JPH0821740A - Inertial navigation apparatus for ship - Google Patents

Abstract

PURPOSE:To shorten the operating time of correcting means by outputting an azimuth angle error from the calculating step at each repeated calculation of the means, and so adding a small correction value to an initial latitude as to gradually converge the error. CONSTITUTION:A speed position calculator 5 fetches the output signals of an accelerometer 11 and a gyro 12 fixed to a ship hull, and sequentially outputs position information and speed information with the input initial latitude and an initial latitude as references. A speed and position comparator 7 compares both the pieces of the information with position and speed information input from the outside, and outputs error information. An error designating unit 8 and a correction calculator 9 fetch the error information, and automatically correct calculating parameters by so repeatedly calculating as to minimize the generated error. Azimuth angle error is fetched from the calculating step at each repeated calculation, and a small correction value is so added to the initial latitude as to gradually converge the azimuth angle error. Thus, the time required for automatically correcting the parameters can be shortened.

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION The present invention is used to accurately measure the position (and speed) of a ship during voyage. The present invention relates to error correction performed when an inertial navigation system is activated.

[0002]

2. Description of the Related Art An inertial navigation system is known as a device for extremely accurately measuring the position of a ship under voyage.
This device takes in the output signals of a three-dimensional accelerometer and a three-dimensional gyro, which are fixed to the hull as hardware, and calculates position information and speed information during navigation based on the initial latitude and initial longitude. It is an arithmetic unit including output software. That is, the initial latitude and the initial longitude when starting this device at the pier of a berth where the latitude and longitude have already been measured accurately, can be read on this device by operating the numeric keypad or by reading a preset magnetic card. When the input is made by, for example, the acceleration received from the hull is integrated over time to calculate speed information, and the speed information is further integrated over time to calculate position information.

Since such an inertial navigation system theoretically obtains position information and velocity information by cumulative calculation, it is inevitable that errors will accumulate if the device is used for a long time. Therefore, as a general rule, when a ship equipped with an inertial navigation system arrives at its destination, it is necessary to reset the already entered initial latitude and initial longitude and newly enter the accurate initial latitude and initial longitude of the berth. is necessary.

A device based on this principle has become widely known as an inertial navigation device for an aircraft, and has become an indispensable device for an autopilot device of an aircraft. Many improvements have been made to make this device usable for ships. One of them is the applicant's prior application (Japanese Patent Application No. 5-52450,
It has not been published at the time of filing this application).

That is, when the aircraft is stationary at a predetermined position at the airport, the aircraft may be in an almost accurate stationary state with respect to the earth. However, even when the ship is moored at the base harbor, the ship is always in a small state of pitching and rolling, and is not in a stationary state. Even if the ship is moored by ropes, its bow direction is slightly swayed. Also, its position relative to the earth is constantly moving, even within a few meters. In other words, the inertial navigation system installed on the ship is always disturbed. Therefore,
With this type of equipment on board a vessel, even if the official reference is entered as initial latitude and initial longitude information while moored in a port where the position is publicly and accurately measured, the disturbance is actually It is unavoidable that there was a small error in the value that was input in the state with.

For this reason, the inertial navigation system for a ship is equipped with a correction means in the calculation means. That is, the correction means includes a comparison means for comparing the position information and the speed information output from the inertial navigation device with the position information and the speed information input from the outside to output error information,
The calculation means of the inertial navigation device automatically corrects the calculation parameter by taking in this error information and repeating the calculation so that the generated error is minimized. Here, the position information and the velocity information input from the outside are information obtained by another device. For example, the latitude and longitude information measured by GPS is one of them, the survey information based on the ground target is one of them, and the ground speed information during voyage is one of them. One of the important information is that the speed should be zero because it is moored.

For example, if the speed of the ship at berth should be zero, but if the output speed information of the inertial navigation device shows a finite value exceeding a certain limit, it is a calculation parameter inside the calculation means. Means that is not correct and the correction must be made. As a general rule, this automatic correction of the calculation parameters is executed by setting a correction mode different from the normal operation mode while the ship is at berth.

FIG. 5 is a block diagram showing an example of a conventional inertial navigation system. This inertial navigation system includes a sensor unit 1
And an accelerometer error correction calculation unit 21, a gyro error correction calculation unit 22, a coordinate conversion unit 3, a coordinate conversion matrix calculation unit 4, a correction calculation unit 9, a speed position calculation unit 5, a posture direction calculation unit 6, and a speed position comparison unit. 7 and an error estimator 8 including a Kalman filter.

Before describing the operation of this conventional example, the coordinates used in this apparatus will be described with reference to FIG. FIG. 6 is a diagram showing coordinates used in the inertial navigation system. The coordinates used in the inertial navigation system include hull coordinates and navigation coordinates (or navigation calculation coordinates). Hull coordinates, roll, by each axial component of the pitch and yaw, the coordinate represented as _{(b x, b y, b} z), the hull center as the origin. On the other hand, the navigation coordinates are also called n coordinates, and are coordinates represented by (x _N , y _E , z _D ) by the respective components in the north, east and height directions. The navigation coordinates also have the center of the hull as the origin, but the coordinates of points on the ship move as the ship moves. The rotation angle around the roll axis is called the roll angle φ, the rotation angle around the pitch axis is called the pitch angle θ, and the direction of the ship is called the azimuth angle ψ. Azimuth ψ is ψ = 0 when facing north, ψ when facing east
= 90 °.

