System and method for detecting long-period wave in real time

Abstract

The present invention relates to a system and a method for detecting a long-period wave in real time, and the system comprises: a quality processing unit for assigning flags to time-series data transmitted from a wave height buoy and removing noise; an imputed data generation unit for generating imputed data on the basis of acceleration data and gyro data corresponding to a missing section of the time-series data; a data interpolation unit for interpolating the imputed data into the time-series data to compensate for the missing section; and a long-period wave detection unit for detecting a long-period wave from the time-series data into which the imputed data has been interpolated.

Description

Real-time long-wave detection system and method

[0001] The present invention relates to a real-time long-wave detection system and method, and more particularly, to a real-time long-wave detection system and method that flags and removes noise in wave height data transmitted in real time from a wave buoy, generates missing supplementary data based on acceleration data and gyro data for missing sections, and then interpolates and supplements the missing data into the wave height data, thereby reducing the frequency of false detection of tsunami components and dramatically increasing the accuracy of information collection in the process of monitoring long-waves in real time.

[0002] Tsunamis can cause serious disasters, including coastal flooding and structural destruction. Earthquakes off the coast of Japan in 1983, 1993, and 2024 triggered tsunamis that reached South Korea's east coast, causing damage.

[0003] Meanwhile, when a tsunami occurs, predicting tsunami height and arrival time is essential to minimize damage. To this end, the Korea Meteorological Administration operates tsunami wave gauges, tide gauge data, and marine surveillance CCTV.

[0004] However, most of the current observation equipment is located along the coast, so there is a limit to the time available to detect tsunamis caused by earthquakes in the coastal waters.

[0005] Therefore, there is an increasing need to install observation equipment in the open sea rather than using tsunami wave gauges and tide gauges located along the coast, extract tsunami components, and perform early detection and warning.

[0006] In relation to this, a method was used in which time series data on sea level displacement measured in real time were extracted at specific time intervals to generate sea level displacement data for analysis, time-domain analysis and spectrum analysis were performed on the sea level displacement time series data for analysis to extract wave information including significant wave heights, and then tidal analysis was performed on the sea level displacement time series data for analysis to calculate the predicted sea level height at the current analysis point in time, and then a method was used to monitor long-period waves in real time.

[0007] However, these conventional methods do not undergo an appropriate quality control process when noise occurs in the wave height data or when the communication environment with the base station deteriorates, resulting in missing sections. Therefore, there is a possibility that the collected information may contain errors, leading to false tsunami warnings.

[0008] The present invention is intended to solve the above-mentioned problem, and provides a real-time long-period wave detection system and method that can reduce the frequency of false detection of tsunami components and dramatically increase the accuracy of information collection in the process of monitoring long-period waves in real time by flagging and removing noise in wave height data transmitted in real time from a wave buoy, generating missing supplementary data based on acceleration data and gyro data for missing sections, and then interpolating and supplementing the missing data into the wave height data.

[0009] A real-time long-wave detection system according to one embodiment of the present invention may include a quality processing unit that removes noise from time series data generated based on real-time wave data transmitted from a wave buoy and assigns a flag, a missing supplementary data generation unit that generates missing supplementary data for time series data classified as a missing target flag based on acceleration data and gyro data corresponding to a missing section among the entire section of time series data classified as a missing target flag, a data interpolation unit that interpolates and supplements the generated missing supplementary data into the time series data classified as the missing target flag, and a long-wave detection unit that detects a long-wave from the time series data into which the missing supplementary data is interpolated.

[0010] The quality processing unit according to one embodiment of the present invention can perform a preset test on time series data and determine a flag to be assigned to the time series data based on the test result.

[0011] The quality processing unit according to one embodiment of the present invention can perform the above tests including an observation time confirmation test, an observation location confirmation test, an observation data format and content reception confirmation test, a spike confirmation test, an instantaneous change rate confirmation test, a constant value observation confirmation test, a step confirmation test, and an observation data volatility axis confirmation test.

