WO2016079848A1 - 状態推定装置 - Google Patents
状態推定装置 Download PDFInfo
- Publication number
- WO2016079848A1 WO2016079848A1 PCT/JP2014/080773 JP2014080773W WO2016079848A1 WO 2016079848 A1 WO2016079848 A1 WO 2016079848A1 JP 2014080773 W JP2014080773 W JP 2014080773W WO 2016079848 A1 WO2016079848 A1 WO 2016079848A1
- Authority
- WO
- WIPO (PCT)
- Prior art keywords
- unit
- value
- covariance matrix
- error covariance
- calculation unit
- Prior art date
Links
Images
Classifications
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01C—MEASURING DISTANCES, LEVELS OR BEARINGS; SURVEYING; NAVIGATION; GYROSCOPIC INSTRUMENTS; PHOTOGRAMMETRY OR VIDEOGRAMMETRY
- G01C13/00—Surveying specially adapted to open water, e.g. sea, lake, river or canal
- G01C13/002—Measuring the movement of open water
- G01C13/004—Measuring the movement of open water vertical movement
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01C—MEASURING DISTANCES, LEVELS OR BEARINGS; SURVEYING; NAVIGATION; GYROSCOPIC INSTRUMENTS; PHOTOGRAMMETRY OR VIDEOGRAMMETRY
- G01C13/00—Surveying specially adapted to open water, e.g. sea, lake, river or canal
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01S—RADIO DIRECTION-FINDING; RADIO NAVIGATION; DETERMINING DISTANCE OR VELOCITY BY USE OF RADIO WAVES; LOCATING OR PRESENCE-DETECTING BY USE OF THE REFLECTION OR RERADIATION OF RADIO WAVES; ANALOGOUS ARRANGEMENTS USING OTHER WAVES
- G01S7/00—Details of systems according to groups G01S13/00, G01S15/00, G01S17/00
- G01S7/02—Details of systems according to groups G01S13/00, G01S15/00, G01S17/00 of systems according to group G01S13/00
- G01S7/41—Details of systems according to groups G01S13/00, G01S15/00, G01S17/00 of systems according to group G01S13/00 using analysis of echo signal for target characterisation; Target signature; Target cross-section
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01S—RADIO DIRECTION-FINDING; RADIO NAVIGATION; DETERMINING DISTANCE OR VELOCITY BY USE OF RADIO WAVES; LOCATING OR PRESENCE-DETECTING BY USE OF THE REFLECTION OR RERADIATION OF RADIO WAVES; ANALOGOUS ARRANGEMENTS USING OTHER WAVES
- G01S13/00—Systems using the reflection or reradiation of radio waves, e.g. radar systems; Analogous systems using reflection or reradiation of waves whose nature or wavelength is irrelevant or unspecified
- G01S13/88—Radar or analogous systems specially adapted for specific applications
- G01S13/95—Radar or analogous systems specially adapted for specific applications for meteorological use
-
- Y—GENERAL TAGGING OF NEW TECHNOLOGICAL DEVELOPMENTS; GENERAL TAGGING OF CROSS-SECTIONAL TECHNOLOGIES SPANNING OVER SEVERAL SECTIONS OF THE IPC; TECHNICAL SUBJECTS COVERED BY FORMER USPC CROSS-REFERENCE ART COLLECTIONS [XRACs] AND DIGESTS
- Y02—TECHNOLOGIES OR APPLICATIONS FOR MITIGATION OR ADAPTATION AGAINST CLIMATE CHANGE
- Y02A—TECHNOLOGIES FOR ADAPTATION TO CLIMATE CHANGE
- Y02A90/00—Technologies having an indirect contribution to adaptation to climate change
- Y02A90/10—Information and communication technologies [ICT] supporting adaptation to climate change, e.g. for weather forecasting or climate simulation
Definitions
- This invention relates to a state estimation device for estimating the state of the sea surface.
- Non-Patent Document 1-4 In order to detect tsunamis far away and prompt warning, it is required to predict tsunamis as quickly and accurately as possible. Therefore, it is necessary to know the accurate tsunami peak value and the accurate flow velocity value in real time. As countermeasures, the following inventions exist as prior art. (For example, see Non-Patent Document 1-4 and Patent Document 1.)
- Non-Patent Document 1 uses a nonlinear shallow water equation, which is a tsunami motion model, to predict a peak value from a flow velocity value obtained from a radar in real time.
- the technique disclosed in Non-Patent Document 2 filters the crest value by filtering the crest value in consideration of the observation error from the crest value of a tsunami meter such as a GPS buoy that is grounded from the coast to the offshore. The prediction is performed in real time.
- the technique disclosed in Patent Document 1 estimates a tsunami wave source based on a tsunami wave height observed at an observation position, and estimates a tsunami wave height from the estimated tsunami wave source.
- Non-Patent Document 3 a tsunami detection using an HF radar and a relational expression between a tsunami flow velocity and a wave height by a shallow water equation are shown. Further, the technique disclosed in Non-Patent Document 4 shows a method of assimilating a flow velocity observation value of an HF radar into a tidal model as a method of estimating a flow velocity in a normal time when no tsunami has occurred.