Next, the operation of each part of the conventional example will be described using the symbols shown in Tables 1 and 2.

[0011]

[Table 1]

[0012]

[Table 2] The sensor unit 1 includes an accelerometer 11 and a gyro 12, the accelerometer 11 measures the motion acceleration of the ship and outputs an acceleration vector, and the gyro 12 measures the motion angular velocity of the ship and outputs a rotation angular velocity vector. . In the strapdown system, the sensor unit 1 is directly attached to the hull reference shaft. The mounting position is set as close to the center of gravity of the ship as possible.

The acceleration vector output from the accelerometer 11 is represented by ship hull coordinates having the accelerations A _bx , A _bx and A _bz in each direction of the roll axis, pitch axis and yaw axis as elements, and the propulsion acceleration vector of the ship, There is the following relationship with the acceleration error vector and the shaking acceleration vector.

[0014]

[Equation 1] The shaking acceleration vector of the third term on the right side becomes a disturbance acceleration during the initial alignment, and has a great influence on the posture determination time and the determination accuracy. Initial alignment refers to determining the initial attitude and azimuth as accurately as possible using the horizontal component of the accelerometer output. At the time of this initial alignment, some of the errors of the gyro and the accelerometer are also corrected. This is called calibration.

The rotational angular velocity vector output from the gyro is also expressed in ship coordinates, and has hull shaft roll angular speed W _bx , hull shaft pitch angular speed W _by , and hull shaft yaw angular speed W _bz as elements. The rotational angular velocity vector has the following relationship with the turning angular velocity vector, the gyro error vector, and the fluctuation angular velocity vector.

[0016]

[Equation 2] The outputs of the accelerometer 11 and the gyro 12 are input to the accelerometer error correction calculation unit 21 and the gyro error correction calculation unit 22, respectively. These calculation units 21 and 22 correct a part of each error by the following correction formula.

[0017]

(Equation 3) In these two mathematical expressions, the first term on the right side is the propulsion acceleration vector and the turning angle vector of the ship, and is represented by the following expressions.

[0018]

[Equation 4] The second term on the right-hand side is the motion acceleration vector and angular velocity vector of the ship, and is expressed by the following equation.

[0019]

(Equation 5) Then, the third term on the right side is the accelerometer error vector and the gyro error vector, and is represented by the following equation.

[0020]

(Equation 6) The last term on the right side is the accelerometer error estimation output and the gyro error estimation output output from the error estimation unit 8, and is a vector represented by the following equation.

[0021]

(Equation 7) The coordinate conversion unit 3 converts the output of the accelerometer error correction calculation unit 21 from the hull coordinates to the n coordinates which are the navigation coordinates according to the following equation.

[0022]

(Equation 8) The coordinate conversion matrix calculation unit 4 calculates the coordinate conversion matrix used for coordinate conversion in the coordinate conversion unit 3 by the following formula.

[0023]

[Equation 9] The coordinate transformation matrix is expressed as follows.

[0024]

[Equation 10] Also, in the formula of [Equation 9], d / dt represents a differential,
The symbol shown by the overline on the right side is a distortion symmetric matrix representation of the relative rotational angular velocity vector with respect to the h-coordinate n coordinate. The relative rotation angular velocity vector is obtained by the following formula.

[0025]

[Equation 11] Therefore, its distortion symmetric matrix expression is as follows.

[0026]

(Equation 12) In the first expression of [Equation 11], the terms represented by the first four symbols on the right side are the output of the gyro. As is well known, the gyro measures not only the angular velocity of the ship (turning angular velocity + swaying angular velocity) but also all angular velocities with respect to absolute stationary coordinates. Specifically, the rotational angular velocity of the earth, the angular velocity of movement of the ship around the round earth, that is, the rate of change in latitude and longitude are also measured. But,
Since the navigation calculation and the attitude and azimuth of the ship are expressed by the motion with respect to the earth and the attitude angle and azimuth, it is necessary to correct the gyro output by subtracting the rotation angular velocity and the moving angular velocity of the earth. What has been corrected is [Equation 11]
Is the first expression of. In this equation, the final term is the correction amount of the attitude angle error and the azimuth angle error due to the gyro error and the accelerometer error, and is represented by the following equation.

[0027]

(Equation 13) The right side is calculated by the correction calculator 9. The subtraction term is a quantity related to the rotation angular velocity of the earth and the movement angular velocity of the ship, and is calculated by the speed position calculation unit 5 as follows.

[0028]

[Equation 14] The velocity position calculation unit 5 also obtains the velocity and latitude / longitude in the equation shown in [Equation 14] by the following calculation based on the output of the coordinate conversion unit 3.