[0012] In one embodiment of the present invention, the quality processing unit determines the observation time point of the time series data during the observation time confirmation test process, determines whether the latitude and longitude values ​​of the time series data exceed a preset range during the observation location confirmation test process, determines whether any one or more of the time, wave height, latitude and longitude values ​​of the time series data are missing during the observation data format and content reception confirmation test process, determines whether any one or more of the time, wave height, latitude and longitude values ​​of the time series data are missing during the spike confirmation test process, calculates the variance within an arbitrary section of the time series data, and classifies as noise any value exceeding n times the variance, and classifies as noise any value exceeding the range of the minimum and maximum values ​​of wave heights that can occur in the relevant sea area during the instantaneous change rate confirmation test process, and classifies as noise any value exceeding the range of the minimum and maximum values ​​of wave heights that can occur in the relevant sea area during the constant value observation confirmation test process, and classifies as noise any value if the fluctuation of n consecutive wave height data falls within a preset threshold value, and classifies as noise any value if the step confirmation test process, compares the wave height average within an arbitrary section of the time series with the wave height average from the start section of the time series to the preceding section of the arbitrary section, and classifies as noise any value if the wave height difference is greater than a preset value, and During the volatility axis verification test process, if the measured wave height shows at least one trend of increasing or decreasing over a preset period of time, it can be classified as noise.

[0013] The quality processing unit according to one embodiment of the present invention may assign a 0 flag if the time series data does not pass the test, a 1 flag if the time series data passes the test, a 3rd flag if the time series data passes the test but residual noise is confirmed, and a 4th flag if the time series data does not pass the test.

[0014] According to one embodiment of the present invention, the quality processing unit may classify time series data corresponding to the fourth flag, among the time series data, in which a missing section is longer than a threshold time, as the missing target flag, and may enable the missing supplementary data generation unit to specify the corresponding time series data as a target for generation of the missing supplementary data.

[0015] The missing supplementary data generation unit according to one embodiment of the present invention can generate the missing supplementary data based on the sea level displacement calculation result for the missing section based on acceleration data and gyro data corresponding to the missing section among the entire section of time series data classified as a missing target flag.

[0016] The missing supplement data generation unit according to one embodiment of the present invention can convert the acceleration data into the sea level displacement using a fast Fourier transform (FFT) and a discrete wavelet transform (DWT) in the process of calculating the sea level displacement for the missing section.

[0017] The data interpolation unit according to one embodiment of the present invention may perform low-pass filtering for a preset time or longer on time series data from which noise has been removed, extract long-period components, add the sea level displacement calculation result to the extracted long-period components, and then add the initial noise section of the time series data from which noise has been removed and the section derived by adding the sea level displacement calculation result to the long-period components to each other, thereby performing the interpolation.

[0018] The long-wave detection unit according to one embodiment of the present invention can detect a long-wave in a preset frequency band by performing band-pass filtering that continuously performs low-pass filtering and high-pass filtering on time series data into which the missing supplementary data is interpolated.

[0019] The long-wave detection unit according to one embodiment of the present invention can set the blocking period of the bandpass filtering to between 1 minute and 3 hours.

[0020] The long-wave detection unit according to one embodiment of the present invention can detect the long-wave by setting the blocking periods of the low-pass filtering and the high-pass filtering to be different from each other.

[0021] A real-time long-period wave detection method according to another embodiment of the present invention may include a step of removing noise from time series data generated based on real-time wave height data transmitted from a wave buoy through a quality processing unit and assigning a flag, a step of generating missing supplementary data for time series data classified as a missing target flag based on acceleration and gyro data corresponding to a missing section among the entire section of time series data classified as a missing target flag through a missing supplementary data generation unit, a step of interpolating and supplementing the generated missing supplementary data into the time series data classified as the missing target flag through a data interpolation unit, and a step of detecting a long-period wave from the time series data into which the missing supplementary data is interpolated through a long-period wave detection unit.

[0022] According to the present invention, noise in wave height data transmitted in real time from a wave buoy is flagged and removed, and missing supplementary data is generated based on acceleration data and gyro data for missing sections and then interpolated into the wave height data to supplement the data, thereby reducing the frequency of false detection of tsunami components in the process of monitoring long-period waves in real time and dramatically increasing the accuracy of information collection.

