CN113391287B - High-frequency ground wave radar sea state data fusion method based on time sequence - Google Patents
High-frequency ground wave radar sea state data fusion method based on time sequence Download PDFInfo
- Publication number
- CN113391287B CN113391287B CN202110649446.7A CN202110649446A CN113391287B CN 113391287 B CN113391287 B CN 113391287B CN 202110649446 A CN202110649446 A CN 202110649446A CN 113391287 B CN113391287 B CN 113391287B
- Authority
- CN
- China
- Prior art keywords
- data
- sea state
- time
- sequence
- equation
- 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.)
- Active
Links
Classifications
-
- 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
- G01S7/418—Theoretical aspects
-
- 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
- G01S7/414—Discriminating targets with respect to background clutter
-
- 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
- G01S7/415—Identification of targets based on measurements of movement associated with the target
-
- 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
Abstract
The invention discloses a high-frequency ground wave radar sea state data fusion method based on a time sequence, and relates to a high-frequency ground wave radar sea state data fusion method. The invention aims to solve the problem that the accuracy rate of obtaining sea state data is low due to abnormal sea state data obtained under the condition of interference of the existing high-frequency ground wave radar. The process is as follows: and (3) a step of: obtaining fused data; and II: all sequences of the fused data meet the stability; thirdly,: fitting a model to the sequence meeting the stability to obtain a time sequence equation; fourth, the method comprises the following steps: judging whether a time sequence equation is valid or not; fifth step: obtaining a predicted value of the next moment; sixth,: obtaining a state transition equation and a noise equation of the self-adaptive Kalman filtering; seventh,: obtaining sea state filtering data; eighth step: and sliding a time window with a certain length m according to the actual situation, performing sea state inversion processing on the echo to be detected to obtain fused data, and repeating two to seven to obtain sea state filtering data. The invention belongs to the field of high-frequency ground wave radar sea state inversion.
Description
Technical Field
The invention belongs to the field of high-frequency ground wave radar sea state inversion, and particularly relates to a high-frequency ground wave radar sea state data fusion method.
Background
In addition to research on beyond-line-of-sight detection of an offshore target, another important research content of the high-frequency ground wave radar is to apply echo data received by the radar to invert ocean information and perform ocean state remote sensing. However, high frequency ground wave radars have very complex electromagnetic operating environments including ionospheric clutter interference, radio station interference, and the like. The complexity of the interference and sea conditions can influence the energy of the first-order and second-order spectral regions of the ocean echo spectrum, the intensity and offset of the first-order Bragg peaks, and further influence the inversion of sea state parameters. Considering that the change of the ocean information is based on a time-dependent slow change process, when radar double-frequency detection is performed, the ocean data processing can be performed by adopting a double-frequency self-adaptive Kalman filtering process based on a time sequence. And (3) carrying out time sequence model fitting on the sea state information, then predicting the sea state information at the future moment, and simultaneously calculating the optimal sea state information through a self-adaptive Kalman filtering algorithm.
Disclosure of Invention
The invention aims to solve the problem that the sea state data obtained by the existing high-frequency ground wave radar under the condition of interference is abnormal, so that the accuracy rate of obtaining the sea state data is low.
The high-frequency ground wave radar sea state data fusion method based on the time sequence comprises the following specific processes:
step one: performing sea inversion processing on echoes received by the high-frequency ground wave radar to obtain carrier frequency 1 sea state data and carrier frequency 2 sea state data, and fusing the carrier frequency 1 sea state data and the carrier frequency 2 sea state data to obtain fused data;
step two: performing stability test on the fused data, and differentiating sequences which do not meet the stability until all the sequences of the fused data meet the stability;
step three: fitting a model to the sequence meeting the stability to obtain a time sequence equation;
step four: subtracting the sequence fitted by the time sequence equation from the fused data in the first step to obtain a residual sequence, observing whether the residual sequence accords with normal distribution, and when the residual sequence meets the normal distribution, enabling the time sequence equation ARIMA (p, i, q) to be effective; when the residual sequence does not meet the normal distribution, the time sequence equation ARIMA (p, i, q) is invalid;
step five: substituting the sea state data into an effective time sequence equation ARIMA (p, i, q) according to the time sequence, and obtaining a predicted value of the next moment;
step six: obtaining a state transition equation and a noise equation of the adaptive Kalman filtering according to coefficients in an effective time sequence equation ARIMA (p, i, q);
step seven: giving an initial value, and performing self-adaptive Kalman filtering on the sea state data to obtain sea state filtering data;
step eight: according to the actual situation, sliding a time window with a certain length m, wherein m is more than or equal to 1 and less than or equal to n, performing sea state inversion processing on echoes received by the high-frequency ground wave radar to obtain carrier frequency 1 sea state data and carrier frequency 2 sea state data, fusing the carrier frequency 1 sea state data and the carrier frequency 2 sea state data to obtain fused data, and repeating the second step to the seventh step to obtain sea state filtering data.
Preferably, in the first step, sea state inversion processing is performed on the echo received by the high-frequency ground wave radar to obtain carrier frequency 1 sea state data and carrier frequency 2 sea state data, and the carrier frequency 1 sea state data and the carrier frequency 2 sea state data are fused to obtain fused data;
the n is the length of the data time window;
the specific process is as follows:
the sequence corresponding to the carrier frequency 1 sea state Data is Data 1 (i)，i＝1,2,3...n；
The sequence corresponding to the carrier frequency 2 sea state Data is Data 2 (i)，i＝1,2,3...n；
The signal to noise ratio of the carrier frequency 1 sea state data is Snr 1 (i),i＝1,2,3...n；
The signal to noise ratio of the carrier frequency 2 sea state data is Snr 2 (i),i＝1,2,3,...n；
The n is the length of the data time window;
the carrier frequency 1 sea state Data and the carrier frequency 2 sea state Data are fused to obtain fused Data (i), i=1, 2,3,..n, and the fused Data (i) has the expression:
where, is the multiplier.
