Apparatus and method of calculating position-velocity-time results of receiver

Patent metadata

Field	Value
Publication number	US-11662476-B2
Application number	US-202117379958-A
Country	US
Kind code	B2
Filing date	Jul 19, 2021
Priority date	Jun 30, 2021
Publication date	May 30, 2023
Grant date	May 30, 2023

Abstract

A PVT calculation device includes a memory; and one or more processors in communication with the memory configured to perform operations including: receiving observations and ephemerides from satellites to obtain PVT data of the satellites and predicted PVT results of the receiver; setting up observation functions respectively corresponding to the satellites; calculating by a least square solution first estimated PVT results of the receiver based on the observation functions; iteratively eliminating by a Random-Sampling Iterative Kalman Filter (RSIKF) algorithm fault observation functions from the observation functions in an inner cluster until no fault observation functions detected in the inner cluster; calculating by the RSIKF algorithm a second estimated PVT results of the receiver using the observation functions in the inner cluster; and outputting final estimated PVT results of the receiver. The PVT calculation device may calculate the PVT results of the receiver with improved accuracy and stability.

First claim

What is claimed is:

1. A Position-Velocity-Time (PVT) calculation device ( 200 ) in a Global Navigation Satellite System (GNSS) receiver ( 100 ), comprising:
a memory; and
one or more processors in communication with the memory, the one or more processors configured to perform operations including:
receiving observations and ephemerides from satellites to obtain PVT data of the satellites and predicted PVT results of the receiver;
setting up observation functions respectively corresponding to the satellites, based on the PVT data of the satellites and the predicted PVT results of the receiver;
calculating by a least square solution first estimated PVT results of the receiver based on the observation functions;
iteratively eliminating, by a Random-Sampling Iterative Kalman Filter (RSIKF) algorithm, fault observation functions from the observation functions in an inner cluster until no fault observation functions detected in the inner cluster;
calculating, by the RSIKF algorithm, second estimated PVT results of the receiver using the observation functions in the inner cluster; and
outputting final estimated PVT results of the receiver.

2. The PVT calculation device of claim 1 , wherein the iteratively eliminating fault functions from the observation functions comprises:
in a first iteration ( 31 ), randomly sampling ( 32 ) a first number of observation functions into the inner cluster, and putting a remaining second number of observation functions in an outlier cluster;
in the first iteration, randomly ranking ( 34 ) inner observation functions in the inner cluster in a sequential order;
in the first iteration, sequentially chi-square detecting ( 35 ) the inner observation functions in the inner cluster in the sequential order, wherein upon detecting one or more fault inner observation functions in the inner cluster, the one or more fault inner observation functions are moved from the inner cluster to the outlier cluster; and
in the first iteration, chi-square detecting the second number of observation functions in the outlier cluster ( 36 ), wherein upon detecting at least a quality outlier observation function in the outlier cluster, the at least quality outlier observation function are moved from the outlier cluster to the inner cluster.

3. The PVT calculation device of claim 2 , wherein the calculating the second estimated PVT results of the receiver using the remaining observation functions comprises:
in the first iteration, responsive to detecting a first inner observation function passing a RSIKF chi-square detection, updating by the first inner observation function the predicted PVT results of the receiver to form an updated PVT results of the receiver,
in the first iteration, responsive to detecting a subsequent inner observation function passing the RSIKF chi-square detection, updating by the subsequent inner observation function the updated PVT results of the receiver, and
in the first iteration, responsive to detecting a last inner observation function passing the RSIKF chi-square detection, updating by the last inner observation function the updated PVT results of the receiver.

4. The PVT calculation device of claim 3 , wherein upon at least one function of the observation functions being found moving between the inner and outlier clusters in a current iteration in the first iteration, a RSIKF convergence detection is set false, and
wherein upon no functions being found moving between the inner and outlier clusters in the first iteration, the RSIKF convergence detection is set true.

5. The PVT calculation device of claim 4 , further comprising:
in a subsequent iteration, if the RSIKF convergence detection is set false after the first iteration, adjusting ( 33 ) the inner cluster and the outlier cluster;
in the subsequent iteration, randomly ranking ( 34 ) by the RSIKF algorithm the inner observation functions in the adjusted inner cluster in a sequential order;
in the subsequent iteration, sequentially chi-square detecting ( 35 ) by the RSIKF algorithm the inner observation functions in the adjusted inner cluster in the sequential order,
wherein a further first inner observation functions passing the chi-square detecting by the RSIKF algorithm updates the predicted PVT results of the receiver to form a further updated PVT results of the receiver,
wherein a further subsequent inner observation function passing the chi-square detecting by the RSIKF algorithm updates the further updated PVT results of the receiver, and
wherein a further last inner observation function passing the chi-square detecting by the RSIKF algorithm updates the further updated PVT results of the receiver;
in the subsequent iteration, upon detecting one or more fault inner observation functions in the adjusted inner cluster, moving the one or more fault inner observation functions from the adjusted inner cluster to the adjusted outlier cluster;
in the subsequent iteration, chi-square detecting ( 36 ) the outlier observation functions in the adjusted outlier cluster, wherein upon detecting at least a quality outlier observation function in the adjusted outlier cluster, the at least quality outlier observation function is moved from the adjusted outlier cluster to the adjusted inner cluster; and
in the subsequent iteration, determining whether to continue to another subsequent iteration based on whether at least an observation function is found moving between the adjusted inner and outlier clusters, wherein iterations continues until no observation functions are found moving between the adjusted inner and outlier clusters.

6. The PVT calculation device of claim 5 , wherein upon finding no observation functions moving between the adjusted inner and outlier clusters in the subsequent iteration, the updated PVT results of the receiver are output as the final estimated PVT results of the receiver to outside, and wherein upon finding at least one observation functions moving between the adjusted inner and outlier clusters in the subsequent iteration, a further subsequent iteration carries out.

7. The PVT calculation device of claim 6 , wherein upon finding at least one observation function moving between the adjusted inner and outlier clusters after a limited number of iterations, the first estimated PVT results of the receiver are output as the final estimated PVT results of the receiver to outside.

8. A computer-implemented method ( 400 ) of calculating Position-Velocity-Time (PVT) results of a receiver ( 100 ) using a PVT calculation device ( 200 ), the method comprising:
receiving observations and ephemerides from satellites to obtain PVT data of the satellites and predicted PVT results of the receiver;
setting up observation functions respectively corresponding to the satellites, based on the PVT data of the satellites and the predicted PVT results of the receiver;
calculating, by a least square solution, first estimated PVT results of the receiver based on the observation functions;
iteratively eliminating, by a Random-Sampling Iterative Kalman Filter (RSIKF) algorithm, fault observation functions from the observation functions in an inner cluster until no fault observation functions detected in the inner cluster;
calculating, by the RSIKF algorithm, second estimated PVT results of the receiver using the observation functions in the inner cluster; and
outputting final estimated PVT results of the receiver.

9. The computer-implemented method of claim 8 , wherein the iteratively eliminating fault functions by the RSIKF algorithm from the observation functions comprises:
in a first iteration ( 31 ), randomly sampling ( 32 ) a first number of observation functions into the inner cluster, and putting a remaining second number o

Assignees

• Beken Corp

Inventors

• Guo Dawei
• Zhang Pengfei

Classifications

• G01S19/396

  Determining accuracy or reliability of position or pseudorange measurements · CPC title

• G01S19/42 Primary

  Determining position · CPC title

• G01S19/22 Primary

  Multipath-related issues · CPC title

• G01S19/40

  Correcting position, velocity or attitude · CPC title

• G01S19/52

  Determining velocity · CPC title