CN107357170A - A kind of Wave Model Forecasting Methodology based on active disturbance rejection state observer - Google Patents
A kind of Wave Model Forecasting Methodology based on active disturbance rejection state observer Download PDFInfo
- Publication number
- CN107357170A CN107357170A CN201710576846.3A CN201710576846A CN107357170A CN 107357170 A CN107357170 A CN 107357170A CN 201710576846 A CN201710576846 A CN 201710576846A CN 107357170 A CN107357170 A CN 107357170A
- Authority
- CN
- China
- Prior art keywords
- wave
- frequency components
- mode
- observer
- monochromatic
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Pending
Links
Classifications
-
- G—PHYSICS
- G05—CONTROLLING; REGULATING
- G05B—CONTROL OR REGULATING SYSTEMS IN GENERAL; FUNCTIONAL ELEMENTS OF SUCH SYSTEMS; MONITORING OR TESTING ARRANGEMENTS FOR SUCH SYSTEMS OR ELEMENTS
- G05B13/00—Adaptive control systems, i.e. systems automatically adjusting themselves to have a performance which is optimum according to some preassigned criterion
- G05B13/02—Adaptive control systems, i.e. systems automatically adjusting themselves to have a performance which is optimum according to some preassigned criterion electric
- G05B13/04—Adaptive control systems, i.e. systems automatically adjusting themselves to have a performance which is optimum according to some preassigned criterion electric involving the use of models or simulators
- G05B13/048—Adaptive control systems, i.e. systems automatically adjusting themselves to have a performance which is optimum according to some preassigned criterion electric involving the use of models or simulators using a predictor
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01W—METEOROLOGY
- G01W1/00—Meteorology
- G01W1/10—Devices for predicting weather conditions
-
- 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
Landscapes
- Engineering & Computer Science (AREA)
- Environmental & Geological Engineering (AREA)
- Life Sciences & Earth Sciences (AREA)
- Atmospheric Sciences (AREA)
- Biodiversity & Conservation Biology (AREA)
- Ecology (AREA)
- Environmental Sciences (AREA)
- Health & Medical Sciences (AREA)
- Artificial Intelligence (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Evolutionary Computation (AREA)
- Medical Informatics (AREA)
- Software Systems (AREA)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Automation & Control Theory (AREA)
- Feedback Control In General (AREA)
Abstract
The present invention relates to a kind of Wave Model Forecasting Methodology based on active disturbance rejection state observer,Belong to the technical field of data processing and Predicting Technique,Fast Fourier Transform (FFT) is carried out to the heave displacement signal in a measured measuring section first,Obtain amplitude and phase of each monochromatic wave to induction signal,By magnitude peak detection and identification is gone out heave movement major frequency components quantity and each mode corresponding to harmonic parameters,Design Kalman's observer,Each mode monochromatic wave of each the mode monochromatic wave and major frequency components of measurement is observed using Kalman's observer,Harmonic prediction parameter corresponding to each mode monochromatic wave of On-line Estimation renewal major frequency components,Heave movement in the standby following certain predicted time section of extra large frock is synthesized based on Prediction Parameters prediction,And then phase correction compensation is carried out to wave compensation system in advance,Prediction effect is comprehensive,It is rigorous accurate.
Description
Technical field
The present invention relates to a kind of sea based on FFT, active disturbance rejection state observer and Parameter Adaptive Compensation
Unrestrained model prediction method, belong to the technical field of data processing and Predicting Technique.
Background technology
For the active-passive composite formula heave compensation system standby applied to extra large frock, it is necessary in real time accurately detection lash ship or
The heave movement posture of towed body, then the Active Compensation towed body displacement of opposite direction or cable tension.Typically use motion sensor
(MRU or IMU etc.) detects the exercise data of ship, and the sensor is made up of accelerometer, gyroscope, can measure ship horizontal stroke
Shake, pitching and heave data, but its output of surging (heave displacement) is to be obtained by acceleration through quadratic integral and filtering process, is deposited
In certain time lag and larger accuracy error.Further, since heave compensator is a large inertia system, system response is deposited
In certain time delay phenomenon, the compensation performance of active wave heave compensation system can be had a strong impact on, or even causes compensation system to lose
Surely.However, the measurement time delay of motion sensor and the dynamic lag dead time of compensation system are relative constancies and can be with
Pick out and in advance, precondition is provided for the time lag correction of heave compensation system.In addition, when although the waveform of wave is presented
The scrambling of change, but the standby heave movement of extra large frock is not confusing, but motion and sea situation depending on ship,
So we may calculate the motion of loaded equipment, it might even be possible to enter in the case where not knowing any ship attribute completely
Row short-term forecast.