[0023] In addition, according to the present invention, it has the advantage of being able to monitor and detect long-wave waves in real time and link them with a tsunami evacuation warning system to minimize damage from tsunamis when an earthquake occurs.

[0024] In addition, according to the present invention, as a solution to the problem of the absence of a detection method for missing data in the existing long-wave detection technology, a missing data supplementation algorithm is added, so that even when an unexpected RTK-GPS buoy malfunctions, correction is performed in real time through acceleration data and interpolation is performed into the missing section, thereby having the advantage of being able to apply the long-wave detection algorithm regardless of whether there is missing data.

[0025] FIG. 1 is a schematic diagram showing the configuration of a real-time long-wave detection system (100) according to one embodiment of the present invention.

[0026] FIG. 2 is a conceptual diagram illustrating the concept of detecting real-time long-period waves through the real-time long-period wave detection system (100) illustrated in FIG. 1.

[0027] Figure 3 is a diagram showing an example of a type of noise observable through a wave buoy.

[0028] Figure 4 is a diagram showing a section of time series data recognized as noise in the quality processing unit (110).

[0029] Figure 5 is a drawing showing a state in which noise is removed through a spike confirmation test of the quality processing unit (110).

[0030] Figure 6 is a drawing showing a state in which noise is removed through a step verification test of the quality processing unit (110).

[0031] Figure 7 is a diagram showing the final time series data derived by interpolating missing supplementary data into time series data.

[0032] Figure 8 is a diagram showing the results of comparison with wave height data that was not supplemented through interpolation of missing supplementary data.

[0033] Figure 9 is a flowchart showing a real-time long-period wave detection method according to one embodiment of the present invention in a series of steps.

[0034] The advantages and features of the present invention, and the methods for achieving them, will become clearer with reference to the embodiments described below together with the accompanying drawings. However, the present invention is not limited to the embodiments disclosed below and may be implemented in various different forms. These embodiments are provided solely to ensure that the disclosure of the present invention is complete and to fully inform those skilled in the art of the scope of the invention, and the present invention is defined solely by the scope of the claims.

[0035] The terms used in this specification will be briefly explained, and the present invention will be described in detail.

[0036] The terms used in this invention have been selected from widely used, current terms, taking into account the functions of the invention. However, these terms may vary depending on the intentions of those skilled in the art, precedents, the emergence of new technologies, etc. Furthermore, in certain cases, terms may be arbitrarily selected by the applicant, in which case their meanings will be described in detail in the relevant description of the invention. Therefore, the terms used in this invention should not be defined simply as names, but rather based on their inherent meanings and the overall content of the invention.

[0037] When a part of the specification is said to "include" a component, this does not mean that other components are excluded, but rather that other components may be included, unless otherwise specifically stated. Furthermore, terms such as "part," "module," and "unit" used in the specification mean a unit that processes at least one function or operation, and may be implemented by software, a hardware component such as an FPGA or ASIC, or a combination of software and hardware. However, terms such as "part," "module," and "unit" are not limited to software or hardware. A "part," "module," and "unit" may be configured to reside on an addressable storage medium, or may be configured to execute one or more processors. Thus, as an example, terms such as "unit," "module," and "unit" include components such as software components, object-oriented software components, class components, and task components, as well as processes, functions, attributes, procedures, subroutines, segments of program code, drivers, firmware, microcode, circuitry, data, databases, data structures, tables, arrays, and variables.

[0038] Below, with reference to the attached drawings, embodiments of the present invention are described in detail so that those skilled in the art can easily practice them. Furthermore, in order to clearly explain the present invention, portions irrelevant to the description are omitted in the drawings.

[0039] Terms including ordinal numbers, such as "first," "second," etc., may be used to describe various components, but the components are not limited by the terms. The terms are used solely to distinguish one component from another. For example, without departing from the scope of the present invention, the first component could be referred to as the second component, and similarly, the second component could also be referred to as the first component. The term "and / or" includes any combination of multiple related items or any one of multiple related items.

