JP2007232443A - Inertia navigation system and its error correction method - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To provide an inertia navigation system which stably improves accuracy of initial data on attitudes and directions regardless of errors of a magnetic direction sensor and disturbances such as winds in heavy weather. <P>SOLUTION: The inertia navigation system for determining the location and attitude of a head part of a passenger of a mobile body through the use of output of a gyrosensor and an accelerometer mounted to a helmet of the passenger comprises: a sensor error correction and computation part for correcting output of the gyrosensor and the accelerometer; a navigation/attitude direction computation part for computing the location and attitude of the head part through the use of output of the gyrosensor and the accelerometer corrected by the sensor error correction and computation part and correcting results of computations of the location and attitude of the head part; and an error estimation and computation part for calibrating sensor error data to be used at the sensor error correction and computation part and calibrating attitude/location error data. <P>COPYRIGHT: (C)2007,JPO&INPIT

Description

  The present invention relates to a sensor that cannot be obtained quantitatively in an inertial navigation device that calculates the orientation and orientation of a user's head using the output of a gyro sensor or an acceleration sensor mounted on the user's head. The present invention relates to an inertial navigation apparatus and an error correction method thereof that can correct errors and determine initial attitude / orientation data more accurately.

  FIG. 4 is a block diagram showing an example of a conventional inertial navigation apparatus and an error correction method thereof, and shows a configuration of a head tracker that is mounted on the head of an aircraft occupant. The head tracker device is a device that is mounted on a helmet of a passenger such as an aircraft and provides the posture and azimuth of the head of the helmet wearer.

  In FIG. 4, 11 is a sensor error correction calculation unit, 12 is a navigation / attitude direction calculation unit, 13 is an error estimation calculation unit, 14 is a reference attitude calculation unit, and 15 is a micro electro mechanical system (MEMS). A MEMS sensor unit 16 including an accelerometer and a gyro sensor used, 16 is an external reference azimuth sensor unit that detects a azimuth from observation of external geomagnetism by a magnetic azimuth sensor. One accelerometer and one gyro sensor are provided for each of the three axes of roll, pitch, and yaw.

The operation flow of the head tracker shown in FIG. 4 will be briefly described. An angular velocity and acceleration are detected by the gyro sensor and the accelerometer of the MEMS sensor unit 15 and output to the sensor error correction calculation unit 11. The sensor error correction calculation unit 11 corrects the input angular velocity and acceleration values in real time based on an error estimation value given in advance, and outputs the corrected value to the navigation / posture direction calculation unit 12. The navigation / posture calculation unit 12 calculates the posture and direction of the head using the corrected angular velocity and acceleration, and outputs them to the pilot and error estimation calculation unit 13. At the time of posture / azimuth calculation, a pre-given posture and azimuth error estimation value is used. The reference posture calculation unit 14 directly uses the acceleration data output from the MEMS sensor unit 15 to calculate the head posture, and outputs the head posture to the error estimation calculation unit 13. In addition, head direction data is output from the external reference direction sensor unit 16 to the error estimation calculation unit 13.
Since the error does not diverge, the outputs of the reference posture calculation unit 14 and the external reference direction sensor unit 16 can be used as reference data for measuring errors included in the calculated values of the head posture / direction data.
The error estimation calculation unit 13 compares the calculated values of the attitude / azimuth data input from the navigation / attitude / azimuth calculation unit 12 with various reference data, thereby calculating errors in the accelerometer and the gyro sensor, navigation / attitude / azimuth calculation. The magnitude of the error included in the posture / orientation data calculated by the unit 12 is estimated in real time. Then, the estimated value of the error is output to the sensor error correction calculation unit 11 and the navigation / posture direction calculation unit 12, and is used for the correction calculation as a new error estimation value.

