Shared state selection and data exchange for collaborative navigation using conditionally independent parallel filters

• US 8,442,765 B1
• Filed: 11/08/2012
• Issued: 05/14/2013
• Est. Priority Date: 11/08/2012
• Status: Expired due to Fees

• Current Assignee: Honeywell International Inc.
• Original Assignee: Honeywell International Inc.
• Inventors: Ingvalson, Ryan
• Primary Examiner(s): Marc, McDieunel
• Application Number:
  US13/672,337
• Time in Patent Office: 187 Days
• Field of Search: 701/36, 701/494, 701/519, 701/532, 701/505, 701/38, 701/48, 701/54, 701/408, 701/470, 701/472, 702/122, 702/188, 702/189, 429/143, 429/419, 429/147, 342/357.28, 342/372, 342/146, 342/357.48, 342/458, 342/463, 706/46, 706/52, 709/202, 709/203, 709/217, 709/219, 709/223, 707/999.104, 707/999.107, 340/995.12, 340/995.19, 340/995.24, 705/53, 705/57, 705/80, 382/103, 382/278, 715/851
• US Class Current: 701/505
• CPC Class Codes: G01S 13/08 Systems for measuring distance only indirect measurement G01S13/46
• CPC Class Codes: G01S 19/47 the supplementary measurement being an inertial measurement, e.g. tightly coupled inertial
• CPC Class Codes: G01S 19/51 Relative positioning
• CPC Class Codes: G01S 5/0072 Transmission between mobile stations, e.g. anti-collision systems
• CPC Class Codes: G06F 16/9535 Search customisation based on user profiles and personalisation
• CPC Class Codes: G06F 16/954 Navigation, e.g. using categorised browsing

Abstract

A method for collaborative navigation comprises initializing a conditionally independent filter on a local platform, propagating the conditionally independent filter forward in time, and when a local measurement has been made, updating the conditionally independent filter with the local measurement. When a common measurement has been made, the conditionally independent filter is updated with the common measurement, and a determination is made whether a remote conditioning node has arrived from a remote platform. If a remote conditioning node has arrived, a determination is made whether the remote conditioning node needs to be fused with a local conditioning node of the conditionally independent filter. If the remote conditioning node needs to be fused, a node-to-node fusion is performed in the conditionally independent filter to merge the remote conditioning node with the local conditioning node to produce a merged conditioning node. A conditional update of the conditionally independent filter is then performed.

Claims

1. A method for collaborative navigation, comprising:

(a) initializing a conditionally independent filter on a local platform;

(b) propagating the conditionally independent filter forward in time;

(c) when a local measurement has been made, updating the conditionally independent filter with the local measurement;

(d) when a common measurement has been made, updating the conditionally independent filter with the common measurement;

(e) determining by a processor whether a remote conditioning node has arrived from a remote platform for conditional update processing;

(f) if a remote conditioning node has arrived, determining by the processor whether the remote conditioning node needs to be fused with a first local conditioning node of the conditionally independent filter;

(g) if the remote conditioning node needs to be fused with the first local conditioning node, performing a node-to-node fusion in the conditionally independent filter to merge the remote conditioning node with the first local conditioning node to produce a merged conditioning node; and

(h) performing a conditional update of the conditionally independent filter.

2. The method of claim 1, further comprising:

after performing the conditional update, determining whether data from the remote platform contains new states to be augmented into the first local conditioning node or the merged conditioning node;

if data from the remote platform contains new states to be augmented;

updating the new states with the first local conditioning node or the merged conditioning node via a conditional update; and

adding the new states to the conditionally independent filter via conditionally independent augmentation.

3. The method of claim 2, further comprising:

determining whether any states in the first local conditioning node or merged conditioning node are unnecessary and removable; and

if there are any states that are unnecessary and removable, marginalizing out the unnecessary states from a joint-distribution of the conditionally independent filter.

4. The method of claim 3, further comprising:

determining whether any data from the remote platform include states for processing an earlier mutual measurement;

updating the conditionally independent filter with the mutual measurement when present;

determining whether the conditionally independent filter is waiting to process any other mutual measurements; and

if there are no other mutual measurements to process, marginalizing states from the conditionally independent filter that were previously factored for processing mutual measurements.

5. The method of claim 4, further comprising:

determining whether conditioning nodes have been received from one or more other remote platforms; and

if conditioning nodes have been received from one or more other remote platforms, repeating the method starting at (f).

6. The method of claim 5, wherein if conditioning nodes have not been received from one or more other remote platforms, the method further comprising:

determining whether to initiate a basic conditional update data exchange with the one or more other remote platforms;

if the basic conditional update data exchange is to be initiated, marginalizing the conditionally independent filter to obtain a distribution of a second local conditioning node; and

sending the second local conditioning node to the one or more other remote platforms.