[0029]

(Equation 15) In the case of a ship, the altitude direction is irrelevant, but since there is a correlation with the horizontal direction, only the calculation is performed as follows.

[0030]

[Equation 16] The final term in each of the two equations of [Equation 16] is a height direction error correction amount for minimizing the influence of the height direction error on the horizontal direction, and is calculated by the correction calculation unit 9 as follows.

[0031]

[Equation 17] The correction calculator 9 also calculates the correction amount of the horizontal velocity error, that is, the final term in the first expression of [Equation 15].

By such calculation, the velocity position calculation unit 5
In the output of, the velocity vector represented by the second equation of [Equation 15] and the position vector represented by the third and fourth equations are obtained. The position vector is represented as follows.

[0033]

(Equation 18) The obtained velocity vector and position vector are output.

In [Equation 15], the n-coordinate expression of the gravity acceleration vector represented by the following equation is used.

[0035]

[Formula 19] Further, the earth radius in the latitude direction and the longitude direction used in [Equation 14] is expressed as follows using the earth radius a at the equatorial plane.

[0036]

(Equation 20) Further, in [Equation 15], values expressed in a matrix by the following equation are used.

[0037]

[Equation 21] On the other hand, the posture azimuth calculation unit 6 calculates the posture angle (φ, θ) and the azimuth angle ψ from the coordinate conversion matrix output from the coordinate conversion matrix calculation unit 4, and outputs the input / output data display device via the input / output circuit. Output. The attitude angle and the azimuth angle are calculated as follows using the elements of the coordinate conversion matrix.

[0038]

[Equation 22] The velocity position and position vector obtained by the velocity position calculation unit 5 are not only output, but also the velocity position comparison unit 7
Is input to Further, the external reference speed and the external reference position are input to the speed position comparison unit 7, respectively. The velocity / position comparison unit 7 compares these and the difference is the error estimation unit 8
To enter.

During the initial alignment, the external reference speed and the external reference position represented by the following expressions are input.

[0040]

(Equation 23) Further, during the initial alignment, the first equation of [Equation 15] is the following equation.

[0041]

[Equation 24] The correction amount of the horizontal velocity error in [Equation 24] is calculated by the correction calculation unit 9 according to the following equation.

[0042]

(Equation 25) That is, two elements of the correction amount vector are obtained by the right side of the second equation of [Equation 25], and another element obtained by [Equation 17] is combined to obtain the correction amount of the horizontal velocity error.

In the U expression of [Equation 14] and the second expression of [Equation 25],

[0044]

(Equation 26) And

[0045]

[Equation 27] Is.

[Equation 25] The coefficient of the second equation is 2 rows 14
It is an optimal control gain represented by a matrix of columns and is determined so as to minimize the following evaluation function.

[0047]

[Equation 28] Minimizing this J means the interval [0 ≦ t ≦ T].
This means minimizing the correction amount used in and the cumulative error at that time. That is, it is energy minimization control.

The two matrices shown in [Equation 27] are the estimated values of the velocity error, the attitude angle error, the azimuth angle error and their accumulated values, and are supplied from the error estimation unit 8 to the correction calculation unit 9. These two matrices have the following relationship.

[0049]

[Equation 29] The correction calculation unit 9 also obtains the correction amount of the attitude angle error and the azimuth angle error by the following formula.

[0050]

[Equation 30] The third symbol on the right side is the same as that shown in [Equation 26], and the second symbol represents the optimum control gain determined so as to minimize the evaluation function of [Equation 28]. The optimum control gain in this case is represented by a matrix of 3 rows and 14 columns.

The error estimating section 8 estimates the velocity error, the position error, the attitude angle error, and the azimuth angle error, and estimates the gyro error and the acceleration error. The model for this estimation is given by

[0052]

[Equation 31] The error estimation unit 8 further models and estimates the external reference speed error and the tidal current speed as follows.

[0053]

[Equation 32] The operation of the error estimator 8 is given by the following equation.

[0054]

[Expression 33] The observation data input to the error estimation unit 8 differs between the initial alignment and the alignment during navigation. In addition, even during alignment during navigation, it will be different when observing the external reference speed, observing the external reference position, or using both the external reference speed observation and the external reference position observation.

The input observation data at the time of external reference velocity observation is as follows.

[0056]

(Equation 34) Therefore, the following equation is obtained.

[0057]

[Equation 35] When observing the external reference position, it becomes as follows.

[0058]

[Equation 36] Therefore, the following equation is obtained.

[0059]

(37) Since the observation data is only velocity data during the initial alignment, external reference position observation is not performed.

Input observation data when the external reference velocity observation and the external reference position observation are used together is as follows.

[0061]

(38) This combined observation is not performed during the initial alignment. That is, only velocity observation is performed at the time of initial alignment.