FIG. 5 is a conceptual diagram of the head tracker. In the rescue helicopter searching for the victim, a pilot who operates the helicopter and a searcher who visually searches for the victim are on board. Searchers are wearing helmets equipped with head trackers. In FIG. 5, the vertical axis XN means north and the horizontal axis YE means east. Roll axis of X _p and Y _p is the helicopter body coordinates, a pitch axis, the X and Y roll axis of the head coordinates of the seeker of the helmet, the pitch axis. Ψ _p is the azimuth angle of the aircraft, and Ψ is the azimuth angle of the helmet.

The head tracker calculates the posture and orientation of the head of the wearer (searcher) and outputs information to the pilot. When the searcher finds a victim, the pilot guides the aircraft in the direction and orientation of the head tracker manually or automatically (autopilot) according to the information calculated by the head tracker. That is, by controlling the aircraft so that the azimuth angle ψ _{p of the} aircraft and the azimuth angle ψ of the helmet coincide, the aircraft can be guided in the direction in which the searcher faces the victim.

  Returning to FIG. 4, the operation of the head tracker will be further described. The MEMS sensor unit 15 is composed of three MEMS accelerometers and three MEMS gyro sensors, and detects the head movement three-axis (roll, pitch, yaw) direction acceleration and angular velocity around the three axes of the head tracker wearer. The digital data is output to the sensor error correction calculator 11 via a data bus such as RS232C.

Ａ_ｂ，Ｗ_ｂは３軸加速度ベクトルと角速度ベクトルであり、初期の静止時にはほぼゼロである。 Ａ_ｂｃ、Ｗ_ｂｃは誤差校正後の加速度および角速度であり、航法・姿勢方位計算部１２へ出力される。 The sensor error correction calculation unit 11 receives error estimation data for calibrating the error (bias stability error) between the accelerometer and the gyro sensor from the error estimation calculation unit 13, and performs correction calculation (calibration calculation In general, this is called calibration.

A _b and W _b are a triaxial acceleration vector and an angular velocity vector, and are substantially zero at the time of initial stationary. A _bc and W _bc are acceleration and angular velocity after error calibration, and are output to the navigation / attitude direction calculator 12.

  The navigation / attitude direction calculator 12 calculates the attitude and direction of the head of the head tracker wearer (searcher) from the corrected acceleration and angular velocity data, and outputs information to the pilot. Further, the posture and orientation data are also output to the error estimation calculation unit 13.

  The navigation / posture direction calculation unit 12 can suppress an increase in error by feeding back the estimated error value of the posture and azimuth angle data input from the error estimation calculation unit 13. Similarly, with respect to the error of the speed and position data, the estimated error value of the speed and the position is fed back from the error estimation calculation unit 13 to the navigation / posture direction calculation unit, and the increase of the error is suppressed.

  The reference posture calculation unit 14 calculates reference posture data from the acceleration data input from the MEMS sensor unit 15 and outputs the reference posture data to the error estimation calculation unit 13.

  The error estimation calculation unit 13 includes the posture and direction data calculated by the navigation / posture direction calculation unit 12, the reference direction data output from the external reference direction sensor unit 16, and the reference posture output from the reference posture calculation unit 14. The data is compared, and the sensor error, navigation error, attitude direction error, etc. of the accelerometer and gyro sensor are estimated by the Kalman filter or the least square method. The accuracy of the error estimate depends on how accurately these errors can be expressed in mathematical formulas.

θ、Ψは航法・姿勢方位計算部で計算された姿勢角と方位角である。ここで＊印項は真値であり、Δ印項は誤差を示す。Δθ_ｇとΔΨ_ｇはジャイロセンサ誤差と加速度計誤差による誤差である。また、添え字ｂが付されている項はジャイロセンサと加速度計のバイアス誤差による誤差で、時間とともに増大する誤差である。添え字ｍが付された項は運動（加速度、角速度）に影響される誤差で、初期の静止時は問題にならない。添え字Ｔが付された項は環境温度変化に影響される誤差である。 The principle of estimating errors in attitude and azimuth, accelerometer and gyro sensor errors from the reference data will be described below.

