Method and system for varying step length estimation using nonlinear system identification

Patent metadata

Field	Value
Publication number	US-10267646-B2
Application number	US-201414764519-A
Country	US
Kind code	B2
Filing date	Jan 30, 2014
Priority date	Feb 1, 2013
Publication date	Apr 23, 2019
Grant date	Apr 23, 2019

Abstract

The present disclosure relates to a method and system for estimating varying step length for on foot motion (such as for example walking or running). The present method and apparatus is able to be used in anyone or both of two different phases depending on the embodiment. The first phase is a model-building phase done offline to obtain the nonlinear model for the step length as a function of different parameters that represent human motion dynamics, the model is built using a nonlinear system identification technique. In the second phase the nonlinear model is used to calculate the step length from the different parameters that represent human motion dynamics used as input to the model. These parameters are obtained from sensors readings from the sensors in the apparatus. This second phase is the more frequent usage of the present method and apparatus for a variety of applications.

First claim

The embodiments in which an exclusive property or privilege is claimed are defined as follows:

1. A method for estimating varying step length for on foot motion of a pedestrian, the method comprising:
a) providing a device moveable with the pedestrian, the device positionable in any orientation, wherein the device comprises a sensor assembly, and wherein the sensor assembly comprises at least one sensor, the at least one sensor configured to provide sensor readings;
b) detecting steps from the sensor readings;
c) obtaining parameters that represent human motion dynamics from the sensor readings for the detected steps; and
d) using the parameters that represent human motion dynamics to:
i) build an improved, previously unknown step length model for on foot motion using a nonlinear system identification technique,
ii) utilize an improved step length model to estimate step length for on foot motion, wherein the step length model was previously unknown and was built using a nonlinear system identification technique, or
iii) build an improved, previously unknown step length model for on foot motion using a nonlinear system identification technique and utilize the step length model built to estimate step length.

2. The method of claim 1 ,
wherein detecting steps from the sensor readings comprises obtaining sensor readings for a plurality of pedestrians and running a step detection technique on the obtained sensor readings to detect steps; and
wherein building the step length model using the nonlinear system identification technique to estimate step length comprises:
obtaining a reference step length for each detected step,
feeding the parameters that represent human motion dynamics and the reference step length to the nonlinear system identification technique and
running the system identification technique to build a model to estimate a step length from the parameters that represent human motion dynamics.

3. The method of claim 1 ,
wherein detecting steps from the sensor readings comprises obtaining the sensor readings and running a step detection technique on the sensor readings to detect a step; and
wherein utilizing the step length model built using the nonlinear system identification technique comprises passing the parameters that represent human motion dynamics to the step length model and estimating an output step length from the model.

4. The method in any one of claim 1 , 2 , or 3 , wherein the method is operable with the device untethered to the pedestrian.

5. The method in any one of claim 1 , 2 , or 3 , wherein the method is operable with the device tethered to the pedestrian.

6. The method in any one of claim 1 , 2 , or 3 , wherein the nonlinear system identification technique is Fast Orthogonal Search.

7. The method in any one of claim 1 , 2 , or 3 , wherein the at least one sensor in the sensor assembly comprises an accelerometer.

8. The method of claim 7 , wherein parameters that represent human motion dynamics are one or more of the following: step frequency, acceleration variance during step, acceleration peak value during step, or peak to peak acceleration value during step.

9. The method in any one of claim 1 , 2 , or 3 , wherein the at least one sensor in the sensor assembly comprises a tri-axial accelerometer.

10. The method of claim 9 , wherein parameters that represent human motion dynamics are one or more of the following: step frequency, acceleration variance during step, acceleration peak value during step, or peak to peak acceleration value during step.

11. A system for estimating varying step length for on foot motion of a pedestrian, the system comprising:
a) a device moveable with the pedestrian, the device in any orientation, the device comprising:
i) an assembly of sensors comprising at least one sensor, the at least one sensor configured to provide sensor readings; and
ii) a receiver for receiving absolute navigational information about the device from an external source, and producing an absolute navigational information output; and

b) a processor, coupled to receive the sensor readings and the absolute navigational information output, and operative to:
i) detect steps from the sensor readings;
ii) obtain parameters that represent human motion dynamics from the sensor readings for the detected steps; and
iii) use the parameters that represent human motion dynamics to:
(A) build an improved, previously unknown step length model for on foot motion using a nonlinear system identification technique,
(B) utilize an improved step length model to estimate step length for on foot motion, wherein the model was previously unknown and was built using a nonlinear system identification technique, or
(C) build an improved, previously unknown step length model for on foot motion using a nonlinear system identification technique and utilize the step length model to estimate step length.

12. The system of claim 11 , wherein the processor is further operative to
obtain the sensor readings for a plurality of pedestrians,
obtain a reference step length for each detected step,
feed the parameters that represent human motion dynamics and the reference step length to the nonlinear system identification technique and
run the system identification technique to build the model to estimate the step length from the parameters that represent human motion dynamics.

13. The system of claim 11 , wherein the processor is further operative to
pass the parameters that represent human motion dynamics to the step length model and estimate the output step length from the model, wherein the step length model was built using the system identification technique.

14. The system of any one of claim 11 , 12 , or 13 , wherein the device is untethered to the pedestrian.

15. The system of any one of claim 11 , 12 , or 13 , wherein the device is tethered to the pedestrian.

16. The system of any one of claim 11 , 12 , or 13 , wherein the at least one sensor in the sensor assembly comprises an accelerometer.

17. The system of any one of claim 11 , 12 , or 13 , wherein the at least one sensor in the sensor assembly comprises a tri-axial accelerometer.

18. The system of any one of claim 11 , 12 , or 13 , wherein the processor is within the device.

19. The system of any one of claim 11 , 12 , or 13 , wherein the processor is not within the device.

Assignees

• Trusted Positioning Inc
• Invensense Inc

Inventors

• Omr Medhat
• Wahdan Ahmed
• Georgy Jacques
• Noureldin Aboelmagd

Classifications

• G01C22/00

  Measuring distance traversed on the ground by vehicles, persons, animals or other moving solid bodies, e.g. using odometers, using pedometers · CPC title

• A61B5/112

  Gait analysis · CPC title

• G01C21/14

  by recording the course traversed by the object (G01C21/16 takes precedence) · CPC title

• A61B5/725

  using specific filters therefor, e.g. Kalman or adaptive filters (specific diagnostics methods using using bioelectric or biomagnetic signals A61B5/316) · CPC title

• G01C22/006 Primary

  Pedometers · CPC title