阅读说明：本技术 一种无人机imu采集数据的预处理方法 (Preprocessing method for IMU (inertial measurement Unit) acquired data of unmanned aerial vehicle ) 是由 周起如 黄桃丽 辛艳峰 张晨 于 2020-09-22 设计创作，主要内容包括：本发明提出了一种无人机IMU采集数据的预处理方法,包括以下步骤：S1、获取IMU采集的无人机飞行状态参数值；S2、基于IMU采集的飞行状态参数值划分滤波器的计算单元；S3、对每个计算单元分别进行三次多项式拟合计算,确定系数矩阵；S4、基于系数矩阵确定矩阵方程,计算无人机对应时刻值滤波后的飞行状态参数值；S5、基于滤波后的数据划分滤波器的下一个计算单元,重复步骤S3至S5,实现对飞行状态参数的迭代滤波处理；S6、输出经过滤波处理的无人机飞行状态参数值。本发明基于多项式的拟合算法处理数据,降低了采集数据中的噪声和峰值,提高无人机飞行姿态解算的稳定性和精确性,促进无人机的进一步的应用与发展。(The invention provides a preprocessing method for IMU (inertial measurement unit) acquired data of an unmanned aerial vehicle, which comprises the following steps: s1, acquiring unmanned aerial vehicle flight state parameter values acquired by the IMU; s2, dividing a calculation unit of the filter based on the flight state parameter value acquired by the IMU; s3, performing cubic polynomial fitting calculation on each calculation unit respectively to determine a coefficient matrix; s4, determining a matrix equation based on the coefficient matrix, and calculating the flight state parameter value after the unmanned aerial vehicle corresponding time value is filtered; s5, dividing the next calculation unit of the filter based on the filtered data, and repeating the steps S3 to S5 to realize iterative filtering processing of the flight state parameters; and S6, outputting the flight state parameter values of the unmanned aerial vehicle after filtering. The data are processed by the fitting algorithm based on the polynomial, so that the noise and peak value in the acquired data are reduced, the stability and accuracy of resolving the flight attitude of the unmanned aerial vehicle are improved, and the further application and development of the unmanned aerial vehicle are promoted.)

1. The preprocessing method for the IMU (inertial measurement Unit) data acquisition of the unmanned aerial vehicle is characterized by comprising the following steps of:

s1, acquiring unmanned aerial vehicle flight state parameter values acquired by the IMU;

s2, dividing a calculation unit of the filter based on the flight state parameter value acquired by the IMU;

s3, performing cubic polynomial fitting calculation on each calculation unit respectively to determine a coefficient matrix;

s4, determining a matrix equation based on the coefficient matrix, and calculating the flight state parameter value after the unmanned aerial vehicle corresponding time value is filtered;

s5, dividing the next calculation unit of the filter based on the filtered flight state parameter values, and repeating the steps S3-S5 to realize iterative filtering processing of the flight state parameters;

and S6, outputting the flight state parameter values of the unmanned aerial vehicle after filtering.

2. The method for preprocessing the data collected by the IMU of the drone of claim 1, wherein in step S1, taking the three-dimensional linear acceleration value of the accelerometer of the drone collected by the IMU as an example, the following are obtained:

acc _ x 1: linear acceleration of the body in m/s along the X-axis of a reference coordinate system²；

Acc _ y 1: linear acceleration of the body in m/s along the Y-axis of a reference coordinate system²；

Acc _ z 1: linear acceleration of the body in m/s along the Z-axis of a reference coordinate system²；

The data acquisition period dt is set to 0.02 in s.

3. The method of claim 2, wherein in Step S2, the window length of the filter is set to L ═ a, and the moving Step size Step of the calculating unit is set to b, then the calculating unit is divided into the following parts:

the first computing unit is the first a three-dimensional linear acceleration values acquired by the IMU;

the second calculating unit is the last b three-dimensional linear acceleration values updated by the filtering of the first calculating unit and is added with the b three-dimensional linear acceleration values immediately behind the first calculating unit;

the m-th computing unit is the last b three-dimensional linear acceleration values updated by the filtering of the m-1-th computing unit, and is added with the b three-dimensional linear acceleration values immediately after the m-1-th computing unit.

4. The method of claim 3, wherein a is 10 and b is 5, ensuring that the computing unit always maintains 10 periods of three-dimensional linear acceleration values as input to a filter, the input to the filter being expressed as:

wherein: w represents a three-dimensional linear acceleration value matrix, Acc — x1, Acc _ y1, Acc _ z 1.