Usual Forecasting Methodology can be divided into qualitative forecasting method, time series forecasting and Causal model predicted method.Traditional is pre-
Survey method generally approx realizes the prediction to the object rule of development by establishing system linearity model.Such as the most frequently used AR
Model prediction, exactly smoothly assume in time series on basis, linear model is established to it, then using model extrapolation
Method predicts its future value.This method is only applicable to the prediction to stationary time series.But actual time series is very
Unstable, traditional Forecasting Methodology can not often obtain preferable prediction effect.Heave movement prediction side of the ship in wave
Method is broadly divided into based on hydrodynamics method and based on non-aqueous dynamic method.Forecasting Methodology based on hydrodynamics mainly uses
The methods of convolution, Kalman filter, neural network, Power estimation meter method mainly had based on non-aqueous dynamic (dynamical) Forecasting Methodology
And analytic approach based on time series etc..It is larger using convolution method amount of calculation because ships data is more, therefore in actual work
Being applied in journey system has certain limitation.Kalman filtering method amount of calculation is small and precision of prediction is higher, but needs accurate
Ship motion state equation, and its precision of prediction is influenceed by wave frequencies and sea situation, therefore Kalman Prediction is straight in practice
Scoop out with being inappropriate.
Chinese patent document (application number CN201410290745.6) discloses a kind of ocean wave based on arma modeling
Wave height Forecasting Methodology；The prediction object of this document is wave, and method is based on arma modeling, and ocean wave power generation system is optimized and controlled
System, but the program is only predicted to sea wave height.
The content of the invention
Goal of the invention：In order to overcome the deficiencies in the prior art, the present invention provides a kind of based on fast Flourier change
Change, the Wave Model Forecasting Methodology of active disturbance rejection state observer and Parameter Adaptive Compensation, with forecast with unerring accuracy Unequal time lag and
Ship heave movement under different sea situations, it can be used for MRU detections and Actuator dynamic time lag in the control of follow-up heave compensation
Dynamic compensates in real time.
Term is explained
Kalman filter：Being one, " optimal recursive data processing algorithm are (optimal
Change autoregression data processing algorithm) ".
FFT (Fast Fourier Transformation), as FFT, is DFT
(DFT) fast algorithm.
To achieve the above object, the technical solution adopted by the present invention is：
A kind of Wave Model prediction based on FFT, active disturbance rejection state observer and Parameter Adaptive Compensation
Method, comprise the following steps：
(1) heave movement is directed to, the heave displacement signal in a measuring section, heave are gathered using measurement sensor
Displacement signal is to be combined under different time signal by multiple monochromatic waves, and quick Fu is carried out to measured heave displacement signal
In leaf transformation, obtain amplitude and phase that each monochromatic wave corresponds to time signal；
(2) the major frequency components quantity for going out heave movement to magnitude peak detection and identification and each mode monochromatic wave institute are passed through
Corresponding harmonic parameters；
(3) Kalman's observer, each mode using Kalman's observer to the major frequency components in step (2) are designed
Monochromatic wave is observed, and draws main observation；Each mode monochromatic wave of measurement in step (1) is entered using Kalman's observer
Row observation, draws measurement observation, according to the deviation of measurement observation and main observation, On-line Estimation renewal main frequency into
The harmonic prediction parameter corresponding to each mode monochromatic wave divided；
(4) the harmonic prediction parameter corresponding to each mode monochromatic wave based on the major frequency components obtained by step (3), in advance
Survey the heave movement synthesized in the standby following certain predicted time section of extra large frock.
According to currently preferred, in step (1), measuring section Δ TFFTValue be 5min-10min.In the range of being somebody's turn to do
Value can ensure will prediction control errors maximum heave amplitude 20% within the premise of, reduce amount of calculation.
According to currently preferred, in step (1), heave movement is expressed as to the superposition of n monochromatic wave：
N be monochromatic wave mode number, Ai、fi、It is amplitude, frequency and the phase corresponding to the monochromatic wave of i-th of mode
Position, h (t) are the initial value after monochromatic wave superposition.
According to currently preferred, described to be detected as in step (2), on the basis of maximum amplitude peak value, magnitude peak is big
In the quantity for being major frequency components, that is, drawing the major frequency components of heave movement of maximum amplitude peak value 80%；Harmonic wave
Parameter includes amplitude, frequency and phase.Calculating time interval Δ TFFTInterior, the change of non-principal frequency content can be ignored, non-
Error caused by major frequency components change can be in the error feedback parameter self-adjusting link of subsequent step (3) state observer
Automatically compensate.
