JPH11344339A5 - - Google Patents

Description

Since each of the magnetic directions θm includes fluctuation, the error also includes fluctuation. In order to eliminate this fluctuation, in the present invention, the error components are subjected to Fourier transform and inverse Fourier transform , and an orientation correction table is created from the error data from which the fluctuation has been eliminated.

[0012]
[Embodiments of the Invention]
Figure 1 is a control block diagram showing one embodiment of the present invention. 1 is a three-axis magnetic sensor equipped with three sets of detection coils arranged on the X-axis, Y-axis, and Z-axis, respectively. 2 is a two-axis tilt sensor that detects pitch and roll as oscillation components. In this example, we used a tilt sensor that measures its own tilt (pitch and roll) from changes in capacitance between multiple capacitor plates protruding from the liquid surface, which changes the liquid level due to tilt.

In this embodiment, if a three-axis magnetic sensor is used in combination with a two-axis tilt sensor that has a mechanism (for example, a gimbal mechanism) that keeps the sensor itself horizontal at all times, the magnetic sensor can be a two-axis sensor consisting of X-axis and Y-axis detection coils, and a two-axis tilt sensor is not required.

[0041]
[Effects of the Invention]
In the present invention, the integrated heading, which has good response over short periods but poor accuracy over long periods, is corrected with a magnetic heading that has good detection accuracy over long periods. This allows for improved long-term accuracy while maintaining the inherent short-term accuracy of the integrated heading. Furthermore, a low-pass filter is used to extract the long-term heading from the magnetic heading data. Inserting this low-pass filter introduces new errors. However, these errors are addressed by offsetting the reference value used to integrate the angular velocity, resulting in a more accurate heading. The second invention determines the true heading, calculates the heading error of the magnetic sensor from the true heading, and creates an error table. To determine the true heading Tdegi, the error of the integrated heading is calculated from the difference between the integrated heading amount when the ship is rotated once and 360°. Based on this error, the true heading for each magnetic heading is calculated. The error for each magnetic heading is then calculated from the difference between the true heading and the magnetic heading. Note that because each of the magnetic headings θm includes a wobble, the error also includes a wobble. In order to eliminate this wobble, in the present invention, the error components are subjected to Fourier transform and inverse Fourier transform , and a more accurate azimuth correction table is created from the error data from which the wobble has been eliminated.

Claims (2)

An electronic magnetic compass equipped with a magnetic compass that detects magnetic direction θm by measuring geomagnetism, and an integrating compass that detects angular velocity around a vertical axis and outputs an integrated direction θs from the integrated value, and obtains an output direction θ by correcting the value of the integrated direction θs with the value of the magnetic direction θm. In the process of the ship turning, each time the magnetic direction θm reaches a predetermined sample angle, the time ti and integrated direction θs at that time are stored, as well as the time tz and integrated direction θdegz for one rotation, the slope a that indicates the error in the integrated direction is calculated from a = (θdegz - 360) / tz, this a is used to calculate the true direction Tdegi relative to the magnetic direction θm, the direction error at each magnetic direction θm is calculated from Tdegi - θm, and a direction correction table is created. 6. An electronic magnetic compass as claimed in claim 5, wherein the components of the heading error in each magnetic heading θm are processed using Fourier transform and inverse Fourier transform techniques to eliminate the wobble error contained in the heading error θm, and a heading correction table is created based on this data.