CN117849395A - Method for calculating carrier acceleration by complementation of high-low frequency noise characteristics - Google Patents

Abstract

The invention discloses a method for calculating carrier acceleration by complementation of high-low frequency noise characteristics, and belongs to the field of navigation. The method firstly utilizes the measured speed to extract the motion acceleration through a sliding recurrence window, and then fuses the motion acceleration with the acceleration measured by an inertial measurement unit through a complementary filter. The window recursion estimation is beneficial to more accurately predicting the motion of which the motion parameters accord with the higher order function and is beneficial to simultaneously reducing the high-frequency noise and the low-frequency noise of different systems. According to the invention, as the technical scheme of window recursion estimation and a complementary filter is adopted, the characteristics that the noise of an inertial navigation signal and a satellite navigation signal is distributed in different frequency ranges are utilized to fuse the two signals, so that the noise is sufficiently reduced; and for scenes in which some motions meet a certain function, the window recursion estimation fitting function is used for improving the prediction precision, so that the integrated navigation precision is improved.

Description

Method for calculating carrier acceleration by complementation of high-low frequency noise characteristics

Technical Field

The invention relates to the field of inertial navigation and satellite navigation. And more particularly to a method for extracting carrier acceleration and fusion acceleration using window recursive estimation based solely on velocity data.

Background

The use of a satellite navigation system or other navigation systems to fuse with an inertial navigation system to estimate acceleration is an important method for improving the accuracy of velocity and position estimation, and a kalman filter is generally used for fusion.

However, the above method has two problems: (1) Some navigation systems, such as GNSS, cannot directly measure acceleration, and have certain limitations in application scenarios. (2) The estimation accuracy is insufficient for situations where the motion exhibits a high-order function. (3) The system characteristics are different, for example, the measurement noise of a satellite navigation system is high-frequency noise, the measurement noise of an inertial measurement unit is low-frequency noise relative to a satellite, and a method for reducing the influence of the two noises on the precision is lacked.

Disclosure of Invention

In view of the above, the present invention proposes an acceleration estimation method using a window recursive estimation and a complementary filter.

The technical scheme of the invention is as follows: a method of calculating carrier acceleration with complementary high and low frequency noise characteristics, the method comprising:

step 1: acquiring acceleration of the carrier by using the gyroscope and the acceleration data;

a _i ＝a _ai -b _a

wherein a is _ai A is the original reading of the accelerometer _i For entering the acceleration values measured in the carrier coordinate system, b _g Zero offset for gyroscope calibrated in advance, b _a Zero offset, a, for a previously calibrated gyroscope _ins For the acceleration measured by the inertial measurement unit under the navigational coordinate system,for the posture matrix of b-series relative to n-series, < >>For +.>t _m-1 T is the last time, t _m The time is the present moment;

step 2: acquiring historical speed data by using a navigation system to estimate acceleration;

setting the speed measured by the navigation system as v, and the noise eta as the undetermined parameter in relation to the sampling time delta t;

wherein l is the window length, k is the sampling frequency, p is the set frequency, I _3×3 The least squares solution for beta, which is the identity matrix, is:

wherein v is _k-l:k For the velocity vector from the k-th moment to the k moment,covariance matrix of velocity vector noise;

calculating a state to be estimated:

wherein v is _k At the speed of the kth moment,for the estimated acceleration, u _Δt Is that

Step 3: fusing the estimated acceleration with the acceleration measured by the inertial measurement unit by using complementary filtering;

calculating a covariance matrix of process noise of the acceleration to be estimated:

wherein,

fusion acceleration:

wherein a is _ins Acceleration measured by an inertial measurement unit, Q _ins Is a process noise matrix of the inertial measurement unit.

According to the invention, as the technical scheme of window recursion estimation and a complementary filter is adopted, the characteristics that the noise of an inertial navigation signal and a satellite navigation signal is distributed in different frequency ranges are utilized to fuse the two signals, so that the noise is sufficiently reduced; and for scenes in which some motions meet a certain function, the window recursion estimation fitting function is used for improving the prediction precision, so that the integrated navigation precision is improved.

Drawings

FIG. 1 is a schematic overall flow chart of the present invention.

FIG. 2 is a graph showing the experimental effect of the present invention, wherein KF represents the Kalman filter method and WRA represents the method of the present invention.

Detailed Description

Step 1: and using the gyroscope and accelerometer data to perform respective angular speeds and accelerations of the inertial measurement units.

Most modern inertial measurement units can directly output the triaxial acceleration and triaxial angular velocity in the carrier coordinate system.

ω _i ＝ω _gi -b _g

a _i ＝a _ai -b _a

Wherein omega _gi Omega as raw reading _i To enter the angular velocity value, a, of the step 2 Kalman filter _ai A is the original reading of the accelerometer _i For entering the acceleration values measured in the carrier coordinate system, b _g Zero offset for a gyroscope calibrated in advance. b _a Zero offset for a gyroscope calibrated in advance. a, a _ins Acceleration measured by the inertial measurement unit under a navigation coordinate system.For +.>The b-series versus n-series gesture matrix. t is t _m-1 T is the last time, t _m The present time.

Step 2: extracting acceleration by using historical speed data measured by a GNSS navigation system;

let the velocity measured by a navigation system, such as a GNSS, be v, with the sampling time Δt, the noise η has the following relation. Beta is a pending parameter.

l is the window length, p is the set number of times, and the least squares solution for β is as follows:

the state to be estimated x can be solved by the following formula;

wherein:

wherein u' _Δt Can be calculated by the following formula:

step 3: solving a covariance matrix of process noise of the acceleration to be estimated:

Q _β from the relationship between β and v.

Fusion acceleration:

wherein a is _ins The acceleration obtained is measured for the inertial measurement unit. Q (Q) _ins Is a process noise matrix of the inertial measurement unit. Therefore, high-frequency noise and low-frequency noise are reduced, and the estimation accuracy of acceleration is improved.

Claims (2)

1. A method of calculating carrier acceleration with complementary high and low frequency noise characteristics, the method comprising:

step 1: acquiring acceleration of the carrier by using the gyroscope and the acceleration data;

a _i ＝a _ai -b _a

wherein a is _ai A is the original reading of the accelerometer _i For entering the acceleration values measured in the carrier coordinate system, b _g Zero offset for gyroscope calibrated in advance, b _a Zero offset, a, for a previously calibrated gyroscope _ins For the acceleration measured by the inertial measurement unit under the navigational coordinate system,for the posture matrix of b-series relative to n-series, < >>For +.>t _m-1 T is the last time, t _m The time is the present moment;

step 2: acquiring historical speed data by using a navigation system to estimate acceleration;

step 3: fusing the estimated acceleration with the acceleration measured by the inertial measurement unit by using complementary filtering;

calculating a covariance matrix of process noise of the acceleration to be estimated:

wherein,

fusion acceleration:

wherein a is _ins Acceleration measured by an inertial measurement unit, Q _ins Is a process noise matrix of the inertial measurement unit.

2. The method for calculating the acceleration of the carrier with complementary high-low frequency noise characteristics according to claim 1, wherein the specific method in the step 2 is as follows:

setting the speed measured by the navigation system as v, and the noise eta as the undetermined parameter in relation to the sampling time delta t;

wherein l is the window length, k is the sampling frequency, p is the set frequency, I _3×3 The least squares solution for beta, which is the identity matrix, is:

wherein v is _k-l:k For the velocity vector from the k-th moment to the k moment,covariance matrix of velocity vector noise;

calculating a state to be estimated:

wherein v is _k At the speed of the kth moment,for the estimated acceleration, u _Δt Is that