It is further preferred that in step (2), the ODE of each mode monochromatic wave of major frequency components is
N ' be detection after major frequency components monochromatic wave quantity, t0For the initial time of selected measuring section, xi
For wiExpression formula after Fast Fourier Transform (FFT), CiFor wiCorresponding xiCoefficient,For xiDifferential form；
System primary condition is
According to currently preferred, in step (3), state observer model is
Formula 4. in x be all xiThe matrix of composition；
Initialization function isx1,0For
One major frequency components is in t0The value at moment；
xoff,0For t0Moment x (t0) one biasing, be the constant value obtained by state observer；C(t0) it is all Ci's
Matrix；
State observer model is discrete to be turned to
K is the sequence number of time of measuring, is the number of w (t) points；A0It is the width of first major frequency components monochromatic wave
Value；
Based on state observer model, each mould using existing discrete type Kalman filter to major frequency components
The amplitude A of state monochromatic wavei,kAnd phaseKalman's observer is designed, Kalman's observer is： For each mode simple harmonic quantity of the major frequency components to step (2)
Ripple observes obtained main observation,For the amplitude observation of each mode monochromatic wave of the major frequency components to step (2),
wkTo observe obtained measurement observation to each mode monochromatic wave of measurement in step (1)；C0It isWithBetween conversion square
Battle array, is constant matrices；Φ is the known matrix of existing discrete type Kalman filter, Φ0It is first element of the matrix,
x0It is first element of each mode monochromatic wave matrix of major frequency components, " each mode monochromatic wave square of major frequency components
Battle array " refer to formula 4. in x；It isFirst element of matrix；
Kalman's observer gain matrix isFormula 8. in P 'kIt is
PkDerivative, 9. obtained by formula；
State variableCovariance matrix beI E R are existing
There is the known constant of discrete type Kalman filter；P ' is covariance matrix P derivative, P0It is first of covariance matrix P
Element；Covariance matrix be formula 9. because formula 9. the inside it is related to P such as Pk、Pk+1All obtain and a matrix is formed after coming
P is called covariance matrix；
To reevaluate value after primary system identification to system noise Q, System Discrimination is the function that existing Kalman filter carries, and Δ T refers to the time step of Kalman's observer observation data
Long, the value is manually set；
For the monochromatic wave of single mode, the harmonic wave of standard is considered in predicted time
WithBy
7. formula is tried to achieve；FormulaIn, only AObs,i,kWithFor unknown number to be asked；tkIt it is the discrete time, when t is continuous
Between, discrete type Kalman filter is being have passed through, the time also becomes discrete；
Assert that frequency is constant, fFFT,i,0For the initial value of i-th of the modal frequency solved after FFT, obtain pre-
Surveying parameter is
fObs,i,k=fFFT,i,0
It is further preferred that Δ T value is 0.001s-1s.
According to currently preferred, in step (4), the span of predicted time section is 0.001s-5s.
According to currently preferred, in the step (4), the standby following certain predicted time section T of extra large frockPredHeave fortune
It is dynamic to predict that synthesis type is
xoffBe formula 5. in constant value.
Beneficial effects of the present invention are as follows：
The present invention is that wave block mold is predicted, more comprehensively；It is provided by the invention to be become based on fast Flourier
Change, the Wave Model Forecasting Methodology of active disturbance rejection state observer and Parameter Adaptive Compensation, it is more rigorous accurate, in existing equipment
On the premise of, the good value of forecasting, Ke Yiyong can be obtained to the lash ship heave movement under Unequal time lag and different sea situations
The real-time dynamic of MRU detections and Actuator dynamic time lag compensates in the control of follow-up heave compensation.
Brief description of the drawings
Fig. 1 is the Forecasting Methodology schematic diagram of the present invention；
Fig. 2 is prediction effect figure of the present invention in the case where simulated experiment delay time is 1s three-level sea situations；
Fig. 3 is prediction-error image of the present invention in the case where simulated experiment delay time is 1s three-level sea situations；
Fig. 4 is prediction effect figure of the present invention in the case where simulated experiment delay time is 5s three-level sea situations；
Fig. 5 is prediction-error image of the present invention in the case where simulated experiment delay time is 5s three-level sea situations.
Embodiment
The present invention is further described with reference to embodiment, experimental example and accompanying drawing, but not limited to this.
