Performance of inertial sensing systems using dynamic stability compensation
AssigneeYost Labs Inc.
InventorPaul Yost
Life of this patent
Estimating the term…
Abstract
A system is provided for improving performance of inertial sensing systems. The system uses dynamic stability compensation to adapt to the changing state of the inertial system such that the system remains ultra-responsive during dynamic events, while also exhibiting ultra stable output during subtle motions or rest conditions. The system uses a method that operates upon output data as it exits the component sensors but before it enters the sensor fusion that produces a final orientation. The method can be used with any sensor fusion algorithm to provide improved noise and stability performance in the output while having little to no impact upon responsiveness of that output during dynamic conditions.
Claims18
An inertial sensing system comprising:
- a plurality of sensors;
- a processing element coupled to the plurality of sensors;
- and
- a memory element coupled to the processing element,
- wherein the processing element is operable to:
- evaluate output measurement data from at least two of the sensors and compute an initial orientation;
- apply a smoothing factor to the output measurement data to produce smoothed output measurement data for the at least two sensors, wherein the smoothing factor is derived from a dynamic smoothing value that dynamically adjusts based on changing output measurement data from the at least two sensors, and wherein applying the smoothing factor further comprises:
- determining the dynamic smoothing value based on application of a filter quaternion to a quaternion of the initial orientation, wherein the filter quaternion is produced by a prior sensor fusion of a sensor fusion filter;
- interpolating between minimum and maximum values for the smoothing factor using the dynamic smoothing value to produce an interpolated value;
- and
- determining the smoothing factor based on the interpolated value;
- and
- perform sensor fusion by the sensor fusion filter with the smoothed output measurement data to produce a resulting orientation from the inertial sensing system.
A method of producing measurements in an inertial sensing system including a plurality of sensors and a processing element coupled to the plurality of sensors, the method comprising:
- generating output measurement data by the plurality of sensors;
- evaluating, via the processing element, the output measurement data from at least two of the sensors and computing an initial orientation;
- applying, via the processing element, a smoothing factor to the output measurement data to produce smoothed output measurement data for the at least two sensors, wherein the smoothing factor is derived from a dynamic smoothing value that dynamically adjusts based on changing output measurement data from the at least two sensors, and wherein applying the smoothing factor further comprises:
- determining the dynamic smoothing value based on application of a filter quaternion to a quaternion of the initial orientation, wherein the filter quaternion is produced by a prior sensor fusion of a sensor fusion filter;
- interpolating between minimum and maximum values for the smoothing factor using the dynamic smoothing value to produce an interpolated value;
- and
- determining the smoothing factor based on the interpolated value;
- and
- performing, via the processing element, sensor fusion by the sensor fusion filter with the smoothed output measurement data to produce a resulting orientation from the inertial sensing system.
Dependencies are inferred from the claim text.