θ and Ψ are the attitude angle and the azimuth calculated by the navigation / attitude direction calculator. Here, the * mark is a true value and the Δ mark indicates an error. Δθ _g and ΔΨ _g are errors due to a gyro sensor error and an accelerometer error. The term with the subscript b is an error due to a bias error between the gyro sensor and the accelerometer, and is an error that increases with time. The term with the subscript m is an error affected by motion (acceleration, angular velocity) and does not cause a problem at the time of initial stationary. The term with the subscript T is an error that is affected by environmental temperature changes.

である。Δθ_ｒは加速度計誤差による姿勢角誤差、ΔΨ_ｒは磁気方位センサ誤差で場所（ｐ）にも影響される。 The reference posture angle and azimuth angle are

It is. Δθ _r is an attitude angle error due to an accelerometer error, and Δψ _r is a magnetic azimuth sensor error and is also influenced by the location (p).

となる。 Comparing equation (2) and equation (3) and taking the difference,

It becomes.

Here, Δθ _g (t) and ΔΨ _g (t) increase with time due to bias errors between the accelerometer and the gyro sensor. In other words, integration processing is performed for navigation calculation, attitude angle and azimuth calculation in the navigation / attitude direction calculator, so if there is a sensor bias error, the speed, position, attitude angle, and azimuth angle errors will pass over time. It increases with.

On the other hand, since the posture angle of the reference posture data of the reference posture calculation unit 14 is directly calculated by the ratio of the accelerometer output and the gravitational acceleration, the reference posture angle error Δθ _r (t) in the equation (4) is increased with time. Does not increase. Further, the reference azimuth angle error ΔΨ _r (t, p) in the equation (4) does not increase with time.

のように、Δθ_ｇ、ΔΨ_ｇを充分大きく取り出すことが出来る。なお、ａ>>ｂは、ａはｂより非常に大きいことを意味する。 Therefore, if the time interval to be compared is sufficiently large according to the operation accuracy,

As described above, Δθ _g and ΔΨ _g can be extracted sufficiently large. Note that a >> b means that a is much larger than b.

If the bias error of the breakdown of Δθ _g and ΔΨ _g , the error due to motion, the error due to temperature, and the like are modeled by mathematical formulas from Equation (2), Δθ _g and ΔΨ _g are estimated by the Kalman filter or the least square method. be able to. Further, since the speed error and the position error are closely related to Δθ _g and ΔΨ _g , it can be similarly estimated from the mathematical model of the speed error and the position error.

Patent No. 3490706

  However, in the error estimation of the calculated value of the azimuth data, the error of the magnetic azimuth sensor of the external reference azimuth sensor unit 16 becomes a correction limit, and the error estimation of the azimuth data greatly depends on this magnetic azimuth sensor.

  Since the reference attitude calculation unit calculates the reference attitude data from the ratio of the accelerometer output and the gravitational acceleration, the reference attitude data can generally be obtained without any problems in the initial stationary state before flight (before the start of movement) During emergency operation in bad weather, it is greatly affected by the movement of the wind.

  Since the external reference direction sensor unit is provided in order to obtain the reference direction data, the configuration of the entire device is complicated and the cost is increased. Further, since the geomagnetism varies depending on the location, the stability of the azimuth reference is poor, and the accuracy of error correction of the azimuth data is low.

  The present invention is intended to solve the problems of such conventional inertial navigation devices, and is more stable without being affected by disturbances such as magnetic direction sensor errors and winds during stormy weather. It is an object of the present invention to realize an inertial navigation apparatus and an error correction method thereof that can improve the accuracy of initial data of attitude / orientation.

