Method for improving dynamic gesture accuracy of inertia-geomagnetic combination

Abstract

The invention provides a method for improving dynamic gesture accuracy of inertia-geomagnetic combination. The method comprises the steps of firstly, extracting appropriate characteristic from a discrete sampling time sequence output from a three-dimensional accelerometer by an artificial neural network with regard to complicated three-dimensional motion such as human body movement; secondly, accurately estimating amplitude of carrier linear acceleration according to the characteristic by the artificial neural network; and finally, adjusting and expanding observation and process covariance ofa kalman algorithm in real time according to an estimation result, and further preventing the influence of the carrier linear acceleration on the algorithm so that the gesture resolution accuracy of the inertia-geomagnetic combination is improved.

Description

technical field [0001] The invention relates to the technical field of sensors, in particular to a method for improving the inertia-geomagnetism combination dynamic attitude determination accuracy by using an artificial neuron network. Background technique [0002] The inertial-geomagnetic combination is a combination of sensors. Its hardware is mainly composed of a three-axis accelerometer, a three-axis gyroscope, a three-axis geomagnetic sensor, and some necessary signal processing units. Among them, the signal processing unit includes a preamplifier, a band-pass filter Devices, analog-to-digital converters, central processing units, power supply circuits, communication circuits, etc., are mainly used to realize the subsequent processing and transmission of sensor signals. The combined software program is run by the central processing unit, mainly including necessary processor initialization program, sensor data acquisition program, sensor error correction program, attitud...

AI Technical Summary

Problems solved by technology

[0005] The technical problem to be solved by the present invention is: in order to solve the problem of the inertia-geomagnetism combined attitude calculation precision drop caused by carrier linear acceleration interference, the present invention provides a method for improving the inertia-geomagnetic combined dynamic attitude determination accuracy, the method first Appropriate features are constructed from the output of the accelerometer, followed by a network of artificial neurons to establish a nonlinear functional relationship between the above features and the magnitude of the linear acceleration of the carrier, and finally an artificial neural network (representing the magnitude of the linear acceleration of the carrier) The output of the meta-network adjusts the process of the extended Kalman algorithm and the observation noise covariance, thereby suppressing the influence caused by the linear acceleration of the carrier and improving the dynamic attitude calculation accuracy of the extended Kalman algorithm

[0006] At present, some targeted methods have been proposed to try to solve the problem of carrier linear acceleration interference

Images