Embodiment 1
Due to (the inertial navigator IMU, or motion measurement list such as active compensation device and heave attitude-measuring sensor
First MRU) regular hour hysteresis be present, the performance of active compensation can be had a strong impact on or even cause compensation system unstability.Cause
This, the forecasting research for heaving state to lash ship has great theory significance and actual application value.It is a kind of base as shown in Figure 1
Illustrate in the Wave Model Forecasting Methodology flow of FFT, active disturbance rejection state observer and Parameter Adaptive Compensation
Figure, comprises the following steps：
(1) heave movement is directed to, the heave displacement signal in a measuring section, heave are gathered using measurement sensor
Displacement signal is to be combined under different time signal by multiple monochromatic waves, and quick Fu is carried out to measured heave displacement signal
In leaf transformation, obtain amplitude and phase that each monochromatic wave corresponds to time signal；
The time interval Δ T wherein calculatedFFTWant Rational choice, Δ TFFTValue is 500s.
Heave movement is expressed as to the superposition of n monochromatic wave：
N be monochromatic wave mode number, Ai、fi、It is amplitude, frequency and the phase corresponding to the monochromatic wave of i-th of mode
Position, h (t) are the initial value after monochromatic wave superposition.
(2) the major frequency components quantity for going out heave movement to magnitude peak detection and identification and each mode monochromatic wave institute are passed through
Corresponding harmonic parameters, n ', A in corresponding diagram 1FFT、fFFT、
On the basis of maximum amplitude peak value, what magnitude peak was more than maximum amplitude peak value 80% is major frequency components,
Draw the quantity of the major frequency components of heave movement；Harmonic parameters include amplitude, frequency and phase.Calculating time interval
ΔTFFTInterior, the change of non-principal frequency content can be ignored, and error caused by non-principal frequency content change can walked subsequently
Suddenly the error feedback parameter self-adjusting link of (3) state observer compensates automatically.
The ODE of each mode of major frequency components is
N ' be detection after major frequency components monochromatic wave quantity, t0For the initial time of selected measuring section, xi
For wiExpression formula after Fast Fourier Transform (FFT), CiFor wiCorresponding xiCoefficient,For xiDifferential form；
System primary condition is
(3) Kalman's observer, each mode using Kalman's observer to the major frequency components in step (2) are designed
Monochromatic wave is observed, and draws main observation；Each mode monochromatic wave of measurement in step (1) is entered using Kalman's observer
Row observation, draws measurement observation, according to the deviation of measurement observation and main observation, On-line Estimation renewal main frequency into
The harmonic prediction parameter corresponding to each mode monochromatic wave divided；
State observer model is
Formula 4. in x be all xiThe matrix of composition；
Initialization function isx1,0For
One major frequency components is in t0The value at moment；
xoff,0For t0Moment x (t0) one biasing, be the constant value obtained by state observer；C(t0) it is all Ci's
Matrix；State observer model is discrete to be turned to
K is the sequence number of time of measuring, is the number of w (t) points；A0It is the width of first major frequency components monochromatic wave
Value.
Based on state observer model, each mould using existing discrete type Kalman filter to major frequency components
The amplitude A of state monochromatic wavei,kAnd phaseKalman's observer is designed,
Kalman's observer is： For to step (2)
Each mode monochromatic wave of major frequency components observes obtained main observation,For the major frequency components to step (2)
The amplitude observation of each mode monochromatic wave, wkMeasurement to observe obtaining to each mode monochromatic wave of measurement in step (1) is observed
Value；C0It isWithBetween transition matrix, be constant matrices；Φ is the known square of existing discrete type Kalman filter
Battle array, Φ0It is first element of the matrix；x0It is first element of each mode monochromatic wave matrix of major frequency components, it is described
" each mode monochromatic wave matrix of major frequency components " refer to formula 4. in x；It isFirst element of matrix.
Kalman's observer gain matrix isFormula 8. in P 'kIt is
PkDerivative, 9. obtained by formula；
State variableCovariance matrix beI E R are existing
There is the known constant of discrete type Kalman filter；P ' is covariance matrix P derivative, P0It is first of covariance matrix P
Element；Covariance matrix be formula 9. because formula 9. the inside it is related to P such as Pk、Pk+1All obtain and a matrix is formed after coming
P is called covariance matrix；
To reevaluate value after primary system identification to system noise Q,
System Discrimination is the work(that existing Kalman filter carries
Energy；Δ T refers to the time step of Kalman's observer observation data, and the value is manually set, and Δ T values are in the present embodiment
0.01s；
For the monochromatic wave of single mode, the harmonic wave of standard is considered in predicted time
WithBy
7. formula is tried to achieve；FormulaIn, only AObs,i,kWithFor unknown number to be asked；tkIt it is the discrete time, when t is continuous
Between, discrete type Kalman filter is being have passed through, the time also becomes discrete；
Assert that frequency is constant, fFFT,i,0For the initial value of i-th of the modal frequency solved after FFT, obtain pre-
Surveying parameter is
fObs,i,k=fFFT,i,0
(4) the harmonic prediction parameter corresponding to each mode monochromatic wave based on the major frequency components obtained by step (3), in advance
Survey the heave movement synthesized in the standby following certain predicted time section of extra large frock；
Such as T in the standby following certain predicted time section of extra large frockPredHeave movement prediction synthesis type be
xoffBe formula 5. in constant value.