- Non-Patent Document 1 an observation error is included in the flow velocity value by the radar, but this observation error is not taken into consideration. Therefore, there is a problem that the accuracy of wave height prediction deteriorates. There is also a problem that two or more radars are assumed to be installed (it is assumed that a flow velocity vector in a two-dimensional space is obtained).
- Non-Patent Document 2 in order to detect tsunami that may be derived from various directions, it is necessary to install a plurality of tsunami meters in a two-dimensional space. There is a problem that the installation cost is large. In addition, since a tsunami meter such as a GPS buoy cannot directly observe a flow velocity value, there is a problem that a flow velocity value calculated from a wave height value according to a tsunami equation of motion has a large error. Moreover, since a tsunami meter using a GPS buoy cannot be installed offshore far from the land, there is also a problem that only the peak value of the nearby sea is known. The technique disclosed in Patent Document 1 also has the same problem as Non-Patent Document 2 because it is premised that the offshore peak value can be directly observed.
- Non-Patent Document 4 does not describe a tsunami estimation method, nor does it describe a method for estimating the sea surface peak value.
- the present invention has been made to solve the above-described problems, and an object of the present invention is to provide a state estimation device that can know an accurate tsunami wave height value and an accurate flow velocity value in real time.
- the state estimation device includes an observation unit that measures a sea surface velocity value by a beam and a coastal crest value, a flow velocity value for each range cell of the beam, a crest difference between boundaries of the range cell, and a coastal value.
- a prediction unit that predicts the state vector at the next time from the state vector consisting of peak values, a prediction error covariance matrix calculation unit that calculates a prediction error covariance matrix from the smooth error covariance matrix, an observation unit and a prediction error
- a smoothing error for calculating a smoothing error covariance matrix from a processing result by a gain matrix calculating unit for calculating a gain matrix from the processing result by the variance matrix calculating unit and an observation unit, a gain matrix calculating unit, and a prediction error covariance matrix calculating unit
- a wave height difference smoothing unit that smoothes the state vector for each wave height difference, Using the processing result of the parts, that will by adding the peak value and the height difference offshore direction, in which a peak value calculation unit for calculating a peak value for each range cell.
- an accurate tsunami wave height value and an accurate flow velocity value can be known in real time.
- FIG. 1 is a diagram showing a configuration of a state estimation apparatus according to Embodiment 1 of the present invention.
- the state estimation apparatus includes an observation unit 1, an initial peak value calculation unit 2, a smoothing error covariance matrix initial value setting unit 3, a smooth value vector initial value setting unit 4, a prediction unit 5, and a pulse height difference smoothing. 6, a gain matrix calculation unit 7, a smoothing error covariance matrix calculation unit 8, a prediction error covariance matrix calculation unit 9, a peak value calculation unit 10, a likelihood calculation unit 11, and an alarm unit 12.
- each function part of a state estimation apparatus is performed by the program process using CPU based on software.
- Observation unit 1 measures sea surface velocity and coastal wave height values.
- the sea surface velocity value is measured in units of cells with a beam using a radar, and the coastal crest value is measured using a tide gauge installed on the coast.
- the cell represents a unit of distance and beam resolution.
- Information indicating the observed values (flow velocity value and peak value) measured by the observation unit 1 includes an initial peak value calculation unit 2, a smooth vector initial value setting unit 4, a peak difference smoothing unit 6, a gain matrix calculation unit 7, and The result is output to the smoothing error covariance matrix calculation unit 8.
- the initial peak value calculation unit 2 calculates the initial value of the peak difference of the target (sea surface) state vector. At this time, based on the flow velocity value measured by the observation unit 1, the initial peak value calculation unit 2 calculates, for each range cell, the difference between the flow velocity values for two times of the flow velocity value at the initial time and the flow velocity value at the next time. Then, the initial value of the wave height difference of the state vector is calculated. Information indicating the peak height difference calculated by the initial peak value calculation unit 2 is output to the smooth value vector initial value setting unit 4.
- the smoothing error covariance matrix initial value setting unit 3 sets the initial value of the smoothing error covariance matrix at the initial time.
- Information indicating the smoothing error covariance matrix set by the smoothing error covariance matrix initial value setting unit 3 is output to the prediction error covariance matrix calculating unit 9.
- the smooth value vector initial value setting unit 4 sets the observation value measured by the observation unit 1 and the wave height difference calculated by the initial peak value calculation unit 2 as the initial value of the target state vector.
- Information indicating the target state vector set by the smooth value vector initial value setting unit 4 is output to the prediction unit 5.
- the prediction unit 5 performs prediction of a target state vector. At this time, the prediction unit 5 applies the target state vector set by the smoothing vector initial value setting unit 4 in the initial phase to the target state vector smoothed by the wave height difference smoothing unit 6 in the tracking phase. On the other hand, the target state vector at the next time is predicted. Information indicating the target state vector predicted by the prediction unit 5 is output to the wave height difference smoothing unit 6 and the likelihood calculation unit 11.