[0062]

An inertial navigation system for a ship is provided with automatic correction means provided in the calculation means of the inertial navigation system as described above, and the calculation parameters set by executing the correction mode are sequentially corrected. However, it is possible to correct the speed information so that there is no contradiction such that the speed information always indicates zero in a predetermined range while the ship is at berth. However, in such a conventional automatic correction means, when the correction mode is executed, it takes several hours to converge the output of the calculation means to a certain allowable value and complete the correction operation. A practical device requires, for example, 4 hours. The vessel must be maintained at zero speed during this correction mode. This is not the problem of the calculation execution speed of the calculation means, but the fact that the ship sways in a cycle of several seconds to several tens of seconds. Even if the hardware of the calculation means is replaced with a high-speed one, this correction Mode execution time is not reduced.

This is because when the inertial navigation system is in operation, its accumulated error has become large, so when the system was reset once and a new initial latitude and initial longitude were set and input,
It means that you can leave the port only after about 4 hours. Ships equipped with conventional devices are actually operated under such a system. In other words, start the inertial navigation device 4 hours or more before the scheduled departure time, input the latitude and longitude information of the anchored port to the inertial navigation device as the initial latitude and initial longitude, and execute the correction mode until immediately before departure. It is operated as follows.

If the correction operation is interrupted for some reason, the correction operation for several hours must be restarted from the beginning. For example, if an operation error is found, such as an error in the initial latitude or initial longitude that was initially input after the passage of time, it will not be possible to leave the port at the scheduled time, and the departure time will be sent over several hours. There may be situations in which the navigation system must be used or inertial navigation equipment cannot be used for the voyage.

On the contrary, when a ship temporarily berths at a berth, even if an attempt is made to reset the accumulated error of the inertial navigation system equipped on the ship, based on accurate latitude and longitude information obtained at the berth. , If the sufficient time for executing the correction mode at the berth cannot be taken, the inertial navigation device will be continuously used without resetting the accumulated error, and the inertial navigation device is originally In some cases, it is not possible to fully utilize the existing performance.

That is, an important function in the initial alignment or calibration is to determine the initial attitude angle and the azimuth angle with high accuracy in a short time. For that purpose, accurate information about latitude is indispensable among the initial input information from the outside. Conventionally, in order to satisfy this purpose, for example, GPS (Global Psitioning Syste
m) and known latitude information of the docking point. However, GPS, which is originally a military purpose, has an error that cannot be corrected unless a special means is intentionally used, the exact position of the berth is unknown, and in reality these information are used. Often not. If accurate latitude information cannot be obtained, problems such as a long initial attitude angle and azimuth angle determination time and an increase in azimuth angle settling error occur. This is because the characteristic equation P (s) of the system at the time of determining the initial attitude angle and the initial azimuth angle depends on the latitude information at that point.

Hereinafter, it will be briefly described that the determination of the initial attitude angle and the azimuth angle depends on the latitude information. Simplifying the polynomial for determining the initial attitude angle and azimuth,

[0068]

[Formula 39]

[0069]

(Equation 40) Becomes Here, δθ is a pitch angle error, δψ is an azimuth angle error, and δλ is a latitude error.
0] is an approximate solution

[0070]

[Formula 41] Becomes Here, s is a Laplace operator. [Outer 1] and [Outer 2] are gains for determining the initial attitude angle and azimuth, and Ω is the earth rotation rate. [Outside 3], [Outside 4]
Is the gyro drift error. The steady-state error δψss is calculated from [Formula 41],

[0071]

(Equation 42) Becomes Here, Ω> 0, cos λ ₀ ≧ 0, (−90 °
≦ λ ₀ ≦ 90 °). That is, the settling error δψss largely depends on the inaccuracy δλ ₀ of the initial latitude input.

[0072]

[Outside 1]

[0073]

[Outside 2]

[0074]

[Outside 3]

[0075]

[Outside 4] Also, the time solution of [Equation 41] is generally

[0076]

[Equation 43] Becomes C, A, and ψ are constants determined by the coefficient parameter of [Formula 41].

The settling time (convergence speed) of the solution is determined by τω _n in [Equation 43]. From equation 41, the characteristic equation P (s) is

[0078]

[Equation 44] Becomes The cos λ ₀ of [Equation 44] or [Equation 4]
The inaccuracy of the initial latitude information also enters cos λ ₀ of 1]. When [Expression 44] is expressed using τ and ω _n of [Expression 43],

[0079]

[Equation 45] Becomes Therefore

[0080]

[Equation 46] Is obtained. Therefore, the effect of δλ ₀ is

[0081]

[Equation 47] Therefore, the transient response (convergence speed) also depends on δλ ₀ . Therefore, the actual ω _n is

[0082]

[Equation 48] Becomes [Outer 5] is true λ ₀ . FIG. 7 shows a settling curve of the azimuth error due to δλ ₀ according to the relationship between [Equation 42] and [Equation 48]. FIG. 7 is a diagram showing a settling curve of a conventional example. The horizontal axis represents time, and the vertical axis represents azimuth error. Figure 7
(A) shows a state in which there is no latitude error, that is, a state of δλ ₀ = 0. In this case, the azimuth angle error δψ converges within the allowable range [δψ] _{lim by the} time limit T _lim .

[0083]

[Outside 5] FIG. 7B shows a state where the correction direction of the latitude error is in the negative direction, that is, a state of δλ ₀ = −δλ ₁ <0. The azimuth error δψ has settled by the time limit T _lim but has not reached the allowable range.

