CN-121994223-A - Personal high-precision inertial navigation method, device, medium and product

CN121994223ACN 121994223 ACN121994223 ACN 121994223ACN-121994223-A

Abstract

The application discloses a personal high-precision inertial navigation method, equipment, medium and product, and relates to the field of inertial navigation, wherein the method comprises the steps of identifying and processing inertial measurement data time sequences by utilizing a deep learning model so as to obtain a motion mode and corresponding confidence coefficient of a user; the method comprises the steps of determining a virtual observed quantity and a corresponding observation matrix at a current moment according to motion constraint corresponding to a motion mode, adaptively adjusting noise covariance of the virtual observed quantity according to confidence coefficient, calculating Kalman gain according to prior error covariance at the current moment, the virtual observed quantity, the corresponding observation matrix and the noise covariance, updating error state estimation and the corresponding covariance according to the Kalman gain to obtain posterior error state and posterior error covariance at the current moment, and determining posterior navigation state at the current moment by combining with prior navigation state. The application can realize the real-time error correction during the continuous motion of the user and improve the inertial navigation precision.

Inventors

WANG XIAOLEI
GE XINPING
DU HAIMING
LOU TAISHAN
JIAO YUZHAO
DING GUOQIANG

Assignees

郑州轻工业大学

Dates

Publication Date: 20260508
Application Date: 20260224

Claims (10)

1. A personal high precision inertial navigation method, comprising: Acquiring inertial measurement data output by an inertial measurement unit at the current moment, wherein the inertial measurement data comprises acceleration and angular velocity; calculating the prior navigation state and the prior error covariance at the current moment by utilizing an inertial navigation solution algorithm according to the optimal navigation state estimation and the posterior error covariance at the last moment and combining the inertial measurement data; According to the inertial measurement data time sequence output by the inertial measurement unit at the current moment and the historical moment, performing recognition processing by using a deep learning model to obtain the current motion mode and the corresponding confidence coefficient of the user, wherein the time length corresponding to the time sequence is a preset time length; determining a motion constraint corresponding to a current motion mode of a user, and further determining a virtual observed quantity and a corresponding observed matrix at the current moment according to the motion constraint; According to the confidence coefficient, the noise covariance of the virtual observed quantity is adaptively adjusted, wherein the higher the confidence coefficient is, the smaller the noise covariance of the virtual observed quantity is; And then updating the error state estimation and the corresponding covariance according to the Kalman gain to obtain a posterior error state and a posterior error covariance at the current moment, and determining the posterior navigation state at the current moment by combining the prior navigation state.
2. A personal high precision inertial navigation method according to claim 1, wherein the navigational state includes position, velocity and attitude.
3. A personal high precision inertial navigation method according to claim 1, wherein the error states include position error, velocity error, attitude error, accelerometer zero bias and gyroscope zero bias.
4. A personal high precision inertial navigation method according to claim 1, wherein the motion patterns include stationary, walking, constant velocity traveling, creeping, going up stairs and going down stairs.
5. The personal high-precision inertial navigation method according to claim 1, wherein if the motion pattern is stationary, the corresponding motion constraint is zero-speed correction and zero-angular velocity correction; If the motion mode is walking, the corresponding motion constraint is zero-speed correction and step length constraint; if the motion mode is uniform-speed running, the corresponding motion constraint is zero-angle-rate constraint; If the motion mode is creeping, the corresponding motion constraint is vertical speed constraint; if the motion mode is ascending stairs, the corresponding motion constraint is a first vertical displacement constraint; If the motion mode is downstairs, the corresponding motion constraint is a second vertical displacement constraint.
6. The personal high-precision inertial navigation method according to claim 1, wherein the adaptively adjusting the noise covariance of the virtual observables according to the confidence level comprises: according to the confidence coefficient, a noise covariance calculation model is utilized to adaptively adjust the noise covariance of the virtual observed quantity, and the mathematical expression of the noise covariance calculation model is as follows: Wherein, the Representation of Noise covariance of time; Representing movement patterns The corresponding confidence level; Representing the minimum noise corresponding to the maximum confidence; Representing the maximum noise corresponding to the minimum confidence.
7. The personal high-precision inertial navigation method according to claim 1, wherein the Kalman gain is calculated by using a Kalman gain calculation model, and the Kalman gain calculation model has a mathematical expression: ; Wherein, the Representation of A Kalman gain at time; Is shown in Time of day and Corresponding a priori error states Covariance matrix of (a), i.e.) A priori error covariance of the moment; Is shown in At the moment only use A priori navigation state determined by the posterior navigation state at the moment; Presetting to 0; Representation of An observation matrix corresponding to the moment virtual observation quantity; Representation of Time noise covariance, superscript Representing the transpose and superscript-1 representing the inverse of the matrix.
8. A computer device comprising a memory, a processor and a computer program stored on the memory and capable of running on the processor, characterized in that the processor executes the computer program to implement a personal high precision inertial navigation method according to any one of claims 1-7.
9. A computer readable storage medium, on which a computer program is stored, characterized in that the computer program, when being executed by a processor, implements a personal high precision inertial navigation method according to any one of claims 1-7.
10. A computer program product comprising a computer program, characterized in that the computer program, when executed by a processor, implements a personal high precision inertial navigation method according to any one of claims 1-7.