- the wave height difference smoothing unit 6 is provided for each wave height difference based on the target state vector predicted by the prediction unit 5, the observed value measured by the observation unit 1, and the gain matrix calculated by the gain matrix calculation unit 7.
- the target state vector is smoothed.
- Information indicating the target state vector smoothed by the peak difference smoothing unit 6 is output to the prediction unit 5 and the peak value calculation unit 10.
- the gain matrix calculation unit 7 calculates a gain matrix from the observation values measured by the observation unit 1 and the prediction error covariance matrix calculated by the prediction error covariance matrix calculation unit 9. Information indicating the gain matrix calculated by the gain matrix calculating unit 7 is output to the wave height difference smoothing unit 6 and the smoothing error covariance matrix calculating unit 8.
- the smoothing error covariance matrix calculation unit 8 includes an observation value measured by the observation unit 1, a gain matrix calculated by the gain matrix calculation unit 7, and a prediction error covariance matrix calculated by the prediction error covariance matrix calculation unit 9. From this, a smooth error covariance matrix is calculated. Information indicating the smoothing error covariance matrix calculated by the smoothing error covariance matrix calculating unit 8 is output to the prediction error covariance matrix calculating unit 9.
- the prediction error covariance matrix calculation unit 9 calculates a prediction error covariance matrix. At this time, the prediction error covariance matrix calculation unit 9 uses the smooth error covariance matrix calculated by the smoothing error covariance matrix initial value setting unit 3 in the initialization phase, and the smoothing error covariance matrix calculation unit in the tracking phase. The prediction error covariance matrix is calculated using the smooth error covariance matrix calculated in step 8. Information indicating the prediction error covariance matrix calculated by the prediction error covariance matrix calculation unit 9 is output to the gain matrix calculation unit 7, the smooth error covariance matrix calculation unit 8, and the likelihood calculation unit 11.
- the crest value calculation unit 10 calculates a crest value for each range cell by adding the crest value and the crest difference in the offshore direction based on the target state vector smoothed by the crest difference smoothing unit 6. It is. Information indicating the peak value calculated by the peak value calculation unit 10 and the flow velocity value of the smoothed target state vector at that time is output to the alarm unit 12.
- the likelihood calculation unit 11 calculates the likelihood for each beam from the target state vector predicted by the prediction unit 5 and the prediction error covariance matrix calculated by the prediction error covariance matrix calculation unit 9, and the likelihood The beam direction with the maximum degree is detected as the arrival direction of the tsunami. Information indicating the beam direction detected by the likelihood calculating unit 11 is output to the alarm unit 12.
- the warning unit 12 outputs a tsunami warning based on the beam direction detected by the likelihood calculation unit 11, the wave height value calculated by the wave height value calculation unit 10, and the magnitude of the flow velocity value at that time.
- the alarm unit 12 predicts the arrival time of the tsunami by performing a fluid simulation using the equation of motion of the tsunami using the peak value calculated by the peak value calculation unit 10 and the flow velocity value at that time as an initial condition. .
- the operation of the state estimation device is divided into an initialization phase and a tracking phase.
- the observation values flow velocity value, peak value
- the initial value setting of the target state vector and the initial value setting of the smoothing error covariance matrix are executed.
- the tracking phase the state vector is updated online at each time as in the normal Kalman filter.
- FIG. 2 is a diagram illustrating measurement of the sea surface velocity value using a radar.
- the initial peak value calculation unit 2 calculates the initial value of the peak difference of the target state vector based on the flow velocity value measured by the observation unit 1.
- the flow velocity value obtained by one radar can detect only the flow velocity component in the line-of-sight direction
- a Kalman filter is formed considering the interaction of waves in the range direction for each beam. For this reason, as shown in FIG. 3 and the following equation (1), the flow velocity value (flow rate) in each range cell from the coast to the offshore, the spatial height difference between the boundary of the range cell, and the coastal peak value are the target states.
- a Kalman filter is executed for each beam as a vector.
- X (k) is the target state vector at time k
- M i (k) is the flow rate of the i-th range cell at time k
- ⁇ 0 (k) is the coastal peak value
- ⁇ i ⁇ 1 is the coastal peak value
- I (k) represents the wave height difference of the i-th range cell.
- the following processing shows the Kalman filter processing when the beam is fixed, but it is actually assumed that Kalman filters are executed in parallel for the number of beams.
- the initial peak value calculation unit 2 receives the flow velocity value for two hours of the flow velocity value at the initial time and the flow velocity value at the next time, and uses the following equations (2) and (3), The initial value of the wave height difference of the state vector is calculated from the difference between the initial time and the flow velocity value at the next time.
- u i is the flow velocity of the i-th range cell
- t is time
- h water depth
- d is the number of range cells
- g gravity acceleration
- ⁇ x is the distance between range cells
- ⁇ t is the sampling interval.
- the water depth h is known from topographic data or the like.
- the smoothing error covariance matrix initial value setting unit 3 sets the smoothing error covariance matrix at the initial time using the following equations (4) to (7).