Predicted time section TPredIt is also known as delay time in an experiment, choosing 1.0s in the present embodiment is predicted calculating,
As a result show that this algorithm can will predict control errors within the 20% of maximum heave amplitude.
Embodiment 2
A kind of Wave Model Forecasting Methodology based on active disturbance rejection state observer, its step as described in Example 1, are distinguished
In, in step (1), Δ TFFTValue is 5min.
Embodiment 3
A kind of Wave Model Forecasting Methodology based on active disturbance rejection state observer, its step as described in Example 1, are distinguished
In, in step (1), Δ TFFTValue is 10min.
Embodiment 4
A kind of Wave Model Forecasting Methodology based on active disturbance rejection state observer, its step as described in Example 1, are distinguished
In in step (3), Δ T values are 0.001s.
Embodiment 5
A kind of Wave Model Forecasting Methodology based on active disturbance rejection state observer, its step as described in Example 1, are distinguished
In in step (3), Δ T values are 1s.
Embodiment 6
A kind of Wave Model Forecasting Methodology based on active disturbance rejection state observer, its step as described in Example 1, are distinguished
In, in step (4), predicted time section TPred0.25s is chosen in an experiment.
Embodiment 7
A kind of Wave Model Forecasting Methodology based on active disturbance rejection state observer, its step as described in Example 1, are distinguished
In, in step (4), predicted time section TPred0.75s is chosen in an experiment.
Embodiment 8
A kind of Wave Model Forecasting Methodology based on active disturbance rejection state observer, its step as described in Example 1, are distinguished
In, in step (4), predicted time section TPred3s is chosen in an experiment.
Embodiment 9
A kind of Wave Model Forecasting Methodology based on active disturbance rejection state observer, its step as described in Example 1, are distinguished
In, in step (4), predicted time section TPred5s is chosen in an experiment.
Experimental example 1
The algorithm steps provided according to embodiment 1, in delay time TPredFor under 1s three-level sea situations, i.e. significant wave height ζ1/3
For 0.9 meter, T average period1For 3.9 seconds, wave period substantial scope 1.4~7.6 seconds, Wave Model is entered in Simulink
Row simulation, in this experimental example, Δ T in step (1)FFTValue is 500s, and Δ T values are 0.01s in step (3), and with this
Invent the step to be predicted, prediction result is as shown in Figure 2.As can be seen that in the case of time delay 1s, tried to achieve using this programme
Predicted value remain to accurately follow actual value.As shown in figure 3, worst error about ± 2.8m, this hair caused by time delay
Bright prediction algorithm maximum prediction error about ± 0.5m, precision of prediction is about 92%.
The time delay error that dotted line represents in Fig. 3 is that the time delay value in Fig. 2 subtracts actual value；The prediction that solid line represents in Fig. 3
Error is that the predicted value in Fig. 2 subtracts actual value.
Experimental example 2
In this experimental example, delay time T is setPredFor 5s, in the case of sea situation is with test example 1, and use of the present invention
Algorithm steps are predicted, and prediction result is as shown in Figure 4.As can be seen that in the case of time delay 5s, tried to achieve using this programme pre-
Measured value can accurately follow actual value.As shown in figure 5, the worst error about ± 0.7m caused by time delay, present invention prediction
Algorithm maximum prediction error about ± 0.2m, precision of prediction is about 90%.
The time delay error that dotted line represents in Fig. 5 is that the time delay value in Fig. 4 subtracts actual value；The prediction that solid line represents in Fig. 5
Error is that the predicted value in Fig. 4 subtracts actual value.
Described above is only the preferred embodiment of the present invention, it should be pointed out that：For the ordinary skill people of the art
For member, under the premise without departing from the principles of the invention, some improvement and retouching can also be made, these are improved also should with retouching
It is considered as protection scope of the present invention.