In order to achieve the above object, according to claim 1 of the present invention, the position of the head of the helmet wearer and the gyro sensor attached to the helmet of the mobile occupant and the output of the accelerometer are used. In an inertial navigation system that determines posture,
A sensor error correction calculator that corrects the outputs of the gyro sensor and the accelerometer using sensor error data estimated in advance;
The head position and orientation are calculated using the output of the gyro sensor and the accelerometer corrected by the sensor error correction calculation unit, and the head position and orientation error data estimated in advance are used. A navigation / attitude direction calculator that corrects the position and orientation calculation results;
The velocity data output from the navigation / orientation calculation unit when the head is in a stationary state is used as a velocity error, the error in the output of the gyro sensor and the accelerometer is estimated, and the sensor error correction calculation unit An error estimation calculation unit that calibrates sensor error data to be used, estimates an error in output of the position and orientation, and calibrates attitude / position error data used in the navigation / attitude direction calculation unit;
It is characterized by having.

  According to a second aspect of the present invention, in the inertial navigation apparatus according to the first aspect of the present invention, the error estimation calculation unit calculates an error estimated value using a Kalman filter.

  According to a third aspect of the present invention, in the inertial navigation apparatus according to the first aspect of the present invention, the error estimation calculation unit calculates an error estimation value using a least square method.

In claim 4, in the error correction method of the inertial navigation apparatus for obtaining the position and posture of the head of the helmet wearer using the output of the accelerometer and the gyro sensor attached to the helmet of the mobile occupant,
A sensor error correction calculation step for correcting the output of the gyro sensor and the accelerometer using sensor error data estimated in advance;
The head position and orientation are calculated using the output of the gyro sensor and the accelerometer corrected in the sensor error correction calculation step, and the head position and orientation error data estimated in advance are used. Navigation / attitude direction calculation step to correct the position and orientation calculation results,
The velocity data output from the navigation / posture direction calculation unit when the head is in a stationary state is used as a velocity error, and an error in the outputs of the gyro sensor and the accelerometer is estimated, and the sensor error correction calculation step An error estimation calculation step for calibrating sensor error data to be used, estimating an error in the output of the position and orientation, and calibrating attitude / position error data used in the navigation / attitude orientation calculation unit;
It is characterized by having.

  According to a fifth aspect of the present invention, in the error correction method of the inertial navigation apparatus according to the fourth aspect, the error estimation calculation step calculates an error estimated value using a Kalman filter.

  According to a sixth aspect of the present invention, in the error correction method for an inertial navigation apparatus according to the fourth aspect of the present invention, the error estimation calculation step calculates an error estimated value using a least square method.

  If the initial attitude and heading data are accurately determined while the inertial navigation device is stationary when the power is turned on, it can contribute to efficient operation. As described above, by correcting the velocity data output from the navigation / orientation calculation unit as a velocity error when the inertial navigation device is in a stationary state, errors in the magnetic azimuth sensor or wind in stormy weather It is possible to realize an inertial navigation apparatus and its error correction method that can improve the accuracy of the initial data of posture / orientation more stably without being influenced by disturbances such as the above.

  By not using the magnetic azimuth sensor, the error estimation of the azimuth data is not affected by the error inherent in the magnetic azimuth sensor or the influence of external geomagnetism. Therefore, a more stable and highly accurate error correction of the gyro sensor and the accelerometer can be realized, and more stable and highly accurate attitude / orientation data can be provided to the pilot. Further, since the magnetic orientation sensor is not required, the configuration of the entire device is simplified, and the cost can be reduced.

  By not using the reference posture data calculated from the output of the accelerometer, stable and highly accurate posture data can be obtained even in the presence of disturbances such as during stormy weather.

  Hereinafter, an inertial navigation apparatus and an error correction method thereof according to the present invention will be described with reference to the drawings.

  FIG. 1 is a block diagram showing an embodiment of the inertial navigation apparatus and its error correction method according to the present invention. In the figure, 1 is a sensor error correction calculation unit, 2 is a navigation / attitude direction calculation unit, 3 is an error estimation calculation unit, and 4 is a MEMS sensor unit including an accelerometer and a gyro sensor using MEMS.

  The sensor error correction calculation unit 1 and the MEMS sensor unit 4 of the present invention example are the same as the conventional example shown in FIG. The configuration and operation of the navigation / posture direction calculation unit 2 and the error estimation calculation unit 3 will be described below.