[0040]

[0041] FIG. 1 is a schematic diagram showing the configuration of a real-time long-wave detection system (100) according to one embodiment of the present invention, and FIG. 2 is a conceptual diagram showing the concept of detecting real-time long-waves through the real-time long-wave detection system (100) shown in FIG. 1.

[0042] Looking at FIGS. 1 and 2, a real-time long-wave detection system (100) according to one embodiment of the present invention may largely include a quality processing unit (110), a missing supplement data generation unit (120), a data interpolation unit (130), and a long-wave detection unit (140).

[0043] The quality processing unit (110) can assign a flag to each time series data generated based on real-time wave height data transmitted from a wave buoy and remove noise. Here, the wave buoy may refer to an RTK-GPS wave buoy. The noise that can be removed by the quality processing unit (110) will be examined in more detail as follows.

[0044]

[0045] FIG. 3 is a diagram showing an example of a type of noise observable through a wave buoy, FIG. 4 is a diagram showing a section of time series data recognized as noise by the quality processing unit (110), FIG. 5 is a diagram showing a state in which noise is removed through a spike confirmation test of the quality processing unit (110), and FIG. 6 is a diagram showing a state in which noise is removed through a step confirmation test of the quality processing unit (110).

[0046] Looking at FIGS. 3 to 6, the types of noise that can be recognized and processed by the quality processing unit (110) according to one embodiment of the present invention can be broadly classified into a flat type as in FIG. 3(a), a spike type as in FIG. 3(b), an outlier type as in FIG. 3(c), and a complex type as in FIG. 3(d).

[0047] Flat type noise has a shape in which some sections of the entire time series are slanted upwards and downwards, while spike type noise has a shape in which some sections are pointed upwards. Outlier type noise has a shape in which some sections are slanted at a certain angle, and complex type noise has a shape that is a complex mixture of flat type, spike type, and outlier type.

[0048] If such noise is found within the time series data, the quality processing unit (110) can delete the interval corresponding to the noise within the entire interval of the discovered time series data. The deleted interval can be supplemented through the interpolation process of the data interpolation unit (130) using missing supplementary data generated by the missing supplementary data generation unit (120) described below.

[0049] Additionally, in one embodiment, the quality processing unit (110) may perform a preset test on time series data and, based on the test results, determine a flag to be assigned to the time series data from which the noise has been removed.

[0050] Here, the preset tests may refer to an observation time confirmation test, an observation location confirmation test, an observation data format and content reception confirmation test, a spike confirmation test, an instantaneous change rate confirmation test, a constant value observation confirmation test, a step confirmation test, and an observation data volatility axis confirmation test.

[0051] In this process, the quality processing unit (110) can determine the observation time point of the time series data during the observation time confirmation test process.

[0052] Additionally, the quality processing unit (110) can determine whether the latitude and longitude values ​​of the time series data exceed a preset range during the observation location confirmation test process.

[0053] In addition, the quality processing unit (110) can determine whether or not any one or more of the time, wave height, latitude, and longitude values ​​of the time series data is missing during the observation data format and content reception confirmation test process.

[0054] Additionally, during the spike verification test process, the quality processing unit (110) can calculate the variance within an arbitrary interval of time series data and classify values ​​exceeding n times the variance as noise. Here, the arbitrary interval may mean, for example, 1 minute, and n may mean, for example, 13.

[0055] Additionally, during the instantaneous change rate verification test process, the quality processing unit (110) can classify values ​​exceeding the minimum and maximum wave height ranges that can occur in the relevant sea area as noise. Here, the maximum wave height may be, for example, 30 m, and the minimum wave height may be -30 m.

[0056] In addition, the quality processing unit (110) can classify as noise if the fluctuation of n consecutive wave height data falls within a preset threshold during the test process for confirming whether a certain value is observed. Here, n can be, for example, 25, and the preset threshold can be 0.01 m.

[0057] In addition, during the step verification test process, the quality processing unit (110) can compare the wave height average within an arbitrary section of the time series with the wave height average from the start section of the time series to the preceding section of the arbitrary section, and classify it as noise if the wave height difference is greater than a preset value. Here, the arbitrary section can be, for example, a section corresponding to 8 minutes, and n can be 0.2.