Claims (9)
1. a kind of Wave Model Forecasting Methodology based on active disturbance rejection state observer, it is characterised in that comprise the following steps：
(1) heave movement is directed to, the heave displacement signal in a measuring section is gathered using measurement sensor, heaves displacement
Signal is to be combined under different time signal by multiple monochromatic waves, and fast Fourier is carried out to measured heave displacement signal
Conversion, obtains amplitude and phase that each monochromatic wave corresponds to time signal；
(2) by going out to magnitude peak detection and identification corresponding to major frequency components quantity and each mode monochromatic wave of heave movement
Harmonic parameters；
(3) Kalman's observer, each mode simple harmonic quantity using Kalman's observer to the major frequency components in step (2) are designed
Ripple is observed, and draws main observation；Each mode monochromatic wave of measurement in step (1) is seen using Kalman's observer
Survey, draw measurement observation, according to measurement observation and the deviation of main observation, On-line Estimation updates major frequency components
Harmonic prediction parameter corresponding to each mode monochromatic wave；
(4) the harmonic prediction parameter corresponding to each mode monochromatic wave based on the major frequency components obtained by step (3), prediction are closed
Heave movement into the standby following certain predicted time section of extra large frock.
2. the Wave Model Forecasting Methodology according to claim 1 based on active disturbance rejection state observer, it is characterised in that step
Suddenly in (1), measuring section Δ TFFTValue be 5min-10min.
3. the Wave Model Forecasting Methodology according to claim 1 based on active disturbance rejection state observer, it is characterised in that step
Suddenly in (1), heave movement is expressed as to the superposition of n monochromatic wave：
N be monochromatic wave mode number, Ai、fi、It is amplitude, frequency and the phase corresponding to the monochromatic wave of i-th of mode, h
(t) it is the initial value after monochromatic wave superposition.
4. the Wave Model Forecasting Methodology according to claim 1 based on active disturbance rejection state observer, it is characterised in that step
Suddenly it is described to be detected as in (2), on the basis of maximum amplitude peak value, magnitude peak be more than maximum amplitude peak value 80% i.e. based on
Frequency content is wanted, that is, draws the monochromatic wave quantity of the major frequency components of heave movement；Harmonic parameters include amplitude, frequency and phase
Position.
5. the Wave Model Forecasting Methodology according to claim 4 based on active disturbance rejection state observer, it is characterised in that step
Suddenly in (2), the ODE of each mode monochromatic wave of major frequency components is
N ' be detection after major frequency components monochromatic wave quantity, t0For the initial time of selected measuring section, xiFor wi
Expression formula after Fast Fourier Transform (FFT), CiFor wiCorresponding xiCoefficient,For xiDifferential form；
System primary condition is
6. the Wave Model Forecasting Methodology according to claim 1 based on active disturbance rejection state observer, it is characterised in that step
Suddenly in (3), state observer model is
Formula 4. in x be all xiThe matrix of composition；
Initialization function is
x1,0It is first major frequency components in t0The value at moment；
xoff,0For t0Moment x (t0) one biasing, be the constant value obtained by state observer；C(t0) it is all CiSquare
Battle array；
State observer model is discrete to be turned to
K is the sequence number of time of measuring, is the number of w (t) points；A0It is the amplitude of first major frequency components monochromatic wave；
Based on state observer model, using each mode letter of the existing discrete type Kalman filter to major frequency components
The amplitude A of harmonic wavei,kAnd phaseKalman's observer is designed, Kalman's observer is：
Obtained main observation is observed for each mode monochromatic wave of the major frequency components to step (2),For to step
(2) the amplitude observation of each mode monochromatic wave of major frequency components, wkFor each mode monochromatic wave to measurement in step (1)
Observe obtained measurement observation；C0It isWithBetween transition matrix, be constant matrices；Φ is existing discrete type karr
The known matrix of graceful wave filter, Φ0It is first element of the matrix, x0It is each mode monochromatic wave matrix of major frequency components
First element；Each mode monochromatic wave matrix of major frequency components be formula 4. in x；It isFirst member of matrix
Element；
Kalman's observer gain matrix is
State variableCovariance matrix beI E R for it is existing from
Dissipate the known constant of type Kalman filter；
To reevaluate value after primary system identification to system noise Q,
System Discrimination is the function that existing Kalman filter carries, and Δ T is
Refer to the time step of Kalman's observer observation data；
For the monochromatic wave of single mode, the harmonic wave of standard is considered in predicted time
With7. tried to achieve by formula；FormulaIn, only AObs,i,kWithFor unknown number to be asked；
Assert that frequency is constant, fFFT,i,0For the initial value of i-th of the modal frequency solved after FFT, Prediction Parameters are obtained
For
7. the Wave Model Forecasting Methodology according to claim 6 based on active disturbance rejection state observer, it is characterised in that step
Suddenly in (3), it is preferred that Δ T value is 0.001s-1s.