7. The method of claim 6, further comprising:

determining whether the local platform has made a mutual measurement to the one or more other remote platforms; and

if the local platform has made a mutual measurement to the one or more other remote platforms, factoring the conditionally independent filter to continue estimation of a necessary subset of a full-state for later processing of the mutual measurement made to the one or more other remote platforms.

8. The method of claim 7, further comprising:

determining whether the one or more other remote platforms has made a mutual measurement to the local platform;

if the one or more other remote platforms have made a mutual measurement to the local platform, then for each of the one or more other remote platforms, marginalizing the conditionally independent filter to obtain a joint-distribution of another local conditioning node and new states to be augmented into the another local conditioning node; and

sending the another local conditioning node and new states to be augmented to each of the one or more other remote platforms, respectively.

9. The method of claim 8, wherein if there are no other remote platforms to consider, repeating the method starting at (b) where the conditionally independent filter is propagated forward in time.

10. The method of claim 1, wherein if a remote conditioning node has not arrived from the remote platform, the method further comprising:

determining whether to initiate a basic conditional update data exchange with one or more other remote platforms;

if the basic conditional update data exchange is to be initiated, then for each of the one or more other remote platforms, marginalizing the conditionally independent filter to obtain a distribution of another local conditioning node; and

sending the another local conditioning node to each of the one or more other remote platforms, respectively.

11. The method of claim 1, wherein the first local conditioning node and the remote conditioning node are defined in a dynamic Bayesian network graph.

12. The method of claim 1, wherein the local platform is located on a manned aircraft, an unmanned aerial vehicle, a ground vehicle, or a person.

13. The method of claim 12, wherein the remote platform is located on a manned aircraft, an unmanned aerial vehicle, a ground vehicle, or a person.

14. A navigation system for a local platform in a collaborative navigation network, the system comprising:

a dead-reckoning unit onboard the local platform that outputs a navigation solution;

a navigation filter in operative communication with the dead-reckoning unit;

a communications unit that shares data with one or more remote platforms in the network; and

a collaborative navigation processing unit in operative communication with the navigation filter, the collaborative navigation processing unit configured to execute processor readable instructions to perform a method comprising;

defining a conditioning node for at least one pair of platforms that includes the local platform and at least one remote platform in the collaborative navigation network;

receiving an estimated joint distribution of the conditioning node from the remote platform at the local platform along with new states to be augmented into the navigation filter;

performing a node-to-node fusion in the local platform to merge the remote platform'"'"'s estimated joint distribution of the conditioning node with an estimated joint distribution of the conditioning node at the local platform to obtain a fused conditioning node estimate; and

performing a conditional update in the navigation filter of the local platform based on the fused conditioning node estimate.

15. The system of claim 14, wherein the dead-reckoning unit comprises an inertial measurement unit that sends inertial data to a strapdown navigation module.

16. The system of claim 14, wherein the communications unit is wired or wireless.

17. The system of claim 14, further comprising:

one or more aiding sensors that send local sensor data to the navigation filter; and

a ranging radio that determines a relative range measurement between the local platform and the at least one remote platform.

18. The system of claim 17, wherein the collaborative navigation processing unit is configured to execute processor readable instructions to perform the method further comprising:

generating a relative range measurement at the local platform; and

performing a measurement update in the local platform with the relative range measurement after performing the conditional update.

19. The system of claim 14, wherein the local platform is located on a manned aircraft, an unmanned aerial vehicle, a ground vehicle, or a person.

20. A computer program product, comprising:

a non-transitory computer readable medium having instructions stored thereon executable by a processor to perform a method for collaborative navigation between at least one pair of platforms, the method comprising;

defining a conditioning node for the pair of platforms based on local states and prior data measurements of the pair of platforms;

generating a mutual measurement at a first platform in the pair of platforms;

receiving an estimated joint distribution of the conditioning node from a second platform in the pair of platforms at the first platform along with new states to be augmented into a navigation filter of the first platform, wherein new states to be added to the conditioning node are maintained in a navigation filter of the second platform;

performing a node-to-node fusion in the first platform to merge the second platform'"'"'s estimated joint distribution of the conditioning node with an estimated joint distribution of the conditioning node at the first platform to obtain a fused conditioning node estimate;

performing a conditional update in the navigation filter of the first platform based on the fused conditioning node estimate, wherein the conditional update is applied to both the first platform'"'"'s current navigation filter and new conditioning nodes states;

augmenting the new states from the second platform into the navigation filter of the first platform and removing any unnecessary states from the navigation filter of the first platform; and

performing a measurement update in the first platform with the mutual measurement.