- P k: k is a smoothing error covariance matrix at time k
- R is an observation error covariance matrix
- Q is a driving noise covariance matrix
- G is expressed by Expressions (5) to (7).
- a driving noise conversion matrix, Id represents a unit matrix having a size of d ⁇ d.
- the smooth value vector initial value setting unit 4 sets the observation value measured by the observation unit 1 and the wave height difference calculated by the initial peak value calculation unit 2 as the initial value of the target state vector.
- the prediction unit 5 performs prediction of a target state vector.
- the prediction unit 5 applies the target state vector set by the smoothing vector initial value setting unit 4 in the initial phase to the target state vector smoothed by the wave height difference smoothing unit 6 in the tracking phase.
- the target state vector at the next time is predicted using the following equations (8) to (10).
- equation (8) X k: k ⁇ 1 represents the predicted state vector at time k.
- a transition matrix A having a size of (2d + 1) ⁇ (2d + 1) representing the motion model is expressed by equations (9) and (10).
- transition matrix A Details of the transition matrix A will be described below. First, from the one-dimensional linear shallow water equation, the relationship between the flow velocity value and the wave height difference can be expressed by the following equations (11) and (12).
- the transition matrix A representing the motion model is expressed by the equations (9) and (10).
- the motion model is derived from the one-dimensional linear shallow water equation.
- smoothing may be performed using a nonlinear filter such as an extended Kalman filter or a particle filter by using a shallow water equation considering a nonlinear term.
- the driving noise may be set according to the depth of the water depth in the measurement region by the observation unit 1.
- the wave height difference smoothing unit 6 is based on the target state vector predicted by the prediction unit 5, the observed value measured by the observation unit 1, and the gain matrix calculated by the gain matrix calculation unit 7.
- the target state vector is smoothed for each wave height difference.
- X k: k is a smoothed state vector at time k
- H is an observation matrix having a size of d ⁇ (2d + 1) represented by Equation (15) from the relationship between the flow rate and the flow velocity
- Z (K) represents an observed value vector of the crest value and the flow velocity value on the coast.
- the gain matrix calculation unit 7 uses the following equation (16) from the observation value measured by the observation unit 1 and the prediction error covariance matrix calculated by the prediction error covariance matrix calculation unit 9 to obtain a gain matrix. Is calculated.
- H t represents the transpose of the matrix H.
- the smoothing error covariance matrix calculation unit 8 includes the observation values measured by the observation unit 1, the gain matrix calculated by the gain matrix calculation unit 7, and the prediction error covariance calculated by the prediction error covariance matrix calculation unit 9.
- a smooth error covariance matrix is calculated from the variance matrix using the following equation (17).
- the prediction error covariance matrix calculation unit 9 calculates a prediction error covariance matrix using the following equation (18). At this time, the prediction error covariance matrix calculation unit 9 uses the smooth error covariance matrix calculated by the smoothing error covariance matrix initial value setting unit 3 in the initialization phase, and calculates the smoothing error covariance matrix in the tracking phase. The prediction error covariance matrix is calculated using the smooth error covariance matrix calculated by the unit 8. In Equation (18), P k: k ⁇ 1 represents a prediction error covariance matrix at time k ⁇ 1.
- the crest value calculation unit 10 adds the crest value and the crest difference in the offshore direction using the following equation (19) based on the target state vector smoothed by the crest difference smoothing unit 6. Then, the peak value for each range cell is calculated (see FIG. 4).
- the likelihood calculation unit 11 calculates the following equations (20) and (21) from the target state vector predicted by the prediction unit 5 and the prediction error covariance matrix calculated by the prediction error covariance matrix calculation unit 9. Is used to calculate the likelihood for each beam, and the beam direction with the maximum likelihood is detected as the arrival direction of the tsunami.
- P represents a probability density function, and is a normal distribution represented by Expression (21).
- N (A, b) represents a normal distribution with an average A and a variance b.
- the warning unit 12 outputs a tsunami warning from the beam direction detected by the likelihood calculation unit 11, the wave height value calculated by the wave height calculation unit 10, and the magnitude of the flow velocity value at that time.
- the arrival time of the tsunami is predicted by performing a fluid simulation using the tsunami equation of motion, with the peak value calculated by the peak value calculation unit 10 and the flow velocity value at that time as initial conditions.
- the crest value is decomposed into the crest difference in the spatial direction, and the flow velocity value (flow rate), the crest value on the coast and the crest difference are smoothed in the time series direction as the target state vector.
- the smoothed wave height value and wave height difference are added in the offshore direction, the accurate wave height value and accurate flow velocity value of the tsunami can be known in real time.
- Non-Patent Document 2 when trying to estimate the offshore peak value, the flow velocity value can be observed, but the peak value other than the coastal peak value cannot be obtained directly. The peak value cannot be estimated accurately.
- FIG. 5 is a diagram showing the propagation of the flow rate (flow velocity value) and the crest value, and the concept of smoothing offshore crest value in the conventional method.
- an arrow in a diamond formed from the peak value and the flow rate represents a state of propagation of the peak value and the flow rate according to the one-dimensional shallow water equation.