8. the Wave Model Forecasting Methodology according to claim 1 based on active disturbance rejection state observer, it is characterised in that step
Suddenly in (4), the span of predicted time section is 0.001s-5s.
9. the Wave Model Forecasting Methodology according to claim 1 based on active disturbance rejection state observer, it is characterised in that institute
State in step (4), the standby following certain predicted time section T of extra large frockPredHeave movement prediction synthesis type be
xoffBe formula 5. in constant value.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201710576846.3A CN107357170A (en) | 2017-07-14 | 2017-07-14 | A kind of Wave Model Forecasting Methodology based on active disturbance rejection state observer |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201710576846.3A CN107357170A (en) | 2017-07-14 | 2017-07-14 | A kind of Wave Model Forecasting Methodology based on active disturbance rejection state observer |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN107357170A true CN107357170A (en) | 2017-11-17 |
Family
ID=60292678
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201710576846.3A Pending CN107357170A (en) | 2017-07-14 | 2017-07-14 | A kind of Wave Model Forecasting Methodology based on active disturbance rejection state observer |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN107357170A (en) |
Cited By (9)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN108877372A (en) * | 2018-06-29 | 2018-11-23 | 山东大学 | A kind of experimental provision of active/passive compensation of undulation |
| CN108875251A (en) * | 2018-07-03 | 2018-11-23 | 广东工业大学 | Wave period prediction technique, device and equipment |
| CN110319838A (en) * | 2019-07-09 | 2019-10-11 | 哈尔滨工程大学 | A kind of adaptive athletic posture frame of reference heave measurement method |
| CN111352447A (en) * | 2018-12-24 | 2020-06-30 | 中国科学院沈阳自动化研究所 | Wave compensation method for take-off and landing platform of unmanned aerial vehicle |
| CN112611382A (en) * | 2020-11-27 | 2021-04-06 | 哈尔滨工程大学 | Strapdown inertial navigation system heave measurement method with phase compensation |
| CN112629525A (en) * | 2020-11-13 | 2021-04-09 | 河北汉光重工有限责任公司 | Method for ship heave phase compensation based on historical data cross-correlation |
| CN114489010A (en) * | 2022-01-25 | 2022-05-13 | 佛山智能装备技术研究院 | ADRC extended observer state observation error real-time prediction method and system |
| CN116231645A (en) * | 2023-05-09 | 2023-06-06 | 中车山东风电有限公司 | Offshore wind farm power generation amount calculation method, calculation system and calculation terminal |
| WO2024026985A1 (en) * | 2022-08-03 | 2024-02-08 | 苏州海希夫智控科技有限公司 | Active heave compensation-based anti-swing method for ship |
Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN104050379A (en) * | 2014-06-25 | 2014-09-17 | 东南大学 | Sea wave height prediction method based on ARMA model |
| KR20160102805A (en) * | 2015-02-23 | 2016-08-31 | 주식회사 동녘 | Ocean weather analysis device and method using thereof |
| CN106599427A (en) * | 2016-12-06 | 2017-04-26 | 哈尔滨工程大学 | Ocean wave information prediction method based on Bayesian theory and hovercraft attitude information |
-
2017
- 2017-07-14 CN CN201710576846.3A patent/CN107357170A/en active Pending
Patent Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN104050379A (en) * | 2014-06-25 | 2014-09-17 | 东南大学 | Sea wave height prediction method based on ARMA model |
| KR20160102805A (en) * | 2015-02-23 | 2016-08-31 | 주식회사 동녘 | Ocean weather analysis device and method using thereof |
| CN106599427A (en) * | 2016-12-06 | 2017-04-26 | 哈尔滨工程大学 | Ocean wave information prediction method based on Bayesian theory and hovercraft attitude information |
Non-Patent Citations (1)
| Title |
|---|
| 李世振: "重型海工装备升沉补偿电液控制系统研究", 《中国博士学位论文全文数据库 信息科技辑》 * |
Cited By (13)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN108877372B (en) * | 2018-06-29 | 2024-02-23 | 山东大学 | Experimental device for active-passive wave compensation |
| CN108877372A (en) * | 2018-06-29 | 2018-11-23 | 山东大学 | A kind of experimental provision of active/passive compensation of undulation |
| CN108875251A (en) * | 2018-07-03 | 2018-11-23 | 广东工业大学 | Wave period prediction technique, device and equipment |
| CN108875251B (en) * | 2018-07-03 | 2022-06-24 | 广东工业大学 | Wave period prediction method, device and equipment |