[0058] Additionally, the quality processing unit (110) may classify as noise if the measured wave height shows an increasing or decreasing trend over a preset period of time during the observation data variability axis confirmation test process. Here, n may be, for example, 13.

[0059] Additionally, in one embodiment, the quality processing unit (110) may assign a 0 flag if the raw data received in real time has not passed the eight tests discussed above, a 1st flag if the time series data passes the test, a 3rd flag if the time series data passes the test but residual noise is confirmed, and a 4th flag if the time series data does not pass the test.

[0060] Accordingly, based on this, the quality processing unit (110) classifies time series data whose missing section is longer than a threshold time among the time series data corresponding to the 0th flag, the 1st flag, the 3rd flag, and the 4th flag, and particularly among the time series data corresponding to the 4th flag, as a missing target flag, so that the corresponding time series data can be specified as a target for generating missing supplementary data through the missing supplementary data generating unit (120) described below.

[0061] Additionally, in one embodiment, the quality processing unit (110) may check the time information of the entire time series during the observation time confirmation test process for the eight test results discussed above, and if an abnormality occurs, assign a fourth flag to the time series data, and assign a first flag in the case of a normal value. Here, the quality processing unit (110) may check whether each time of the time series data is sequentially received, and whether a time past or future than the time when the data is transmitted is not indicated, etc.

[0062] Additionally, the quality processing unit (110) can check the latitude and longitude positions of the entire time series during the observation location confirmation test process, assigning a fourth flag if the position exceeds a preset range, and assigning a first flag if the position is normal. At this time, the latitude and longitude ranges can be changed at any time, taking into account the characteristics of the observation equipment.

[0063] In addition, the quality processing unit (110) may assign a fourth flag if it is determined that there is a missing value in the time, wave height, latitude, and longitude values ​​of the entire time series during the observation data format and content reception confirmation test process, and may assign a first flag if the values ​​are normal. In this process, the quality processing unit (110) may check whether there is a missing section in each time series data.

[0064] Additionally, the quality processing unit (110) may calculate the variance within the processing interval during the spike confirmation test process, determine a value above a threshold as an outlier, assign a fourth flag, and assign a first flag to a normal value. Here, the threshold may be set in consideration of the overall quality of the time series data, and the quality processing unit (110) may detect temporarily high or low values ​​as noise during the spike confirmation test process.

[0065] Additionally, the quality processing unit (110) can designate the maximum and minimum wave heights that can be achieved in a region during the instantaneous change rate verification test process, assigning a fourth flag to outliers that fall outside the range, and assigning a first flag to normal values. At this time, the maximum and minimum values ​​can be changed at any time to reflect the characteristics of the sea area where wave height measurement equipment is installed.

[0066] In addition, the quality processing unit (110) can assign a third flag to the data in the corresponding time period if a constant value is input without a change in wave height for a set threshold time period during the test process for confirming whether a constant value is observed, and can assign a first flag in the case of a normal value. At this time, the quality processing unit (110) can assign a third flag if, for example, the fluctuation value of 10 consecutive wave height data is 0.01 m or less by reflecting the characteristics of the observation equipment. The quality processing unit (110) can detect noise that does not cause a change in wave height for a certain period of time through the test for confirming whether a constant value is observed.

[0067] Additionally, during the step verification test process, if the time series data continues to rise or fall momentarily, the quality processing unit (110) may determine it as an abnormal value and assign a fourth flag, and assign a first flag to a normal value. At this time, the quality processing unit (110) may assign a fourth flag, reflecting the characteristics of the observation equipment, for example, if the average wave height for 8 minutes is 0.2 m higher than the average wave height for the previous section.

[0068] In addition, during the observation data variability axis confirmation test process, the quality processing unit (110) checks whether there is a wave height that continuously rises or falls for a critical period of time, and if a continuous increase or decrease is confirmed for a critical period of time, it can be determined as an abnormal value and assigned a fourth flag, and a normal value can be assigned a first flag. At this time, the quality processing unit (110) can consider the characteristics of the sea area where the observation equipment is installed and, for example, if the wave height rises or falls continuously for 25 seconds, it can be distinguished as noise.

