JPH10122860A - Tidal-wave meter - Google Patents

Abstract

PROBLEM TO BE SOLVED: To provide a tidal-wave meter in which data on only tidal waves can be extracted by a method in which a change signal at a cycle in a range which can be considered to be the cycle of the tidal waves is fetched by using a digital filter. SOLUTION: Change data, with reference to the passage of time on a water- level height, obtained by a tidal-wave meter 1 is converted 2 into a digital signal so as to be input to a digital filter 3. The filter 3 is operated so as to output only change data on a water-level height having a tidal-wave cycle sorted as the cycle of waves, that of tidal waves, that of high tides and that of astronomical tides so as to be sent to a display 4. Thereby, waveform data on only the tidal waves can be extracted so as to be displayed. At this time, in a digital processing operation by, e.g. a finite impulse response (FIR)-type filter 3, impulses can be discriminated by a response when they are input in a time series at a value of 1 in a certain time and at a value of 0 before or after it. The response of the impulses is inversely Fourier-transformed so as to obtain a frequency response. Inversely, when a frequency is given so as to be Fourier-transformed, an impulse response is obtained.

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention belongs to the field of technology for separating and extracting a tsunami component from fluctuation data with respect to a lapse of time of a water level obtained from a wave height meter.

[0002]

2. Description of the Related Art Vertical fluctuations of the sea surface are classified into wind waves (waves), tsunamis, storm surges (meteorological tides), and astronomical tides (tides) according to their fluctuation periods. This is illustrated in FIG. Storm surges are sea-level rises caused by atmospheric pressure.
Astronomical tide is a sea level change caused by astronomical influences such as the sun and the moon, and its main cycle is 12 hours and 24 hours. The actual tide level is the tide level where the tsunami, storm surge, and astronomical tide overlap (this is called the absolute tide level), which is an important judgment target for disasters and evacuation. It is the most important measurement target.

[0003] In addition, conventionally, in addition to such an absolute tide level, a tide level obtained by subtracting an astronomical tide level from an absolute tide level, that is, a tide level in which a tsunami and a storm surge are combined is known to know the situation at the time of a tsunami. And had The separation of the astronomical tide is theoretically simple. This is because the astronomical tide can be predicted by performing a harmonic analysis of the measured tide level data for an appropriate period. As a practical matter, it is necessary to measure the tide level in advance at the installation site for an appropriate period. By subtracting the astronomical tide level thus obtained from the absolute tide level, a combined component of the tsunami and the storm surge can be obtained.

[0004]

However, the period of the tsunami is on the order of tens of minutes, and there is not enough time for evacuation. On the other hand, the cycle of storm surge is in a wide range for more than one hour. The main cycle of the astronomical tide is 12 hours and 24 hours. Therefore, from the point of time for evacuation, it can be said that storm surge and astronomical tide have more margin than tsunami and belong to the same range. On the other hand, absolute tide level display is most useful as tide level information for avoiding damage. Therefore, there is a problem that data obtained by subtracting only the astronomical tide level from the absolute tide level is not so useful data as in the conventional case, and that the data of the tsunami itself cannot be obtained due to the inclusion of the storm surge component.

[0005] Therefore, subtracting from the absolute tide level to the astronomical tide level as well as the storm tide, it may be possible to obtain data on only the tsunami. The storm surge is a sea level change caused by the effect of atmospheric pressure, and the atmospheric pressure varies depending on weather conditions. Therefore, there is a problem in that there is no regularity such as the astronomical tide caused by the movement of the sun or the moon, the prediction cannot be performed like the astronomical tide, and data to be subtracted cannot be obtained.

SUMMARY OF THE INVENTION An object of the present invention is to solve the above-mentioned problems of the prior art, and focuses on the fact that the vertical movement of the sea surface is classified by the period as shown in FIG. An object of the present invention is to provide a tsunami meter capable of obtaining data of only a tsunami by extracting a fluctuation signal having a cycle within a range considered as a tsunami cycle by using the tsunami meter.

[0007]

To achieve the above object, a tsunami meter of the present invention comprises the following means. (B) A wave height meter that measures the fluctuation of the water level over time and outputs the water level data as an analog signal. (B) An A / D converter that converts the water level data from the wave height meter to a digital signal. Digital filter that receives digital water level data from the / D converter and selectively outputs tsunami cycle data among the data contained in it. (D) Receives the output data of the digital filter and responds to the passage of time. Display that displays water level data on the screen in a waveform

[0008]

DESCRIPTION OF THE PREFERRED EMBODIMENTS As a wave height meter in a tsunami meter of the present invention, an ultrasonic wave height meter, a pressure wave height meter, and a Foos type tide level meter which are conventionally used are considered. Of these, the Foos type tide gauge requires correction because the thin tube connecting the well and the sea acts as a low-pass filter, causing a time delay with respect to the tsunami cycle. The A / D converter may be of an existing technology.