| CN111352447A (en) * | 2018-12-24 | 2020-06-30 | 中国科学院沈阳自动化研究所 | Wave compensation method for take-off and landing platform of unmanned aerial vehicle |
| CN110319838A (en) * | 2019-07-09 | 2019-10-11 | 哈尔滨工程大学 | A kind of adaptive athletic posture frame of reference heave measurement method |
| CN112629525A (en) * | 2020-11-13 | 2021-04-09 | 河北汉光重工有限责任公司 | Method for ship heave phase compensation based on historical data cross-correlation |
| CN112611382A (en) * | 2020-11-27 | 2021-04-06 | 哈尔滨工程大学 | Strapdown inertial navigation system heave measurement method with phase compensation |
| CN112611382B (en) * | 2020-11-27 | 2022-06-21 | 哈尔滨工程大学 | Strapdown inertial navigation system heave measurement method with phase compensation |
| CN114489010A (en) * | 2022-01-25 | 2022-05-13 | 佛山智能装备技术研究院 | ADRC extended observer state observation error real-time prediction method and system |
| WO2024026985A1 (en) * | 2022-08-03 | 2024-02-08 | 苏州海希夫智控科技有限公司 | Active heave compensation-based anti-swing method for ship |
| CN116231645B (en) * | 2023-05-09 | 2023-08-11 | 中车山东风电有限公司 | Offshore wind farm power generation amount calculation method, calculation system and calculation terminal |
| CN116231645A (en) * | 2023-05-09 | 2023-06-06 | 中车山东风电有限公司 | Offshore wind farm power generation amount calculation method, calculation system and calculation terminal |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN107357170A (en) | A kind of Wave Model Forecasting Methodology based on active disturbance rejection state observer | |
| DE102008024513B4 (en) | Crane control with active coast sequence | |
| Messineo et al. | Offshore crane control based on adaptive external models | |
| CN110806692A (en) | Wave compensation prediction method based on CNN-LATM combined model | |
| Ueno et al. | Estimation and prediction of effective inflow velocity to propeller in waves | |
| Yang et al. | A novel delta operator Kalman filter design and convergence analysis | |
| CN108469736B (en) | Marine crane anti-swing positioning control method and system based on state observation | |
| CN109917657B (en) | Anti-interference control method and device for dynamic positioning ship and electronic equipment | |
| Nielsen | Introducing two hyperparameters in Bayesian estimation of wave spectra | |
| CN110374804B (en) | Variable pitch control method based on gradient compensation of depth certainty strategy | |
| CN109739248B (en) | Ship-borne three-degree-of-freedom parallel stable platform stability control method based on ADRC | |
| Fortuna et al. | A roll stabilization system for a monohull ship: modeling, identification, and adaptive control | |
| CN107065569B (en) | Ship dynamic positioning sliding mode control system and method based on RBF neural network compensation | |
| Li et al. | Compensated model-free adaptive tracking control scheme for autonomous underwater vehicles via extended state observer | |
| CN110083057A (en) | PID control method based on hydrofoil athletic posture | |
| CN113297798A (en) | Robot external contact force estimation method based on artificial neural network | |
| Abdelrahman et al. | Observer-based unknown input estimator of wave excitation force for a wave energy converter | |
| Ianagui et al. | Robust output-feedback control in a dynamic positioning system via high order sliding modes: Theoretical framework and experimental evaluation | |
| CN116026325A (en) | Navigation method and related device based on neural process and Kalman filtering | |
| Davis et al. | Systematic identification of drag coefficients for a heaving wave follower | |
| CN115933413A (en) | Self-adaptive inversion sliding mode control method based on preset performance and aiming at network control system | |
| Cavallo et al. | Gray-box identification of continuous-time models of flexible structures | |
| Cholette et al. | Precedent-free fault isolation in a diesel engine exhaust gas recirculation system | |
| Fan et al. | Attitude Adaptive Robust Strong Tracking Kalman Filter Based on Unmanned Surface Vehicle Navigational Radar Target Tracking | |
| Xu et al. | A sliding mode predictive anti-pitching control for a high-speed multihull |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| WD01 | Invention patent application deemed withdrawn after publication | ||
| WD01 | Invention patent application deemed withdrawn after publication |
Application publication date: 20171117 |