- the initial peak value calculated using the equations (2) and (19) is smoothed using the ambient flow rate as the initial value of the Kalman filter. Therefore, the influence of the error of the initial peak value is large, and even if smoothing is performed in time series, the estimation accuracy of the peak value is not improved.
- the flow rate used for smoothing offshore peak values is limited to the flow rate in the triangular region of FIG.
- the crest value is decomposed into a crest difference in the spatial direction, smoothed for each crest difference, and the crest value is obtained by addition processing of the crest difference in the spatial direction.
- FIG. 6 is a diagram showing the concept of smoothing the wave height difference in the present invention.
- the flow rate (flow velocity value) used for smoothing the nearest wave height difference from the coastal wave height value is the flow rate within the broken line in the figure.
- the offshore peak value is represented by the addition of the peak difference
- the number of flow rates used in the calculation of the offshore peak value is the number of flow rates at all past times of all range cells. From this, it can be expected that the estimation accuracy of the crest value is improved as compared with the conventional method.
- any component of the embodiment can be modified or any component of the embodiment can be omitted within the scope of the invention.
- the state estimation device can be used to know an accurate tsunami wave height value and an accurate flow velocity value in real time, and is suitable for use in a state estimation device that estimates the state of the sea surface.
Landscapes
- Engineering & Computer Science (AREA)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Radar, Positioning & Navigation (AREA)
- Remote Sensing (AREA)
- Life Sciences & Earth Sciences (AREA)
- Hydrology & Water Resources (AREA)
- Computer Networks & Wireless Communication (AREA)
- Radar Systems Or Details Thereof (AREA)
Abstract
Description
また、非特許文献2に開示された技術は、沿岸から沖合に向かって接地されたGPSブイ等の津波計の波高値から観測誤差を考慮して波高値のフィルタリングを行うことで、波高値の予測をリアルタイムに行うものである。
また、特許文献1に開示された技術は、観測位置において観測される津波波高に基づいて津波波源を推定し、推定された津波波源から津波波高を推定するものである。
また、非特許文献4に開示された技術では、津波が起きていない平常時の流速の推定方法として、HFレーダの流速観測値を潮汐のモデルに同化する方法が示されている。
また、特許文献1に開示された技術についても、沖合の波高値を直接観測できることが前提となっているため、非特許文献2と同様の課題がある。
また、非特許文献4では津波の推定方式について記載がなく、また、海面の波高値の推定方法についても記載がない。
実施の形態1.
図1はこの発明の実施の形態1に係る状態推定装置の構成を示す図である。
状態推定装置は、図1に示すように、観測部1、初期波高値算出部2、平滑誤差共分散行列初期値設定部3、平滑値ベクトル初期値設定部4、予測部5、波高差平滑部6、ゲイン行列算出部7、平滑誤差共分散行列算出部8、予測誤差共分散行列算出部9、波高値算出部10、尤度算出部11及び警報部12から構成されている。なお、状態推定装置の各機能部は、ソフトウェアに基づくCPUを用いたプログラム処理によって実行される。
状態推定装置の動作は、初期化フェーズと追尾フェーズとに分けられる。初期化フェーズでは、2時刻分の観測値(流速値、波高値)を蓄積した上で、目標の状態ベクトルの初期値設定及び平滑誤差共分散行列の初期値設定を実行することで、カルマンフィルタの初期値を設定する。また、追尾フェーズでは、通常のカルマンフィルタと同様に、時刻毎に状態ベクトルの更新をオンラインで実行する。
式(1)において、X(k)は時刻kにおける目標の状態ベクトル、Mi(k)は時刻kにおけるi番目のレンジセルの流量、η0(k)は沿岸の波高値、Δηi-1,i(k)はi番目のレンジセルの波高差を表す。
式(2),(3)において、uiはi番目のレンジセルの流速、tは時間、hは水深、dはレンジセル数、gは重力加速度、Δxはレンジセル間距離、Δtはサンプリング間隔を表す。また、水深hは、地形データ等から既知であると仮定する。
式(4)において、Pk:kは時刻kにおける平滑誤差共分散行列、Rは観測誤差共分散行列、Qは駆動雑音共分散行列、Gは式(5)~(7)で表される駆動雑音変換行列、Idはd×dのサイズの単位行列を表す。ここで、式(5)~(7)では、津波が運動する際に波高差が正規分布に従って揺らぐことを想定しているが、波高差の時間微分が正規分布に従って揺らぐとして定式化を行ってもよい。