[0069]

[0070] Returning to FIGS. 1 and 2 again, the missing supplementary data generation unit (120) can generate missing supplementary data for the time series data classified as the missing target flag based on acceleration data and gyro data corresponding to the missing section among the entire section of the time series data classified as the missing target flag through the quality processing unit (110).

[0071] More specifically, the missing supplementary data generation unit (120) can generate missing supplementary data based on the sea level displacement calculation result for the missing section based on acceleration data and gyro data corresponding to the missing section among the entire section of time series data classified as a missing target flag.

[0072] At this time, the missing supplementary data generation unit (120) can utilize the three-directional acceleration and three-directional rotation angle in the real-time wave height data collected through the wave height buoy. In this process, the missing supplementary data generation unit (120) performs coordinate axis transformation of the acceleration information measured in the navigation coordinate system into the geographic coordinate system using quaternions. Thereafter, the missing supplementary data generation unit (120) removes mechanical noise using the moving average method for the processed z-direction acceleration.

[0073] In addition, the missing supplement data generation unit (120) can convert acceleration data into sea level displacement using a fast Fourier transform (FFT) and a discrete wavelet transform (DWT) to reduce acceleration integration error due to drift in the process of calculating sea level displacement for a missing section.

[0074]

[0075] The data interpolation unit (130) can interpolate missing supplementary data generated through the missing supplementary data generation unit (120) into time series data classified by the corresponding missing target flag to supplement it.

[0076] More specifically, the data interpolation unit (130) performs low-pass filtering for a preset time period or longer on the time series data from which noise has been removed, extracts long-term components, adds the sea level displacement calculation result to the extracted long-term components, and then adds the initial noise section of the time series data from which noise has been removed and the section derived by adding the sea level displacement calculation result to the long-term components to perform interpolation.

[0077]

[0078] The long-wave detection unit (140) can detect long-waves from time series data into which missing supplementary data has been interpolated.

[0079] More specifically, the long-wave detection unit (140) can detect long-waves in a preset frequency band by performing band-pass filtering, which continuously performs low-pass filtering and high-pass filtering on time series data into which missing supplementary data has been interpolated.

[0080] At this time, the long-wave detection unit (140) can perform band-pass filtering by setting the cutoff period of band-pass filtering to between 1 minute and 3 hours so that only the wind wave component can be acquired by considering the wave characteristics of the sea area.

[0081] Additionally, in one embodiment, the long-wave detection unit (140) can set the cutoff periods for low-pass filtering and high-pass filtering to 1 minute and 3 hours, respectively. Through this, after long-wave detection, only wave components between 1 minute and 3 hours remain in the wave height data.

[0082] Figure 7 is a diagram showing the final time series data derived by supplementing missing supplementary data through interpolation into time series data, and Figure 8 is a diagram showing the result of comparison with wave height data for which supplementation through interpolation of missing supplementary data was not performed.

[0083] Looking at Figure 7, before missing supplementary data is interpolated into time series data, noise and missing sections are formed in a specific section as in Figure 7(a), whereas when missing supplementary data generated through the missing supplementary data generation unit (120) is interpolated into time series data and supplemented, the corresponding noise and missing sections can be supplemented as in Figures 7(b) and 7(c).

[0084] In addition, as shown in FIG. 8, the long-wave detection unit (140) detects long-wave components of time series data generated through missing supplementation of missing supplementation data and time series data that has only had noise removed, thereby showing a large difference when compared to time series data that has not had missing supplementation performed, thereby drastically reducing the possibility of false alarms due to earthquakes and tsunamis.

[0085]

[0086] Next, we will examine in order the entire process of detecting real-time long-period waves from time series data using the real-time long-period wave detection system (100) discussed above.

[0087] Figure 9 is a flowchart showing a real-time long-period wave detection method according to one embodiment of the present invention in a series of steps.