First, the navigation / attitude direction calculator 2 will be described.

ここで、Ｖ_ｎは航法座標系で表現された速度ベクトルであり、北方向成分ｖ_Ｎ、東方向成分ｖ_Ｅ、鉛直方向成分ｖ_Ｄに分解される。
また、Ω_ｅｎは航法座標系での地球自転速度の北軸周り、東軸周り、鉛直軸周り、各成分からなる３行３列の行列である。Ω_ｐｎは航法座標系での地球の周りを移動する移動速度の北軸周り、東軸周り、鉛直軸周り、各成分からなる３行３列の行列である。Ｇ_ｎは重力加速度ベクトルの航法座標表現であり、Ａ_ｎは移動加速度ベクトルの航法座標表現で加速度計出力である。Ｒ_λ、Ｒ_Λは緯度方向と経度方向の地球半径である。ｈは飛行高度である。λ、Λは緯度と経度である。 The speed and position data of the helmet wearer is calculated by the following formula.

Here, V _n is a velocity vector expressed in the navigation coordinate system, and is decomposed into a north direction component v _N , an east direction component v _E , and a vertical direction component v _D.
Further, Ω _en is a 3 × 3 matrix composed of components around the north axis, around the east axis, around the vertical axis, and around the earth rotation speed in the navigation coordinate system. Ω _pn is a matrix of 3 rows and 3 columns consisting of components around the north axis, around the east axis, around the vertical axis of the moving speed of moving around the earth in the navigation coordinate system. G _n is the navigation coordinate representation of the gravitational acceleration vector, A _n is an accelerometer output in the navigation coordinate representation of the movement acceleration vector. R _λ and R _Λ are the earth radii in the latitude and longitude directions. h is the flight altitude. λ and Λ are latitude and longitude.

Ｋ_ｖｎ、Ｋ_λ、Ｋ_Λは速度誤差、緯度誤差、経度誤差を最小にするフィードバックゲインである。 U _vn , u _λ , and u _Λ are feedback control amounts obtained by the following equations using estimated values such as velocity error and position error obtained by error estimation calculation of the error estimation calculation unit 3.

K _vn , K _λ , and K _Λ are feedback gains that minimize the speed error, latitude error, and longitude error.

ここで、Ｃ_ｎｈは３行３列のヘルメット座標ｈから航法座標ｎへの座標変換行列である。

Ψ、θ、φは頭部の方位角、ピッチ角、ロール角である。

Ｋ_ｗｈは姿勢および方位データの誤差を最小にするフィードバックゲインである。 The posture and azimuth calculation of the helmet wearer can be calculated by the following formula.

Here, C _nh is a coordinate transformation matrix from helmet coordinates h to navigation coordinates n in 3 rows and 3 columns.

Ψ, θ, and φ are the head azimuth, pitch, and roll angles.

_Kwh is a feedback gain that minimizes errors in attitude and orientation data.

Next, the error estimation calculation unit 3 will be described.
How to estimate the gyro sensor error, the accelerometer error, the attitude angle error, the speed error, etc. from the speed data output from the navigation / attitude direction calculator 2 will be described with the following simple model. It's actually more complex but the idea is the same.
In the following, (XN, YE) is the navigation coordinate indicating the north axis and the east axis, and (X, Y) is the roll axis and the pitch axis of the head coordinates of the helmet. The accelerometer and gyro sensor are directly attached to X and Y of the head reference axis in a strap-down manner. Ψ is an azimuth angle.
FIG. 2 shows a coordinate system of navigation coordinates and head coordinates.

Ａ_Ｎは北方向の加速度であり、初期状態の静止時のため０である。
δａ_Ｎは加速度誤差である。結局、初期は静止時ゆえ計算された速度出力は速度誤差そのものであり、式（９）は、

のように表現できる。ここが、本発明のポイントである。 The north speed data output of the navigation calculation output of the navigation / attitude direction calculation unit 2 is expressed by the following equation.

