated error transversely with respect to the range direction.

2. The method as claimed in claim 1, wherein:

a desired variance of the position estimate transversely with respect to the range direction is defined by presetting a reference value {circumflex over (σ

)}²_ref; and

the nominal variances are determined in the range direction (R²_m) and transversely thereto (C²_m) such that the estimation device on average determines position estimates with desired variances transversely with respect to the range direction, and produces matching estimates.

3. The method as claimed in claim 2, wherein a reference value {circumflex over (σ

)}²_ref is determined such that a bias-free method is achieved by the associated nominal variances in the range direction (R²_ref) and transversely thereto (C²_ref) with a consistent estimate of the estimated error variance transversely with respect to the range direction, with minimized nominal variance of the psuedo-measurements in the range direction.

4. The method as claimed in claim 3, wherein R²_ref is calculated:

in a first step, with the reference value {circumflex over (σ

)}²_ref being determined such that a bias-free method is specified by the associated nominal variances in the range direction (R²_ref) and transversely thereto (C²_ref), with a consistent estimate of the estimated error variance transversely with respect to the range direction, and with nominal variance of the psuedo-measurements in the range direction being as small as possible; and

in a second step, the reference variable R²_ref multiplied by a real factor q is used as the nominal variance R²_m in the range direction.

5. The method as claimed in claim 4, wherein the factor q is greater than unity.

6. The method as claimed in claim 5, wherein the factor q is limited by a value q_max.