FIG. 7C shows a state in which the correction direction of the latitude error is toward the plus direction, that is, the state of δλ ₀ = δλ ₂ > 0. The azimuth error δψ has not reached the allowable range by the time limit T _lim .

That is, the influence of δλ ₀ deteriorates the static characteristics even if the dynamic characteristics are improved, and conversely, the dynamic characteristics become extremely deteriorated even if the static characteristics are improved. This indicates deterioration of convergence.

The present invention has been made against such a background, and when the inertial navigation system is reset and new initial latitude and initial longitude are input, the correction means necessary for automatically correcting the calculation parameters is provided. The purpose is to reduce the operation time. The present invention is required until the inertial navigation device can be used after the inertial navigation device is reset and restarted, even if the ship at berth has a relatively large cycle and a wobbling with a complicated amplitude. The purpose is to reduce the time. It is an object of the present invention to reduce the preparation time required for a ship equipped with an inertial navigation system to leave a port. The present invention aims to improve the practical accuracy of an inertial navigation system. SUMMARY OF THE INVENTION It is an object of the present invention to provide an inertial navigation system which can reset accumulated errors frequently and can be operated in a state where accumulated errors are small. An object of the present invention is to provide an inertial navigation system that can allow the generation of cumulative error in a range larger than that of conventional devices.

[0087]

DISCLOSURE OF THE INVENTION The present invention is based on the discovery that if there is an error in the input initial latitude, the correction operation will take a long time. That is, when the inertial navigation device is activated, the latitude and longitude of the anchorage are input as the initial latitude and the initial longitude, but the time required for the correction operation is greatly affected by the error of the initial latitude. This is because the calculation means of the inertial navigation system is fixed to the hull.
When the output signals of the accelerometer in the three directions and the gyro in the three directions are taken and the horizontal reference calculation (leveling) and the heading calculation (gyro compassing) are executed, the gyro input is the angular velocity of the earth as a function of latitude. This is due to the fact that

According to the present invention, the azimuth angle error (δψ (ti), δψ (ti + 1), from the calculation process for each iterative calculation sequentially executed from the time ti by the correction means in the correction mode.
δψ (ti + 2) ...) are sequentially taken out, and the initial latitude (δ (t ₀ )) is gradually adjusted so that the azimuth error converges.
Small correction value (δλ · δ, including positive or negative sign)
It is characterized by including an initial latitude input estimation calculation means for adding.

That is, according to the present invention, the output signals of the accelerometer and the gyro fixed to the hull are taken in, and the position information and the velocity information are sequentially output with reference to the initial latitude (λ (t ₀ )) and the initial longitude input by the operation. And a comparing means for comparing the position information and the speed information output from the calculating means with the position information and the speed information input from the outside to output error information, respectively. An inertial navigation system for a ship, which includes correction means for automatically correcting calculation parameters by repetitive calculations so as to minimize an error that occurs when taking in.

Here, the feature of the present invention is that, for each iterative calculation (executed sequentially from time ti) of the correction means, the azimuth error (δψ (ti),
δψ (ti + 1), δψ (ti + 2) ...) are taken out,
Where the initial latitude input estimation calculation means for adding a small correction value (δλ · δ, including a positive or negative sign) to the initial latitude (λ (t ₀ )) so that the azimuth angle error gradually converges It is in.

The initial latitude input estimation calculation means calculates an average value of the azimuth angle error information for n consecutive operations of the iterative calculation of the correction means to obtain a primary average value, and the primary average value thereof. It is preferable to include means for calculating an average value of the values m times to obtain a secondary average value, and identifying a convergence of the azimuth angle error information based on the secondary average value. As a result, it is possible to eliminate the influence of ship sway.

The initial latitude input estimation calculation means calculates the secondary average value each time the calculation means completes the calculation, and the secondary average value and the preceding secondary average value are calculated. It is desirable to stop the calculation assuming that the azimuth angle error information has converged when the difference in absolute value of is within a preset error range (ε). As a result, it is possible to determine that the desired specifications are satisfied, so that the calculation can be stopped.

The initial latitude input estimation calculation means calculates the second average value every time the calculation means completes the calculation, and the second average value is corrected in spite of the predetermined Q corrections of the initial latitude. When the difference between the absolute values of the secondary average value and the preceding secondary average value does not fall within the preset error range (ε), it is desirable to stop the calculation. This makes it possible to determine that it is no longer possible to satisfy the desired specifications and stop the operation.

The correction value (δλ · δ) is a fixed value δλ (degrees) set in advance to a value of about several thousandths of a degree,
A positive / negative sign set in a direction to converge to a value obtained by dividing the difference between the secondary average value executed at that time and the secondary average value executed at the previous time by the error range (ε). It is desirable to be the product of the parameter δ (anonymous number) with.

When it is judged that the direction of the correction direction is opposite, the direction can be corrected by changing the sign of δ.