A _N is acceleration in the north direction, and is 0 because the stationary state is the initial state.
δa _N is an acceleration error. After all, since the initial speed is stationary, the calculated speed output is the speed error itself, and equation (9) becomes

It can be expressed as This is the point of the present invention.

で表される。 Here, the northward acceleration error δa _N is

It is represented by

ここで、Δω_ｘ，Δω_ｙはＸジャイロセンサとＹジャイロセンサのバイアス誤差である。 Here, Δa _x and Δa _y are bias errors of the X accelerometer and the Y accelerometer, and δθ _E is an angle error around the east axis expressed by the following equation. Ψ ₀ , g is the initial azimuth angle and gravitational acceleration.

Here, Δω _x and Δω _y are bias errors of the X gyro sensor and the Y gyro sensor.

  The purpose of the error estimation calculation unit 3 is to estimate the accelerometer bias error, the gyro sensor bias error, the angle error, the speed error, and the like from the integral speed data input of Equation (10). A case where the error is obtained using the least square method will be described below. The error can be obtained in the same way using a Kalman filter.

となる。 The initial azimuth angle Ψ ₀ is a constant as long as there is no disturbance. Therefore, the time integration of equation (12) is

It becomes.

The integration of equation (10) is given by the following output data.

と書ける。ここでｈ(ｔ)は１行６列の観測行列で時間ｔの関数、Θが推定すべき未知パラメータで６行１列の行列である。 When Expression (14) is expressed as a matrix,

Can be written. Here, h (t) is a 1-by-6 observation matrix, a function of time t, and Θ is an unknown parameter to be estimated and is a 6-by-1 matrix.

ここで、ω_ｘ，ω_ｙはＸジャイロセンサとＹジャイロセンサの出力で、

である。Ωは地球自転速度で１５°／Ｈｒである。 The initial azimuth angle Ψ ₀ and the initial latitude λ ₀ are first estimated simply by the following equation.

Here, ω _x and ω _y are the outputs of the X gyro sensor and the Y gyro sensor,

It is. Ω is 15 ° / Hr in terms of earth rotation speed.

で与えられる。±の決定は北半球にいるか南半球にいるかで決まる。 λ ₀ is the initial latitude, from equation (17)

Given in. The decision of ± depends on whether you are in the northern or southern hemisphere.

Since Equation (15) has six unknown parameters, six or more equations are required to obtain all these unknown parameters. Therefore, the time _{t 1, t 2, t 3} , ..., to prepare the following equation from the output of each t _n.

で求まる。 Applying the least squares method to equation (19), the estimation of the unknown parameter Θ is

It is obtained by.

で求まる。

The estimation of angular error and velocity error is based on Equation (13) and Equation (14)

It is obtained by.

The accelerometer bias error and the estimated value of the gyro sensor bias error obtained as described above are output to the sensor error correction calculation unit 1.
In addition, the estimated values of the angle error and the speed error are output to the navigation / attitude direction calculator 2 and used for the speed error and attitude / azimuth angle error feedback control of the equations (7) and (8).

  Here, only the estimation of the angle error and the speed error has been described, but in reality, estimation of the speed error, the position error, the attitude / orientation error, etc. is executed based on the same concept. In the actual estimation of the present invention, a Kalman filter is applied in order to minimize the influence of noise, but even in that case, only the mathematical expression of the error is different.

  FIG. 3 is a diagram comparing the conventional example and the present invention with respect to errors in the direction data calculated by the navigation / posture direction calculation unit 2. FIG. 3A shows a conventional example, and FIG. 3B shows the result of the present invention. The horizontal axis is time, and the vertical axis is the direction error. In the conventional example of (a), since the magnetic direction sensor is used, the correction limit is determined by the error of the magnetic direction sensor, and the error cannot be made smaller than a certain level. Also, the magnitude of the error is not stable due to changes in geomagnetism. On the other hand, in the present invention of (b), it is possible to make the azimuth error as close to zero as possible by utilizing the fact that the reference speed is zero at rest.