5. The method of pre-processing data acquired by an unmanned IMU of claim 1, wherein in step S3, the cubic polynomial function is y (x) - ω₀+ω₁x+ω₂x²+ω₃x³(ii) a Order to

Wherein: x represents a time value corresponding to the filtering data, X represents a coefficient matrix obtained by a cubic polynomial expression, M represents a fitting coefficient matrix of the cubic polynomial, and y represents a linear acceleration value along a certain dimension direction in the coordinate system X, Y, Z;

the cubic polynomial function can be converted into a linear algebraic form to obtain a coefficient matrix:

y(x,M)＝XM。

6. the method for preprocessing IMU data acquired by the UAV of claim 1, wherein in step S5, the filtered last 5 flight state parameter values are used as the first 5 flight state parameter values of the next calculation unit, and the steps S3 to S5 are repeated to realize the iterative filtering processing of the flight state parameters.

Technical Field

The invention relates to the technical field of unmanned aerial vehicles, in particular to a preprocessing method for IMU (inertial measurement unit) acquired data of an unmanned aerial vehicle.

Background

With the rapid development of unmanned aerial vehicle technology, the unmanned aerial vehicle is gradually and widely applied to the military and civil fields. An Inertial Measurement Unit (IMU) is a measurement unit for acquiring the state parameter information of a carrier in real time, such as an accelerometer, a gyroscope, a magnetometer and the like, and the IMU has the characteristics of low price and capability of stably and quickly acquiring the state parameter information of a machine body. The application of IMU on unmanned aerial vehicle has reduced manufacturing cost, has further promoted unmanned aerial vehicle's popularization and application.

Along with each field of wide application of unmanned aerial vehicle, the service environment that leads to IMU has the uncertainty, and elements such as gyroscope and accelerometer are easily influenced by external environment such as temperature, vibration, lead to the data of gathering to have drift or the great problem of noise. Therefore, preprocessing of IMU data by the drone before performing attitude resolution is necessary.

Disclosure of Invention

Based on the technical problems in the background art, the invention provides a preprocessing method for IMU (inertial measurement unit) data acquisition of an unmanned aerial vehicle.

The invention provides a preprocessing method for IMU (inertial measurement unit) acquired data of an unmanned aerial vehicle, which comprises the following steps:

s1, acquiring unmanned aerial vehicle flight state parameter values acquired by the IMU;

s2, dividing a computing unit of the filter based on the flight state parameter value of the unmanned aerial vehicle acquired by the IMU;

s3, performing cubic polynomial fitting calculation on each calculation unit respectively to determine a coefficient matrix;

s4, determining a matrix equation based on the coefficient matrix, and calculating the flight state parameter value after the unmanned aerial vehicle corresponding time value is filtered;

s5, dividing the next calculation unit of the filter based on the filtered flight state parameter values, and repeating the steps S3-S5 to realize iterative filtering processing of the flight state parameters;

and S6, outputting the flight state parameter values of the unmanned aerial vehicle after filtering.

Preferably, in step S1, taking the three-dimensional linear acceleration value of the accelerometer of the drone collected by the IMU as an example, the following are obtained:

acc _ x 1: linear acceleration of the body in m/s along the X-axis of a reference coordinate system²；

Acc _ y 1: linear acceleration of the body in m/s along the Y-axis of a reference coordinate system²；

Acc _ z 1: linear acceleration of the body in m/s along the Z-axis of a reference coordinate system²；

The data acquisition period dt is set to 0.02 in s.

Preferably, in Step S2, the calculating unit is divided by the window length L of the filter, and the moving Step of the calculating unit is set to b, and the calculating unit is divided as follows:

the first computing unit is the first a three-dimensional linear acceleration values acquired by the IMU;

the second calculating unit is the last b three-dimensional linear acceleration values updated by the filtering of the first calculating unit and is added with the b three-dimensional linear acceleration values immediately behind the first calculating unit;

the m-th computing unit is the last b three-dimensional linear acceleration values updated by the filtering of the m-1-th computing unit, and is added with the b three-dimensional linear acceleration values immediately after the m-1-th computing unit.

Preferably, setting a to 10 and b to 5 ensures that the computing unit always maintains 10 cycles of three-dimensional linear acceleration values as the input to the filter, which is expressed as:

wherein: w represents a three-dimensional linear acceleration value matrix, Acc — x1, Acc _ y1, Acc _ z 1.

Preferably, in step S3, the cubic polynomial function is y (x) ═ ω₀+ω₁x+ω₂x²+ω₃x³(ii) a Order to

Wherein: x represents a time value corresponding to the filtering data, X represents a coefficient matrix obtained by a cubic polynomial expression, M represents a fitting coefficient matrix of the cubic polynomial, and y represents a linear acceleration value along a certain dimension direction in the coordinate system X, Y, Z;

the cubic polynomial function can be converted into a linear algebraic form to obtain a coefficient matrix:

y(x,M)＝XM。

preferably, in step S5, the last 5 filtered flight state parameter values are used as the first 5 flight state parameter values of the next calculating unit, and steps S3 to S5 are repeated to realize the iterative filtering process for the flight state parameters.