[0096]

The initial latitude (λ (t ₀ )) input by taking in the output signals of the accelerometer and gyro fixed to the hull
And the position information and the speed information are sequentially output based on the initial longitude, and the position information and the speed information are compared with the position information and the speed information input from the outside, respectively, and the error information is output. The calculation parameters are automatically corrected by iterative calculation so that the error generated by taking in this error information is minimized.

At this time, azimuth angle errors (δψ (ti), δψ (ti + 1), δ
ψ (ti + 2) ...), and adds a small correction value (δλ · δ, including positive or negative sign) to the initial latitude (λ (t ₀ )) so that the azimuth angle error gradually converges. To do.

As a result, the initial latitude can approach the actual latitude within the range satisfying the desired specifications.

The average value of the azimuth angle error information is calculated as the primary average value for n consecutive operations of repetitive calculations, and the average value of the primary average values is calculated m times and the second average is calculated. A secondary average value may be used, and it may be identified that the azimuth angle error information converges based on the secondary average value. As a result, it is possible to remove the influence of the fleet of the hull as compared to the case where the calculation is performed using only the first average value.

The secondary average value is calculated each time one arithmetic operation is completed, and the difference in absolute value between the secondary average value and the preceding secondary average value is set to a preset error range (ε ), It is preferable to stop the calculation assuming that the azimuth angle error information has converged. As a result, it can be determined that the initial latitude has approached the actual latitude within a range satisfying the desired specifications.

The secondary average value is calculated each time the calculation is completed, and the secondary average value and the previous secondary average value are calculated despite the correction of the initial latitude by a predetermined Q times. When the difference of does not fall within the preset error range (ε), it is preferable to stop the calculation. As a result, it is possible to avoid wasting calculation time. The cause may be that the correction direction is opposite.

The correction value (δλ · δ) is a fixed value δλ (degree) preset to a value of about several thousandths, a secondary average value executed at that time, and the previous time. The product of the difference between the absolute value and the second-order average value, which is executed in the above, and the parameter δ (anonymous number) with positive and negative signs set in the direction of convergence to the value obtained by dividing by the error range (ε). Good to have. As described above, since the correction direction is opposite, if the initial latitude is determined to be difficult to correct within a predetermined time, the correction can be performed again by changing the sign of the parameter δ.

[0103]

DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS The configuration of an embodiment of the present invention will be described with reference to FIG. FIG. 1 is a block diagram of an apparatus according to the present invention.

The present invention is based on the accelerometer 11 fixed to the hull.
Also, the speed position calculation unit 5 as a calculation means for sequentially outputting position information and speed information with reference to the initial latitude (λ (t ₀ )) and the initial longitude that are input by taking in the output signal of the gyro 12 and this speed, The speed position calculation unit 5 includes a speed position comparison unit 7 as a comparison unit that compares the position information and the speed information output from the position calculation unit 5 with the position information and the speed information input from the outside and outputs error information. Reference numeral 5 denotes a ship inertial navigation system including an error estimation unit 8 and a correction calculation unit 9 as correction means for automatically correcting calculation parameters by repetitive calculation so as to minimize the error generated by taking in this error information. is there.

Here, the feature of the present invention is that the azimuth angle error (δψ (ti)) is calculated from the calculation process for each iterative calculation (executed sequentially from time ti) of the error estimation unit 8 and the correction calculation unit 9. , Δψ (ti + 1), δψ (ti +
2) ......) is taken out, and a small correction value (δλ · δ, including a positive or negative sign) is added to the initial latitude (λ (t ₀ )) so that the azimuth error gradually converges. There is provided an initial latitude input estimation calculation unit 10 as a latitude input estimation calculation means.

The initial latitude input estimation / calculation section 10 calculates the average value of the azimuth angle error information for n consecutive operations of the iterative calculation of the error estimation section 8 and the correction calculation section 9 to obtain a primary average value. , Further about the primary average value m
Means for identifying the convergence of the azimuth angle error information based on the secondary average value by calculating the average value of the times to obtain the secondary average value.

The initial latitude input estimation / calculation section 10 calculates the second average value each time the velocity position calculation section 5 completes the calculation operation, and the second average value and the previous second average value are calculated. When the difference between the absolute value and the next average value falls within a preset error range (ε), the calculation is stopped assuming that the azimuth angle error information has converged.

Furthermore, the initial latitude input estimation calculation unit 10
The second average value is calculated each time the speed position calculation unit 5 completes the calculation, and the second average value and the previous second average value are calculated despite the correction of the initial latitude by a predetermined Q times. When the difference between the absolute value and the average value does not fall within the preset error range (ε), the calculation is stopped.

The correction value (δλ · δ) is a fixed value δλ (degrees) set in advance to a value of several thousandths of a degree,
The difference between the absolute values of the secondary average value executed that time and the secondary average value executed the previous time was divided by the error range (ε), and the difference was set to converge. It is the product of the parameter δ (anonymous number) with positive and negative signs.