  In the present invention, the sensor is directly attached to the moving body. However, the sensor may be attached to a horizontal plane (platform method) realized by mechanical servo control. In addition, when a disturbance such as wind cannot be ignored at rest, it can be applied by providing a filter for removing the disturbance. In this case, since the disturbance speed is superimposed on the speed data of the present invention, a disturbance removal filter is provided in the preceding stage of the error estimation calculation unit 3, and the speed data after disturbance removal is used for the error estimation calculation.

  In the present invention, the error of the gyro sensor or the accelerometer is corrected with the estimated value, and the navigation error such as the speed / position data and the error of the attitude / azimuth data are corrected by the feedback control based on the estimated value. This feedback control is mainly optimized control, but may be PID (Proportional Integral Differential) control.

FIG. 1 is a block diagram showing an embodiment of an inertial navigation apparatus and its error correction method according to the present invention. FIG. 2 is a diagram showing a coordinate system of navigation coordinates and head coordinates. FIG. 3 is a diagram comparing a conventional example and the present invention with respect to an error in azimuth data. FIG. 4 is a configuration diagram illustrating an example of a conventional inertial navigation apparatus and an error correction method thereof. FIG. 5 is a conceptual diagram of a head tracker mounted on the head of an aircraft passenger.

Explanation of symbols

DESCRIPTION OF SYMBOLS 1,11 Sensor error correction calculation part 2,12 Navigation / attitude direction calculation part 3,13 Error estimation calculation part 4,15 MEMS sensor part 14 Reference attitude calculation part 16 External reference direction sensor part

Claims (6)

In an inertial navigation device that uses the output of a gyro sensor and an accelerometer attached to the helmet of a mobile occupant to determine the position and posture of the head of the helmet wearer,
A sensor error correction calculator that corrects the outputs of the gyro sensor and the accelerometer using sensor error data estimated in advance;
The head position and orientation are calculated using the output of the gyro sensor and the accelerometer corrected by the sensor error correction calculation unit, and the head position and orientation error data estimated in advance are used. A navigation / attitude direction calculator that corrects the position and orientation calculation results;
The velocity data output from the navigation / orientation calculation unit when the head is in a stationary state is used as a velocity error, the error in the output of the gyro sensor and the accelerometer is estimated, and the sensor error correction calculation unit An error estimation calculation unit that calibrates sensor error data to be used, estimates an error in output of the position and orientation, and calibrates attitude / position error data used in the navigation / attitude direction calculation unit;
An inertial navigation device comprising:   The inertial navigation apparatus according to claim 1, wherein the error estimation calculation unit calculates an error estimation value using a Kalman filter.   The inertial navigation apparatus according to claim 1, wherein the error estimation calculation unit calculates an estimated value of an error using a least square method. In an error correction method for an inertial navigation device that uses the output of a gyro sensor and an accelerometer attached to a helmet of a mobile occupant to determine the position and posture of the head of the helmet wearer,
A sensor error correction calculation step for correcting the output of the gyro sensor and the accelerometer using sensor error data estimated in advance;
The head position and orientation are calculated using the output of the gyro sensor and the accelerometer corrected in the sensor error correction calculation step, and the head position and orientation error data estimated in advance are used. Navigation / attitude direction calculation step to correct the position and orientation calculation results,
The velocity data output from the navigation / posture direction calculation unit when the head is in a stationary state is used as a velocity error, and an error in the outputs of the gyro sensor and the accelerometer is estimated, and the sensor error correction calculation step An error estimation calculation step for calibrating sensor error data to be used, estimating an error in the output of the position and orientation, and calibrating attitude / position error data used in the navigation / attitude orientation calculation unit;
An error correction method for an inertial navigation apparatus, comprising:   5. The error correction method for an inertial navigation apparatus according to claim 4, wherein the error estimation calculation step calculates an error estimation value using a Kalman filter. 5. The error correction method for an inertial navigation apparatus according to claim 4, wherein the error estimation calculation step calculates an error estimated value using a least square method.

Images