[0088] Looking at Figure 9, first, the quality processing unit assigns a flag to each time series data generated based on real-time wave height data transmitted from the wave buoy and removes noise (S901).

[0089] In step S901, if such noise is found within the time series data, the quality processing unit can delete the interval corresponding to the noise within the entire interval of the discovered time series data. For the deleted interval, the missing supplementary data generated by the missing supplementary data generation unit can be supplemented through the interpolation process of the data interpolation unit. Furthermore, in step S901, the quality processing unit can also assign a flag.

[0090] Next, the missing supplementary data generation unit generates missing supplementary data for the time series data classified as the missing target flag based on acceleration and gyro data corresponding to the missing section among the entire section of the time series data classified as the missing target flag (S902).

[0091] In step S902, the missing supplement data generation unit can use the three-directional acceleration and three-directional rotation angle in real-time wave height data collected through the wave height buoy, and in the process of calculating sea level displacement for the missing section, the acceleration data can be converted into sea level displacement using a fast Fourier transform (FFT) to reduce acceleration integration error due to drift.

[0092] Next, the data interpolation unit interpolates the previously generated missing supplementary data into the time series data classified with the corresponding missing target flag to supplement it (S903), and the long-period wave detection unit detects long-period waves from the time series data into which the missing supplementary data has been interpolated (S904).

[0093] In step S903, the data interpolation unit may perform low-pass filtering for a preset time period or longer on the time series data from which noise has been removed, extract long-term components, add the sea level displacement calculation result to the extracted long-term components, and then perform interpolation by adding the initial noise section of the time series data from which noise has been removed and the section derived by adding the sea level displacement calculation result to the long-term components.

[0094] In addition, in step S904, the long-period wave detection unit can detect long-period waves in a preset frequency band by performing band-pass filtering that continuously performs low-pass filtering and high-pass filtering on time series data with interpolated missing supplementary data.

[0095]

[0096] A description of the code of a real-time long-wave detection system according to one embodiment of the present invention is as follows.

[0097] 100: Real-time long-wave detection system 110: Quality processing unit

[0098] 120: Missing supplement data generation unit 130: Data interpolation unit

[0099] 140: Long-wave detection unit

[0100]

[0101] Those skilled in the art will appreciate that the embodiments of the present invention can be implemented in modified forms without departing from the essential characteristics of the above description. Therefore, the disclosed methods should be considered illustrative rather than restrictive. The scope of the present invention is indicated by the claims, not the detailed description thereof, and all differences within the scope equivalent thereto should be construed as being included within the scope of the present invention.

[0102] The real-time long-wave detection system according to one embodiment of the present invention is a technology that accurately detects long-waves in real time based on ocean data collected through wave buoys. It can be effectively applied to various marine monitoring industries, such as earthquake and tsunami early warning systems, marine disaster prevention systems, and national marine observation infrastructure. In particular, it includes real-time data processing and missing data supplementation algorithms, thereby overcoming the limitations of existing systems, enhancing the accuracy of information collection, and contributing to the prevention of damage caused by natural disasters. Therefore, the present invention has high applicability in industries where marine technology, disaster prevention technology, and information and communication technology converge.

Claims

1. A quality processing unit that flags time series data generated based on real-time wave height data transmitted from a wave buoy and removes noise; A missing supplementary data generation unit that generates missing supplementary data for time series data classified as a missing target flag based on acceleration data and gyro data corresponding to a missing section among the entire section of time series data classified as a missing target flag; A data interpolation unit that interpolates the generated missing supplementary data into the time series data classified by the corresponding missing target flag to supplement it; and A long-period wave detection unit that detects long-period waves from time series data into which the above missing supplementary data is interpolated; A real-time long-wave detection system including:

2. In paragraph 1, The above quality processing department, It is to perform a pre-set test on time series data and determine the flag to be assigned to the time series data based on the test results. Real-time long-wave detection system.

3. In paragraph 2, The above quality processing department, The above tests are conducted including an observation time confirmation test, an observation location confirmation test, an observation data format and content reception confirmation test, a spike confirmation test, an instantaneous change rate confirmation test, a constant value observation confirmation test, a step confirmation test, and an observation data volatility axis confirmation test. Real-time long-wave detection system.