式(8)において、Xk:k-1は時刻kにおける予測された状態ベクトルを表す。また、運動モデルを表す(2d+1)×(2d+1)のサイズの遷移行列Aは式(9),(10)で表される。
式(14)において、Xk:kは時刻kにおける平滑化された状態ベクトル、Hは流量と流速との関係から式(15)で表されるd×(2d+1)のサイズの観測行列、Z(k)は沿岸の波高値と流速値の観測値ベクトルを表す。
式(16)において、Htは行列Hの転置を表す。
式(18)において、Pk:k-1は時刻k-1における予測誤差共分散行列を表す。
式(20)において、Pは確率密度関数を表し、式(21)で表される正規分布とする。また、式(21)において、N(A,b)は平均A、分散bの正規分布を表す。
従来方式では、カルマンフィルタの初期値として、式(2),(19)を用いて算出した初期波高値を、周囲の流量を用いて平滑化している。そのため、初期波高値の誤差の影響が大きく、時系列に平滑化を実施しても波高値の推定精度は改善されない。これに加え、沖合の波高値の平滑化に利用される流量が、図5の三角形の領域の流量に限られるため、波高値の推定精度が劣化する。
この図6に示すように、沿岸の波高値から一番近傍の波高差の平滑化に使用される流量(流速値)は、図中の破線内の流量である。そして、沖合の波高値は波高差の加算で表されることから、沖合の波高値の算出の際に利用される流量数は、全レンジセルの過去の全時刻の流量数となる。このことから、従来方式よりも波高値の推定精度が向上することが期待できる。
Claims (7)
- ビームによる海面の流速値の計測、及び沿岸の波高値の計測を行う観測部と、
前記ビームのレンジセル毎の流速値、当該レンジセルの境界間の波高差及び沿岸の波高値から成る状態ベクトルに対し、次時刻における当該状態ベクトルを予測する予測部と、
平滑誤差共分散行列から予測誤差共分散行列を算出する予測誤差共分散行列算出部と、
前記観測部及び前記予測誤差共分散行列算出部による処理結果から、ゲイン行列を算出するゲイン行列算出部と、
前記観測部、前記ゲイン行列算出部及び前記予測誤差共分散行列算出部による処理結果から、平滑誤差共分散行列を算出する平滑誤差共分散行列算出部と、
前記観測部、前記予測部及び前記ゲイン行列算出部による処理結果から、波高差毎に前記状態ベクトルを平滑化する波高差平滑部と、
前記波高差平滑部による処理結果を用い、沖合方向に波高値と波高差を加算していくことで、前記レンジセル毎の波高値を算出する波高値算出部と
を備えた状態推定装置。 - 前記予測部は、浅水方程式による運動モデルを用い、次時刻における前記状態ベクトルを予測する
ことを特徴とする請求項1記載の状態推定装置。 - 前記予測部及び前記予測誤差共分散行列算出部による処理結果から、前記ビーム毎の尤度を算出し、当該尤度が高いビーム方向を津波の到来方向として検出する尤度算出部を備えた
ことを特徴とする請求項1記載の状態推定装置。 - 前記尤度算出部により検出されたビーム方向、前記波高値算出部により算出された波高値、前記波高差平滑部により平滑化された前記状態ベクトルの流速値から、津波の到来時間を予測する警報部を備えた
ことを特徴とする請求項1記載の状態推定装置。 - 前記観測部により計測された流速値に基づいて、前記レンジセル毎に、初期時刻における流速値と次時刻における流速値の2時刻分の流速値との差から、前記状態ベクトルにおける波高差の初期値を算出する初期波高値算出部を備え、
前記予測部は、初期フェーズでは、前記観測部により観測された前記レンジセル毎の流速値及び沿岸の波高値、及び前記初期波高値算出部により算出された前記レンジセルの境界間の波高差から成る状態ベクトルから、次時刻における当該状態ベクトルを予測する
ことを特徴とする請求項1記載の状態推定装置。 - 前記予測部は、非線形浅水方程式による運動モデルを用い、次時刻における前記状態ベクトルを予測する
ことを特徴とする請求項1記載の状態推定装置。 - 前記予測部は、前記浅水方程式による運動モデルにおいて、前記観測部による計測領域内での水深に応じて、駆動雑音の大きさを設定する
ことを特徴とする請求項2記載の状態推定装置。
Priority Applications (5)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
JP2016559757A JP6132990B2 (ja) | 2014-11-20 | 2014-11-20 | 状態推定装置 |
US15/525,154 US9964408B2 (en) | 2014-11-20 | 2014-11-20 | State estimation device |
PCT/JP2014/080773 WO2016079848A1 (ja) | 2014-11-20 | 2014-11-20 | 状態推定装置 |
DE112014007193.5T DE112014007193B4 (de) | 2014-11-20 | 2014-11-20 | Zustandsschätzvorrichtung |
MYPI2017701819A MY167374A (en) | 2014-11-20 | 2014-11-20 | State estimation device |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
PCT/JP2014/080773 WO2016079848A1 (ja) | 2014-11-20 | 2014-11-20 | 状態推定装置 |
Publications (1)
Publication Number | Publication Date |
---|---|
WO2016079848A1 true WO2016079848A1 (ja) | 2016-05-26 |
Family
ID=56013454
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
PCT/JP2014/080773 WO2016079848A1 (ja) | 2014-11-20 | 2014-11-20 | 状態推定装置 |
Country Status (5)
Country | Link |
---|---|
US (1) | US9964408B2 (ja) |
JP (1) | JP6132990B2 (ja) |
DE (1) | DE112014007193B4 (ja) |
MY (1) | MY167374A (ja) |
WO (1) | WO2016079848A1 (ja) |
Cited By (5)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN106970373A (zh) * | 2017-05-10 | 2017-07-21 | 公安部第三研究所 | 基于水面状态连续成像系统的小波浪浪高提取方法 |
CN108549961A (zh) * | 2018-05-02 | 2018-09-18 | 河海大学 | 一种基于cmip5预估海浪有效波高的方法 |
JP6641532B1 (ja) * | 2019-01-24 | 2020-02-05 | 三菱電機株式会社 | 状態予測装置および状態予測方法 |
WO2020071327A1 (ja) * | 2018-10-01 | 2020-04-09 | 東電設計株式会社 | 津波予測装置、方法、及びプログラム |
US11035953B2 (en) * | 2016-08-25 | 2021-06-15 | Mitsubishi Electric Corporation | Radar apparatus |