Description

Personal high-precision inertial navigation method, device, medium and product Technical Field The application relates to the technical field of inertial navigation, in particular to a personal high-precision inertial navigation method, equipment, medium and product. Background Conventional Personal Inertial Navigation Systems (PINS) drift rapidly after long periods of operation due to the accumulation of Inertial Measurement Unit (IMU) sensor errors. This accumulation is especially severe in indoor, underground, or urban canyon environments where the GNSS signals are occluded. The prior art relies primarily on zero speed correction (ZUPT) applied at stationary moments. However, there is a lack of effective, reliable real-time correction of errors while the user is in continuous motion. Disclosure of Invention The application aims to provide a personal high-precision inertial navigation method, equipment, media and products, which can realize real-time error correction during continuous motion of a user and improve the inertial navigation precision. In order to achieve the above object, the present application provides the following solutions: in a first aspect, the present application provides a personal high precision inertial navigation method, comprising: Acquiring inertial measurement data output by an inertial measurement unit at the current moment, wherein the inertial measurement data comprises acceleration and angular velocity; calculating the prior navigation state and the prior error covariance at the current moment by utilizing an inertial navigation solution algorithm according to the optimal navigation state estimation and the posterior error covariance at the last moment and combining the inertial measurement data; According to the inertial measurement data time sequence output by the inertial measurement unit at the current moment and the historical moment, performing recognition processing by using a deep learning model to obtain the current motion mode and the corresponding confidence coefficient of the user, wherein the time length corresponding to the time sequence is a preset time length; determining a motion constraint corresponding to a current motion mode of a user, and further determining a virtual observed quantity and a corresponding observed matrix at the current moment according to the motion constraint; According to the confidence coefficient, the noise covariance of the virtual observed quantity is adaptively adjusted, wherein the higher the confidence coefficient is, the smaller the noise covariance of the virtual observed quantity is; And then updating the error state estimation and the corresponding covariance according to the Kalman gain to obtain a posterior error state and a posterior error covariance at the current moment, and determining the posterior navigation state at the current moment by combining the prior navigation state. In a second aspect, the present application provides a computer device comprising a memory, a processor to store a computer program on the memory and executable on the processor, the processor executing the computer program to implement the steps of a personal high precision inertial navigation method of the first aspect. In a third aspect, the present application provides a computer readable storage medium having stored thereon a computer program which when executed by a processor implements the steps of a personal high precision inertial navigation method of the first aspect. In a fourth aspect, the present application provides a computer program product comprising a computer program which, when executed by a processor, implements the steps of a personal high precision inertial navigation method of the first aspect. According to the specific embodiment provided by the application, the application has the following technical effects: the application provides a personal high-precision inertial navigation method, equipment, medium and product, which are characterized in that a deep learning model is utilized to identify an inertial measurement data time sequence so as to obtain a motion mode and corresponding confidence coefficient of a user, then a virtual observed quantity and a corresponding observation matrix at the current moment are determined according to motion constraint corresponding to the motion mode, the noise covariance of the virtual observed quantity is adaptively adjusted according to the confidence coefficient, finally, kalman gain is calculated according to the prior error covariance at the current moment, the virtual observed quantity and the corresponding observation matrix and the noise covariance, error state estimation and the corresponding covariance are updated according to the Kalman gain, the posterior error state and the posterior error covariance at the current moment are obtained, and the posterior navigation state at the current moment is determined by combining with the prior navigation state. Compared with the