Digital filters are generally used most often, so-called FIR (Finite Impulese Respo).
nse) type or IIR (Infinite Inpulese Response) type, but the off-line type which takes out the data and analyzes it after the tsunami hits and calms down
An IIR or FIR type Kalman filter is suitable for collecting tsunami information in real time during a tsunami attack (real-time type). If a certain time delay from the input time from the crest meter to the output of the filter is allowed (a pseudo real-time type), the symmetric FIR type is suitable.

[0010]

DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS An embodiment of the tsunami meter of the present invention will be described below with reference to the drawings. FIG. 1 is a block diagram showing a configuration of an embodiment of a tsunami meter of the present invention. The fluctuation data of the water level with respect to the passage of time obtained by the tsunami meter 1 is sent to the A / D converter 2 and converted into a digital signal, and then converted into a digital filter 3.
Is input to The digital filter 3 operates so as to output only the water level fluctuation data having the tsunami cycle of the cycle classification shown in FIG. As described above, the digital filter may be of the FIR type or the IIR type.

The FIR type is an abbreviation for Finite Impulese Response, and has a finite-length impulse response as it is. An impulse in digital processing is a time series whose value is 1 at a certain time and whose time series values before and after it are zero. The characteristics of the filter can be distinguished by a response (impulse response) when an impulse is input.
When this impulse response is Fourier-transformed, a frequency response is obtained. Conversely, when an inverse Fourier transform is performed by giving a frequency response, an impulse response is obtained.

[0012] If the impulse response of the filter is obtained, the output is calculated as follows. That is, the time series of the input can be regarded as an impulse train having the magnitude of the sample value. If the sample time interval is Δt, the impulse train is input with a shift of Δt.
Therefore, if each impulse response (time series obtained by multiplying the time series of the impulse by the sample value) of this impulse train is shifted in time, a filter output can be obtained. This operation is called folding.

There are three methods for obtaining the impulse response. The first is a method of directly providing an impulse response. For example, there is a method of giving this weight used for a moving average of rectangular weights and a moving average of triangular weights. The inverted left and right weights are actually impulse responses. Therefore, a method of giving a rectangular or triangular impulse response corresponds to this. Although this method is simple, a filter having a desired frequency response characteristic cannot be formed.

The second is a technique for obtaining an impulse response by giving a frequency response. Conventionally, it has been known that good results cannot be obtained if a frequency response is given and a simple Fourier transform is performed to obtain an impulse response. This is because if the frequency response is given and the impulse response is simply calculated, the impulse response has an infinite length. For this reason, when calculating an impulse response from a frequency response, a technique of multiplying by a window function, that is, a “window function method” is used. This is a technique in which both ends of the impulse response are smoothly converged to zero using a window function for the impulse response having an infinite length. In this case, filtering is performed correctly. However, since the filter length is limited, the desired frequency response may not completely match. In general, the longer the length of this filter, the closer to the desired frequency response. However, if the length is increased, the amount of calculation is increased. Therefore, the optimum length is determined through trial and error.

[0015] A third method for obtaining an impulse response is as follows.
In principle, an equation that should be this way is made, and the current value is predicted so that the error is minimized. That is, it is a prediction type. The impulse response of this filter can be obtained by creating and solving simultaneous equations. The filter thus obtained becomes a Kalman filter. In conclusion, "F
Creating an IR digital filter creates an impulse response. "

When a frequency response is given and an impulse response is obtained, if a frequency characteristic that does not cause a phase shift is given, the impulse response shows a response in the past and in the future. Moreover, it shows a symmetric response to zero time. By using this impulse response, an ideal filter having no phase shift can be obtained. But responding in the past, on the contrary,
Since the output of the current point is affected by future data,
It means that it is not determined by the current point. That is, the output is delayed with the time corresponding to the length of the impulse response toward the past. This digital filter has a symmetric FIR
Called type. Such a filter that responds in the past is sometimes called an FIR filter that does not satisfy the causality law.

On the other hand, there is an asymmetric FIR type. This is used to correct the phase of the measuring instrument.
In this case, the frequency response characteristic to be corrected is given, and can be similarly obtained. If a real-time filter with no delay is desired, a filter having an impulse response only in the future may be created. The simultaneous equations can be created and the impulse response can be obtained (Kalman's method). filter). When created in this way, the current output value is determined only by past data. This is sometimes called an FIR filter that satisfies the causality law. As a real-time filter, a method using the impulse response of a filter made of an electric circuit is considered.
R-type filters are rarely made. The reason for this is that it is easier to create with an IIR type digital filter described below, and the amount of calculation is so small that it cannot be compared.