Families Citing this family (3)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US11425395B2 (en) * | 2013-08-20 | 2022-08-23 | Google Llc | Encoding and decoding using tiling |
FR3082666B1 (fr) * | 2018-06-19 | 2021-05-21 | Thales Sa | Procede de mesure de la hauteur de vagues a l'aide d'un radar aeroporte |
CN111457901A (zh) * | 2020-03-31 | 2020-07-28 | 北京航天广通科技有限公司分公司 | 海浪高度的检测方法、装置、设备及存储介质 |
Citations (7)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
JP2001059869A (ja) * | 1999-08-24 | 2001-03-06 | Oki Electric Ind Co Ltd | レーダ受信信号の比例尺度による物理量観測システム |
JP2003185743A (ja) * | 2001-12-19 | 2003-07-03 | Yokohama Tlo Co Ltd | 波高算出装置、波高算出方法、記録媒体、及び船舶 |
JP2003307563A (ja) * | 2002-04-16 | 2003-10-31 | Michio Nobeoka | レーダ波高計測装置、レーダ波高補正係数作成方法、及びレーダ波高補正係数作成装置 |
JP2005003611A (ja) * | 2003-06-13 | 2005-01-06 | Japan Radio Co Ltd | レーダ波浪観測装置 |
JP2009229424A (ja) * | 2008-03-25 | 2009-10-08 | Mitsubishi Electric Corp | 津波監視装置 |
US20100315284A1 (en) * | 2009-09-02 | 2010-12-16 | Trizna Dennis B | Method and apparatus for coherent marine radar measurements of properties of ocean waves and currents |
JP2011064677A (ja) * | 2009-08-21 | 2011-03-31 | Univ Of Tokyo | 水面形状計測装置、及び水面形状計測方法 |
Family Cites Families (1)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
JP5007391B2 (ja) | 2006-09-29 | 2012-08-22 | 国立大学法人京都大学 | 津波波源推定方法及び津波波高予測方法並びにその関連技術 |
-
2014
- 2014-11-20 JP JP2016559757A patent/JP6132990B2/ja active Active
- 2014-11-20 WO PCT/JP2014/080773 patent/WO2016079848A1/ja active Application Filing
- 2014-11-20 DE DE112014007193.5T patent/DE112014007193B4/de active Active
- 2014-11-20 US US15/525,154 patent/US9964408B2/en active Active
- 2014-11-20 MY MYPI2017701819A patent/MY167374A/en unknown
Patent Citations (7)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
JP2001059869A (ja) * | 1999-08-24 | 2001-03-06 | Oki Electric Ind Co Ltd | レーダ受信信号の比例尺度による物理量観測システム |
JP2003185743A (ja) * | 2001-12-19 | 2003-07-03 | Yokohama Tlo Co Ltd | 波高算出装置、波高算出方法、記録媒体、及び船舶 |
JP2003307563A (ja) * | 2002-04-16 | 2003-10-31 | Michio Nobeoka | レーダ波高計測装置、レーダ波高補正係数作成方法、及びレーダ波高補正係数作成装置 |
JP2005003611A (ja) * | 2003-06-13 | 2005-01-06 | Japan Radio Co Ltd | レーダ波浪観測装置 |
JP2009229424A (ja) * | 2008-03-25 | 2009-10-08 | Mitsubishi Electric Corp | 津波監視装置 |
JP2011064677A (ja) * | 2009-08-21 | 2011-03-31 | Univ Of Tokyo | 水面形状計測装置、及び水面形状計測方法 |
US20100315284A1 (en) * | 2009-09-02 | 2010-12-16 | Trizna Dennis B | Method and apparatus for coherent marine radar measurements of properties of ocean waves and currents |
Cited By (12)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US11035953B2 (en) * | 2016-08-25 | 2021-06-15 | Mitsubishi Electric Corporation | Radar apparatus |
CN106970373A (zh) * | 2017-05-10 | 2017-07-21 | 公安部第三研究所 | 基于水面状态连续成像系统的小波浪浪高提取方法 |
CN106970373B (zh) * | 2017-05-10 | 2019-09-03 | 公安部第三研究所 | 基于水面状态连续成像系统的小波浪浪高提取方法 |
CN108549961A (zh) * | 2018-05-02 | 2018-09-18 | 河海大学 | 一种基于cmip5预估海浪有效波高的方法 |
CN108549961B (zh) * | 2018-05-02 | 2021-10-15 | 河海大学 | 一种基于cmip5预估海浪有效波高的方法 |
WO2020071327A1 (ja) * | 2018-10-01 | 2020-04-09 | 東電設計株式会社 | 津波予測装置、方法、及びプログラム |
JP2020056649A (ja) * | 2018-10-01 | 2020-04-09 | 東電設計株式会社 | 津波予測装置、方法、及びプログラム |