According to the preprocessing method for the IMU data acquired by the unmanned aerial vehicle, data are processed based on a polynomial fitting algorithm, noise and peak values in the acquired data are reduced, the stability and accuracy of unmanned aerial vehicle flight attitude calculation are improved, and further application and development of the unmanned aerial vehicle are promoted; the internal characteristics of the unmanned aerial vehicle state parameters of the acquired data are reserved by adopting the data association iterative computation of the front and back computation; the processing data of the latter computing unit comprises the data after the filtering processing of the former computing unit, the connectivity between the former computing unit and the latter computing unit is ensured, the precision and the stability are effectively improved, and the basis is set for the stable flight cushion of the unmanned aerial vehicle.

Drawings

Fig. 1 is a flowchart of a preprocessing method for data acquired by an unmanned aerial vehicle IMU according to the present invention.

Detailed Description

Referring to fig. 1, the invention provides a preprocessing method for data acquired by an unmanned aerial vehicle IMU, which comprises the following steps:

s1, acquiring unmanned aerial vehicle flight state parameter values acquired by the IMU, and acquiring the three-dimensional linear acceleration value of the accelerometer as an example:

acc _ x 1: body along a reference coordinate systemLinear acceleration value of X axis in m/s²；

Acc _ y 1: linear acceleration of the body in m/s along the Y-axis of a reference coordinate system²；

Acc _ z 1: linear acceleration of the body in m/s along the Z-axis of a reference coordinate system²；

The data acquisition period dt is set to 0.02 in s.

S2, a calculating unit of a three-dimensional linear acceleration value dividing filter of the unmanned aerial vehicle accelerometer based on IMU acquisition;

in the method, a calculation unit is divided by taking the window length of a filter as L (10) (a state parameter value of a sampling period), and the moving Step length Step of the calculation unit is set as 5, then the calculation unit is divided as follows:

the first computing unit is the first 10 three-dimensional linear acceleration values acquired by the IMU;

the second computing unit adds the last 5 three-dimensional linear acceleration values updated by the filtering of the first computing unit to the 5 three-dimensional linear acceleration values immediately behind the first computing unit;

the m-th computing unit is the last 5 three-dimensional linear acceleration values updated by the m-1 th computing unit by filtering, and adds the 5 three-dimensional linear acceleration values immediately after the m-1 th computing unit, so as to ensure that the computing unit always keeps 10 cycles of three-dimensional linear acceleration values as the input of the filter, and the input of the filter is expressed as:

wherein: w represents a three-dimensional linear acceleration value matrix, Acc — x1, Acc _ y1, Acc _ z 1.

S3, performing cubic polynomial fitting calculation on each calculation unit respectively to determine a coefficient matrix;

the function of the cubic polynomial being y (x) ω₀+ω₁x+ω₂x²+ω₃x³(ii) a Order to

Wherein: x represents a time value corresponding to the filtering data, X represents a coefficient matrix obtained by a cubic polynomial expression, M represents a fitting coefficient matrix of the cubic polynomial, and y represents a linear acceleration value along a certain dimension direction in the coordinate system X, Y, Z;

the cubic polynomial function can be converted into a linear algebraic form to obtain a coefficient matrix:

y(x,M)＝XM；

setting a quality evaluation function, selecting E_RMS(root mean square error) as follows:

wherein: n represents the filter window length;

based on the minimum error rule, a linear regression equation is obtained.

And S4, determining a matrix equation based on the coefficient matrix, and calculating the three-dimensional linear acceleration value after the unmanned aerial vehicle corresponding time value is filtered.

S5, dividing the next calculation unit of the filter based on the filtered three-dimensional linear acceleration value, and repeating the steps S3 to S5 to realize iterative filtering processing of the three-dimensional linear acceleration value;

and taking the filtered last 5 three-dimensional linear acceleration values as the first 5 three-dimensional linear acceleration values of the next calculation unit to ensure the connectivity between the previous and next calculation units, wherein half of the three-dimensional linear acceleration values are the three-dimensional linear acceleration values subjected to filtering in the next calculation, so that the accuracy of the three-dimensional linear acceleration values calculated by polynomial fitting of the current calculation unit is higher.

S6, outputting the three-dimensional linear acceleration value of the unmanned aerial vehicle accelerometer subjected to filtering processing;

the three-dimensional linear acceleration value acquired by the IMU is continuously subjected to iterative computation and updating, so that the preprocessing computation of the three-dimensional linear acceleration value is completed, and a solid foundation is established for the subsequent state estimation.

The above description is only for the preferred embodiment of the present invention, but the scope of the present invention is not limited thereto, and any person skilled in the art should be considered to be within the technical scope of the present invention, and the technical solutions and the inventive concepts thereof according to the present invention should be equivalent or changed within the scope of the present invention.