The IIR type filter is based on Infinite Impules.
An abbreviation of e Pesponse, which can handle infinitely long impulse responses. Here, the generally used IIR type will be described. Considering the filter made of electric circuit, it is as follows. The current output is what has been processed since the power was turned on. Further, the storage device is a capacitor or a coil, and the number thereof is very small. If a digital filter corresponding to the same circuit is made, it should be able to be made with little memory.

There are Butterworth-type and Chebyshev-type general low-pass and high-pass filters made of electric circuits. These are primary and secondary depending on the cutoff characteristics.
Next, there is something called .... These are characterized by having a feedback circuit for adding a part of the output to the input side. These are represented by the Laplace transform (represented by the variable s). On the other hand, digital filters are represented by a z-transform. A conversion formula for converting s into z has been proposed, and a digital filter is created by substituting this conversion formula. This is a simple calculation.

A feature of this filter is that the impulse response only works in the future. A filter that satisfies causality. In addition, one having an infinite impulse response can be created. The filtering calculations are very small and can be performed very quickly. However, there is always a phase distortion as in the case of the electric filter.

Considering the use of these filters,
It can be divided into offline type, real-time type, and pseudo real-time type. The offline type is to post-process data. Therefore, an FIR type that can create an ideal filter (can specify both gain and phase) is used.
The real-time type processes at the same time as the input, and outputs almost at the same time as the input, like an analog filter of an electric circuit. The time delay is only the processing time. Therefore, the IIR type or the FIR type (FI
R-type Kalman filter) is used. Although these have differences in characteristics, it is difficult to completely satisfy both the gain and the phase. The pseudo real-time type is a filter used when the output time can be delayed to some extent from the input time without shifting the phase. In this case, a symmetric FIR type having a short filter length is used.

From the above, in the present embodiment, a Kalman filter is employed for real-time display during the tsunami attack. This is because the filter is a real-time filter with little phase distortion. In the case where another analysis is performed after the tsunami hits and calms down, a symmetric FIR filter was used. This is because the post-processing does not cause a time delay and has an ideal characteristic as a filter. The above description has been given of the case where tsunami data is extracted. However, the present invention is not limited to this. If the response characteristic of the digital filter is set within the range of the storm surge cycle, data on the water surface fluctuation of the storm surge can be taken out and the cycle of the wave Wave data can be extracted by setting the range.

FIG. 2 shows waveform data obtained by performing real-time tsunami extraction using a Kalman filter. The horizontal axis is time, and the vertical axis is water level. The data that fluctuates in the upper sawtooth shape is the data obtained from the wave height meter, and is the data in which the tsunami, the storm surge, and the astronomical tide are combined. The gentle curves overlaid represent the overlapping tides of astronomical tides and storm surges. The lower waveform is the extracted tsunami-only waveform.

FIG. 3 shows tsunami extraction using a symmetric FIR filter. The upper waveform is a waveform from a wave gauge, and includes a storm surge, an astronomical tide, and a wave having a period of 200 seconds or less. The lower waveform is a waveform obtained by extracting only the tsunami from the upper waveform by a digital filter.

FIG. 4 shows a data obtained by extracting astronomical tide + storm surge data, which also uses a symmetric FIR filter. The upper data is the wavemeter data, and the lower curve is the combined water level fluctuation data of the astronomical tide and storm surge extracted from the upper data. As described above, the waveforms of FIGS. 3, 4, and 5 are displayed on the screen of the display 4. The digital filter is performed by a personal computer.

[0026]

As described above, the tsunami meter of the present invention
Since the data obtained by the wave height meter is sent to the display after passing through a digital filter that is set to pass only the signal of the tsunami cycle, the waveform data of only the tsunami must be extracted and displayed. There is an advantage that can be.

[Brief description of the drawings]

FIG. 1 is a block diagram illustrating a configuration of an embodiment of a tsunami meter of the present invention.

FIG. 2 is a waveform data diagram obtained by performing real-time tsunami extraction using a Kalman filter in the embodiment of the present invention.

FIG. 3 is an extracted waveform data diagram of a tsunami using a symmetric FIR filter in the embodiment of the present invention.

FIG. 4 is a waveform diagram in which overlapping fluctuations of astronomical tide and storm surge are extracted using a symmetrical FIR filter in the embodiment of the present invention.

FIG. 5 is a diagram showing the relative magnitude of energy by classifying the vertical fluctuation of the sea surface according to its fluctuation cycle.

[Explanation of symbols]

 1 Crest meter 2 A / D converter 3 Digital filter 4 Display

Claims (1)

[Claims] 1. A tsunami meter comprising the following components. (B) A wave height meter that measures the fluctuation of the water level over time and outputs the water level data as an analog signal. (B) An A / D converter that converts the water level data from the wave height meter to a digital signal. Digital filter that receives digital water level data from the / D converter and selectively outputs tsunami cycle data among the data contained in it. (D) Receives the output data of the digital filter and responds to the passage of time. Display that displays water level data on the screen in a waveform