4. In paragraph 3, The above quality processing department, In the above observation time confirmation test process, the observation time of the time series data is determined, In the above observation location confirmation test process, it is determined whether the latitude and longitude values ​​of the time series data exceed the preset range. In the process of testing the receipt of the above observation data format and content, it is determined whether there is a missing value in one or more of the time, wave height, latitude, and longitude values ​​of the time series data. In the above spike verification test process, the variance within an arbitrary section of time series data is calculated, and values ​​exceeding n times the variance are classified as noise. In the above instantaneous change rate verification test process, values ​​exceeding the minimum and maximum range of wave heights that can occur in the relevant sea area are classified as noise, In the test process to confirm whether the above constant value is observed, if the fluctuation of n consecutive wave height data falls within the preset threshold, it is classified as noise. In the above step verification test process, the wave height average within an arbitrary section of the time series and the wave height average from the start section of the time series with noise removed to the preceding section of the arbitrary section are compared with each other, and if the wave height difference is greater than a preset value, it is classified as noise. In the above observation data volatility axis confirmation test process, if the measured wave height shows at least one trend of increasing or decreasing over a preset period of time, it is classified as noise. Real-time long-wave detection system.

5. In paragraph 2, The above quality processing department, If the time series data does not pass the above test, a 0 flag is assigned. If the time series data passes the above test, the first flag is assigned. If the time series data passes the above test but residual noise is confirmed, a third flag is assigned. If the time series data fails the above test, a fourth flag is assigned. Real-time long-wave detection system.

6. In paragraph 5, The above quality processing department, Among the time series data corresponding to the above fourth flag, the time series data for which the missing section is longer than the threshold time is classified as the missing target flag, and the time series data is specified as the target for generating the missing supplementary data through the missing supplementary data generating unit (120). Real-time long-wave detection system.

7. In paragraph 1, The above missing supplement data generation unit is, Based on the acceleration data and gyro data corresponding to the missing section among the entire section of time series data classified as a missing target flag, the missing supplementary data is generated based on the result of calculating sea level displacement for the missing section. Real-time long-wave detection system.

8. In paragraph 7, The above missing supplement data generation unit is, In the process of calculating sea level displacement for the above missing section, the acceleration data is converted into the sea level displacement using a fast Fourier transform (FFT) and a discrete wavelet transform (DWT). Real-time long-wave detection system.

9. In paragraph 7, The above data interpolation part is, After performing low-pass filtering for a preset time period or longer on the time series data from which noise has been removed, long-term components are extracted, and the sea level displacement calculation result is added to the extracted long-term components. Then, the initial noise section of the time series data from which noise has been removed and the section derived by adding the sea level displacement calculation result to the long-term components are added together to perform the interpolation. Real-time long-wave detection system.

10. In paragraph 1, The above long-wave detection unit, Band-pass filtering is performed by continuously performing low-pass filtering and high-pass filtering on time series data into which the above missing supplementary data is interpolated to detect long-period waves in a preset frequency band. Real-time long-wave detection system.

11. In paragraph 10, The above long-wave detection unit, The cutoff period of the above bandpass filtering is set to between 1 minute and 3 hours. Real-time long-wave detection system.

12. In paragraph 10, The above long-wave detection unit, The long-wave is detected by setting the blocking cycles of the low-pass filtering and high-pass filtering to be different from each other. Real-time long-wave detection system.

13. A step of flagging and removing noise from time series data generated based on real-time wave height data transmitted from a wave buoy through a quality processing unit; A step of generating missing supplementary data for time series data classified as a missing target flag based on acceleration and gyro data corresponding to the missing section among the entire section of time series data classified as a missing target flag through a missing supplementary data generation unit; A step of interpolating the generated missing supplementary data into the time series data classified with the corresponding missing target flag through data interpolation; and A step of detecting a long-period wave from time series data into which the missing supplementary data is interpolated through a long-period wave detection unit; A real-time long-period wave detection method including:

Images