US20210396515A1 (en) * | 2018-10-01 | 2021-12-23 | Tokyo Electric Power Services Co., Ltd. | Tsunami prediction device, method and computer-readable storage medium |
JP7156613B2 (ja) | 2018-10-01 | 2022-10-19 | 東電設計株式会社 | 津波予測装置、方法、及びプログラム |
TWI834735B (zh) * | 2018-10-01 | 2024-03-11 | 日商東電設計股份有限公司 | 海嘯預測裝置、方法、及記錄媒體 |
JP6641532B1 (ja) * | 2019-01-24 | 2020-02-05 | 三菱電機株式会社 | 状態予測装置および状態予測方法 |
WO2020152824A1 (ja) * | 2019-01-24 | 2020-07-30 | 三菱電機株式会社 | 状態予測装置および状態予測方法 |
Also Published As
Publication number | Publication date |
---|---|
JP6132990B2 (ja) | 2017-05-24 |
DE112014007193T5 (de) | 2017-08-24 |
US20170314920A1 (en) | 2017-11-02 |
DE112014007193B4 (de) | 2021-09-23 |
JPWO2016079848A1 (ja) | 2017-04-27 |
MY167374A (en) | 2018-08-16 |
US9964408B2 (en) | 2018-05-08 |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
JP6132990B2 (ja) | 状態推定装置 | |
JP5905646B2 (ja) | 津波監視システム | |
CN104035083B (zh) | 一种基于量测转换的雷达目标跟踪方法 | |
CN102221448B (zh) | 一种确定海上搜救区域的方法、系统及搜救模拟器 | |
Alford et al. | A real-time system for forecasting extreme waves and vessel motions | |
JP6521777B2 (ja) | 津波監視システム | |
CN103389077B (zh) | 一种基于mbes的海底沙波地貌运动探测方法 | |
JP5007391B2 (ja) | 津波波源推定方法及び津波波高予測方法並びにその関連技術 | |
CN107607092A (zh) | 一种基于无人机遥感的波浪与浮体运动预报系统 | |
JP6440912B2 (ja) | レーダ装置 | |
Brakenhoff et al. | Spatio‐temporal characteristics of small‐scale wave–current ripples on the Ameland ebb‐tidal delta | |
Nielsen | Response-based estimation of sea state parameters—influence of filtering | |
JP7156613B2 (ja) | 津波予測装置、方法、及びプログラム | |
Rosman et al. | Extracting Reynolds stresses from acoustic Doppler current profiler measurements in wave-dominated environments | |
WO2020152824A1 (ja) | 状態予測装置および状態予測方法 | |
JP6340067B2 (ja) | 表層流推定装置、表層流推定システム、海洋モデル推定装置、及び危険度判定装置 | |
Stone et al. | Ship Shoal as a prospective borrow site for barrier island restoration, coastal south-central Louisiana, USA: numerical wave modeling and field measurements of hydrodynamics and sediment transport | |
Popielarczyk | RTK water level determination in precise inland bathymetric measurements | |
Amador et al. | ADCP bias and Stokes drift in AUV-based velocity measurements | |
JP6610898B2 (ja) | 横メタセンタ高さ推定装置及び横メタセンタ高さ推定方法 | |
Le Coz et al. | Video-based hydrometry: a Bayesian camera calibration method for uncertainty analysis | |
WO2024189755A1 (ja) | 推定装置、推定システムおよび推定方法 | |
Mil-Homens et al. | Re-assessment and update of bulk longshore sediment transport formulas | |
Maxim et al. | Comparison of Models and Volumetric Determination for Catusa Lake, Galati | |
CN116449388A (zh) | 一种利用机载激光雷达海面点云反演水深的方法 |
Legal Events
Date | Code | Title | Description |
---|---|---|---|
121 | Ep: the epo has been informed by wipo that ep was designated in this application |
Ref document number: 14906379 Country of ref document: EP Kind code of ref document: A1 |
|
ENP | Entry into the national phase |
Ref document number: 2016559757 Country of ref document: JP Kind code of ref document: A |
|
WWE | Wipo information: entry into national phase |
Ref document number: 15525154 Country of ref document: US |
|
WWE | Wipo information: entry into national phase |
Ref document number: 112014007193 Country of ref document: DE |
|
122 | Ep: pct application non-entry in european phase |
Ref document number: 14906379 Country of ref document: EP Kind code of ref document: A1 |