The initial latitude input estimation / calculation unit 10 corrects the latitude information λ (t ₀ ) at least Q times to obtain λ _q (t ₀ ) = λ _q-1 (t ₀ ) + Δλ · δ _q ( However, q = 1, 2, ... Q) δ _q = [-J _q-1 (t _i )] /
Let ε ¹ lim be about this J _q

[0111]

[Equation 49] (However, [outer 6] is calculated from the previous average value [outer 7],
[Outside 8] is based on this average value [Outside 9]

[0112]

[Equation 50] The value is determined by _Jq (t _i ) <0 and | J _q (t _i ) | <ε ¹ lim.

[0113]

[Outside 6]

[0114]

[Outside 7]

[0115]

[Outside 8]

[0116]

[Outside 9] Next, the operation of the embodiment of the present invention will be described with reference to FIGS. FIG. 2 is a flow chart showing the operation of the embodiment of the present invention. FIG. 3 is a diagram for explaining the operation principle of the embodiment of the present invention. It is already known that the inaccuracy of the initial latitude input λ (t ₀ ) = λ ₀ δλ (t ₀ ) = δλ ₀ has a great influence on the determination of the initial attitude angle and the azimuth angle. Hereinafter, focusing on the specified value [outside 10] of the bearing error for the inaccuracy of the initial latitude input, it is corrected to a more accurate value by the following method to explain the improvement of the responsiveness and the settling error [outside 11]. To do.

Let the initial initial latitude input be λ (t ₀ ) = λ ₀ . Based on this λ ₀ , leveling and gyrocompassing for determining the initial attitude angle and azimuth angle are executed (S
1) The following operations are executed within tens of seconds after the start.
[Outer 10] The following equation is calculated using the information (S2). δψ
Because it is greatly affected by noise and upset [outside 12]
To use. When the average of N data of [outer 12] is calculated,

[0118]

(Equation 51) Then, when [outer 13] is calculated from [Equation 51] at time t _i ,

[0119]

[Equation 52] (S3). The calculation process and the motion of the hull will be described with reference to FIG. In FIG. 3A, λ (t ₀ ) = λ ₀ , and if this is corrected by an appropriate amount δλ · δ ₁ ,

[0120]

[Equation 53] Using this λ ₁ (t ₀ ), the initial attitude angle and azimuth angle are determined again from t ₀ to time t _i (S4). However, δ ₁ = 1.

When the same calculation is executed using [outer 10] (S5), as shown in FIG.

[0122]

[Equation 54] (S6). When the δλ correction reference formula is calculated from this [outer 14] and the previous [outer 13],

[0123]

[Equation 55] (S7). If this J ₁ (t _i ) satisfies J ₁ (t _i ) <0 and | J ₁ (t _i ) | <ε ¹ lim, the correction of δλ ends as a specification pass (S8). This ε ¹ lim is a predetermined specification value. If J ₁ (t _i ) <0 and | J ₁ (t _i ) |> ε ¹ lim (S12), the correction of δλ is still insufficient, and if q ≧ Q (S11). , S4 and so on again. At this time, if J ₁ (t _i )> 0 (S12), it means that the correction direction is opposite, so the sign of δ is converted (S13), and if q ≧ Q, correction is made.

[0124]

[Equation 56] And redo (S4). Using this λ ₂ (t ₀ ), t
Performing a determination of the initial attitude angle and azimuth angle from ₀ to t _i (S10). And, like the last time, [Formula 51]

[0125]

[Equation 57]

[0126]

[Equation 58] Is calculated (S5 to S7), and if J ₂ (t _i ) <0 and | J ₂ (t _i ) | <ε ¹ lim are satisfied, the latitude error δλ is corrected (S9).

[0127]

[Outside 10]

[0128]

[Outside 11]

[0129]

[Outside 12]

[0130]

[Outside 13]

[0131]

[Outside 14]

[0132]

[Outside 15]

[0133]

[Outside 16]

[0134]

[Outside 17] If the condition is not satisfied (S8), the process is repeated until it is satisfied within the range of Q times (S11). Further, when the above-mentioned conditions are too strict, that is, J _q (q = 1, 2, ..., Q) <0 is satisfied, but | J _q (t _i ) | ≦ ε ¹ lim may not be satisfied. When ε ¹ lim → ε ² lim, (ε ¹ lim <ε ² li
m) is corrected. The time required for the above correction calculation is the sampling number N, 4 so that it can be processed within the time limit T _lim .
Data mean below or above and εli
Determine the m value.

The calculation result of the embodiment of the present invention will be described with reference to FIG. FIG. 4 is a diagram showing a settling curve of the apparatus according to the embodiment of the present invention. The horizontal axis represents time, and the vertical axis represents azimuth error. When the latitude error δλ is corrected as described above, the result approaches δλ ₀ = 0 as shown in FIG. That is, FIG. 4A shows a state where there is no latitude error, that is, a state where δλ ₀ = 0. In this case, the azimuth angle error δψ converges within the allowable range [δψ] _{lim by the} time limit T _lim .

FIG. 4B shows a state in which the correction direction of the latitude error is in the negative direction, that is, the state of δλ ₀ = −δλ ₁ <0. Even in such a case, according to the apparatus of the present invention, the azimuth angle error δψ reaches the allowable range by the time limit T _lim .

