# US 8,280,110 B2 · State estimating apparatus and state estimating program

> Agents: this data is available as JSON from the patentgrep API and MCP server — see https://patentgrep.com/llms.txt

- Number: US 8,280,110 B2
- Document ID: US8280110B2
- Status: Granted
- Kind: B2
- Application number: 12/475,779
- Filed: 2009-06-01
- Published: 2009-12-10
- Granted: 2012-10-02
- Earliest claimed: 2008-06-09
- Assignees: Honda Motor Co., Ltd.
- Inventors: Tadaaki Hasegawa
- Web page: https://patentgrep.com/patent/US8280110B2

## CPC

- G06T 7/277: involving stochastic approaches, e.g. using Kalman filters
- G06T 2207/10016: Video; Image sequence
- G06T 2207/30224: Ball; Puck
- G06T: IMAGE DATA PROCESSING OR GENERATION, IN GENERAL

## Abstract

A state estimating apparatus permits efficient, highly accurate estimation of the state of an object. A particle in a state variable space defined by a second state variable preferentially remains or increases as the likelihood thereof relative to a current measured value of a first state variable is higher, while a particle is preferentially extinguished as the likelihood thereof is lower. A particle which transitions in the state variable space according to a state transition model with a high probability of being followed by an object (a high-likelihood model) as a next model tends to increase. On the other hand, although in a small quantity, there are particles having models (low-likelihood models) which are different from the high-likelihood model as their unique models.

## Claims

### Claim 1

A state estimating apparatus comprising:
a first processor element configured to: measure a value of a first state variable indicative of a state of an object; disperse at least some particles among a plurality of particles around previous positions in a state variable space defined by a second state variable; transit each of the plurality of particles to a current position according to one current model selected from a plurality of models, such that a plurality of particle groups which each include particles having a common current model are formed, wherein the current model common to the particles within each of said plurality of particle groups is a unique current model for that particle group which fluidly changes for each particle within said particle group; assess a likelihood of each particle relative to a current measured value of the first state variable; and estimate a current state of the object on the basis of the likelihood; and
a second processor element configured to: allow particles from among the plurality of particles which have been assessed by the first processor element to have a likelihood greater than a predetermined high-likelihood threshold to remain or increase; extinguish particles from among the plurality of particles which have been assessed by the first processor element to have a likelihood less than a predetermined low-likelihood threshold, said low-likelihood threshold being equal to or less than the high-likelihood threshold; and then determine a next model from the plurality of models for each particle on the basis of the current model for each particle.

### Claim 2 (depends on 1)

The state estimating apparatus according to claim 1, wherein
the second processor element is configured to determine a current model for each particle according to one region to which a current value of a model variable, which is unique to each particle and which fluidly changes, belongs among a plurality of regions defined in correspondence with each of the plurality of models.

### Claim 3 (depends on 2)

The state estimating apparatus according to claim 2, wherein
the second processor element is configured to change a next value such that a perturbation amount from a current value to a next value of the model variable falls within a permissible range, and to determine a next model from among the plurality of models for each particle according to a region to which the next value belongs.

### Claim 4 (depends on 1)

The state estimating apparatus according to claim 1, wherein
the first processor element is configured to estimate a previous state transition model of the object as the current model common to one of the particle groups whose total sum of the likelihoods is greater than that of all other particle groups.

### Claim 5 (depends on 1)

The state estimating apparatus according to claim 1, wherein
the first processor element is configured to estimate a weighted mean value, which uses the likelihood of the current position of each particle as the weight, as a current value of the second state variable of the object.

### Claim 6

A state estimating program recorded on a non-transitory computer-readable medium, the state estimating program configured to cause a computer to function as a state estimating apparatus, the state estimating apparatus comprising:
a first element which measures a value of a first state variable indicative of a state of an object, disperses at least some particles among a plurality of particles around previous positions in a state variable space defined by a second state variable, transits each of the plurality of particles to a current position according to a one current model selected from a plurality of models, such that a plurality of particle groups which each include particles having a common current model are formed, and each of said plurality of particle groups have a unique current model which fluidly changes for each particle within said particle group, assesses a likelihood of each particle relative to a current measured value of the first state variable, and estimates a current state of the object on the basis of the likelihood; and
a second element which allows particles from among the plurality of particles which have been assessed by the first element to have a likelihood greater than a predetermined high-likelihood threshold to remain or increase, and extinguishes particles which have been assessed by the first element to have a likelihood less than a predetermined low-likelihood threshold, and then determines a next model from the plurality of models each particle on the basis of the current model from the plurality of models each particle.