FIG. 4C shows a state where the latitude error is corrected in the positive direction, that is, δλ ₀ = δλ ₂ > 0. According to the device of the present invention, the azimuth angle error δψ reaches the allowable range by the time limit T _lim .

[0138]

As described above, according to the present invention,
When the inertial navigation device is reset and a new initial latitude and initial longitude are input, it is possible to shorten the operation time of the correction means required to automatically correct the calculation parameter.
The present invention is required until the inertial navigation device can be used after the inertial navigation device is reset and restarted, even if the ship at berth has a relatively large cycle and a wobbling with a complicated amplitude. The time can be shortened. INDUSTRIAL APPLICABILITY The present invention can shorten the preparation time required for the departure of a ship equipped with an inertial navigation system. The present invention can improve the practical accuracy of an inertial navigation system. INDUSTRIAL APPLICABILITY The present invention can realize an inertial navigation device which can reset accumulated errors frequently and can be operated in a state where accumulated errors are small. INDUSTRIAL APPLICABILITY The present invention can realize an inertial navigation system capable of allowing the occurrence of accumulated error in a range larger than that of conventional devices.

[Brief description of drawings]

FIG. 1 is a block configuration diagram of an apparatus according to an embodiment of the present invention.

FIG. 2 is a flowchart showing the operation of the embodiment of the present invention.

FIG. 3 is a diagram for explaining the operation principle of the embodiment of the present invention.

FIG. 4 is a diagram showing a settling curve of an apparatus according to an embodiment of the present invention.

FIG. 5 is a block diagram showing an example of a conventional inertial navigation system.

FIG. 6 is a diagram showing coordinates used in an inertial navigation system.

FIG. 7 is a diagram showing a settling curve of a conventional example.

[Explanation of symbols]

 1 sensor part 2 sensor error correction part 3 coordinate conversion part 4 coordinate conversion matrix calculation part 5 speed position calculation part 6 attitude azimuth calculation part 7 speed position comparison part 8 error estimation part 9 correction calculation part 10 initial latitude input estimation calculation part 11 acceleration Total 12 Gyro 21 Accelerometer error correction calculator 22 Gyro error correction calculator

─────────────────────────────────────────────────── ───

[Procedure amendment]

[Submission date] August 5, 1994

[Procedure Amendment 1]

[Document name to be amended] Statement

[Correction target item name] 0088

[Correction method] Change

[Correction content]

According to the present invention, the azimuth angle error (δψ (ti), δψ (ti + 1), from the calculation process for each iterative calculation sequentially executed from the time ti by the correction means in the correction mode.
δψ (ti + 2) ...) are sequentially taken out, and the initial latitude ( λ (t ₀ )) is gradually adjusted so that the azimuth error gradually converges.
Small correction value (δλ · δ, including positive or negative sign)
It is characterized by including an initial latitude input estimation calculation means for adding.

Claims (5)

[Claims] 1. An initial latitude (λ (t) input by taking in output signals of an accelerometer fixed to a hull and a gyro.
₀ )) and the initial longitude to calculate the position information and the speed information that are sequentially output, and the position information and the speed information output by the calculation means are compared with the position information and the speed information input from the outside, respectively. An inertial navigation system for a ship, which comprises a comparing means for outputting error information, and the calculating means includes a correcting means for automatically correcting a calculation parameter by repetitive calculation so as to minimize an error generated by taking in the error information. In the device, the azimuth angle error (δψ (ti), δψ (ti + 1), δψ (ti
+2) ...) and add a small correction value (δλ · δ, including a positive or negative sign) to the initial latitude (λ (t ₀ )) so that the azimuth error converges gradually. An inertial navigation system for ships, characterized by comprising latitude input estimation calculation means. 2. The initial latitude input estimation calculation means calculates an average value of the azimuth angle error information for n consecutive operations of the iterative calculation of the correction means to obtain a primary average value, and first The ship according to claim 1, further comprising means for calculating an average value of m times of the secondary average value to obtain a secondary average value, and identifying that the azimuth angle error information converges based on the secondary average value. Inertial navigation system. 3. The initial latitude input estimation calculation means calculates the second average value each time one calculation execution of the calculation means is completed, and the second average value and the previous second average value. The inertial navigation system for a ship according to claim 1, wherein when the difference from the value falls within a preset error range (ε), the calculation is stopped assuming that the azimuth angle error information has converged. 4. The initial latitude input estimation calculation means calculates the second average value every time one calculation execution of the calculation means is completed, and the second average value is calculated despite the correction of the initial latitude a predetermined Q times. The inertial navigation device for a ship according to claim 1 or 2, wherein when the difference between the second average value and the previous second average value does not fall within a preset error range (ε), the calculation is stopped. 5. The fixed value (δλ · δ) is a fixed value δλ (degree) set in advance to a value of about tens of thousands.
And a positive / negative value set in a direction to converge to a value obtained by dividing the difference between the secondary average value executed at that time and the secondary average value executed at the previous time by the error range (ε). The inertial navigation device for a ship according to claim 2, which is a product of a parameter δ (unnamed number) with a sign of.