Preferably, in the second step, the stability of the fused data is checked, and the sequences which do not meet the stability are differentiated until all the sequences of the fused data meet the stability; the method comprises the following specific steps:
performing stability test on the fused Data (i) by adopting a time sequence diagram method;
if all sequences of the fused Data meet the stability, obtaining a sequence data_stable (i), i=1, 2,3,..n, wherein data_stable (i) =data (i), i=1, 2,3,..n;
if a certain sequence of the fused data does not meet the smoothness, the sequence which does not meet the smoothness is compared with the sequence which does not meet the smoothness
Data (i), i=1, 2, 3..n, the sequence after the difference being data_diff (i), i=1, 2, 3..n-1, wherein
Data_diff(i)＝Data(i+1)-Data(i),i＝1,2,3,..n-1；
Continuously performing stability test on the sequence after the difference, if the stability is not met, continuously performing the difference until the stability is met to obtain a sequence data_stable (i), i=1, 2,3,..m', and recording the number of times i of the difference;
where data_stable (i) =data_diff (i), i=1, 2,3,..m'.
Preferably, in the third step, fitting a model to the sequence satisfying the stationarity to obtain a time sequence equation; the method comprises the following specific steps:
step three, fitting the sequence data_stable (i) meeting stationarity, i=1, 2,3,..n to a time sequence equation ARIMA (p, i, q);
wherein p is the number of AR model parameters, and q is the number of MA model parameters;
step three, an AIC (automatic identification) order determining criterion is used for determining an ARIMA (p, i, q) model, and the specific process is as follows:
the AIC order criteria are as follows:
different p and q are selected, and corresponding AIC values are calculated, so that the number of coefficients corresponding to the AIC reaching the minimum value is p, and the q takes the best value;
thirdly, estimating the coefficient value of the ARIMA (p, i, q) model after p and q are determined to obtain the coefficient value of the ARIMA (p, i, q) model after p and q are determined;
and thirdly, taking the expression of the time sequence equation ARIMA (p, i, q) based on the number of the coefficients and the coefficient values, and obtaining the determined time sequence equation ARIMA (p, i, q).
Preferably, the expression of the time series equation ARIMA (p, i, q) in the third step is:
x(t)＝φ 1 x(t-1)+φ 2 x(t-2)+...+φ p x(t-p)+a(t)+θ 1 a(t-1)+θ 2 a(t-2)+...θ q a(t-q)
wherein x (t) is sea state data at time t, t is time phi 1 、φ 2 、φ p Is sea state data coefficient, a is noise, theta 1 、θ 2 、θ q For the noise coefficient, p is the number of AR model parameters, and q is the number of MA model parameters.
Preferably, in the third step
Where Q is the sum of the remaining squares of the ARIMA (p, i, Q) model, N is the length of the sequence,is the residual variance of the sequence, μ is the mean of the sequence.
Preferably, in the sixth step, a state transition equation and a noise equation of the adaptive kalman filter are obtained according to coefficients in an effective time sequence equation ARIMA (p, i, q); the method comprises the following specific steps:
the time series equation model is obtained by the effective time series equation ARIMA (p, i, 0) as follows:
x(t)＝φ 1 x(t-1)+φ 2 x(t-2)+...+φ p x(t-p)+a(t)
let t=t+1 to give
Let x 1 (t)＝x(t),x 2 (t)＝x(t-1),...,x p (t) =x (t-p+1), then there is
Due to x 2 (t+1)＝x 1 (t),x 3 (t+1)＝x 2 (t),...,x p (t+1)＝x p-1 (t)
Can obtain
Wherein x is 1 (t) is the value of x (t) at the time of t, and x p (t) is the value of the time x (t) of p; x is x 1 (t+1) is the sea state data value at time t+1, x 2 (t+1) is the sea state data value at time t, x p (t+1) is a sea state data value at the time t-p+2;
where Φ is the state transition matrix and Γ is the noise matrix.
Preferably, the step seven is given an initial value, and adaptive kalman filtering is performed on the sea state data to obtain sea state filtering data; the method comprises the following specific steps:
(1) Obtaining a predicted state X (t|t-1) =ΦX (t-1) at time t from the state at time t-1
Wherein X (t|t-1) is a predicted state at time t and X (t-1) is a state at time t-1;
(2) Prediction P (t|t-1) =Φp (t-1) Φ, giving covariance matrix T +ΓQ(t)Γ T
Wherein Q is a system noise covariance matrix, P (t|t-1) is a value at the time of T-1 predicted according to the value at the time of T-1, P (T-1) is a value at the time of T-1, and T is a transposition;
(3) Introducing an adjustment update parameter for the adaptive Kalman filtering:
wherein b is a forgetting factor;
(4) Calculating the distance between the predicted value and the two observed values, and selecting the observed value with the short distance as the observed value of Kalman filtering;
the observation error is as follows:
v(t)＝Z(t)-H(t)X(t|t-1)
wherein H is an observation matrix, and Z is an observation value;
(5) Updating the observed noise covariance matrix:
R(t)＝(1-d t )R(t-1)+d t ([I-H(t)K(t-1)]v(t)v T (t)[I-H(t)K(t-1)] T +H(t)P(t-1)H T (t))
wherein I is an identity matrix, R (t-1) is an observation noise covariance matrix at the time t-1, K (t-1) is a Kalman gain at the time t-1, and v (t) is an observation error;
(6) Calculation of Kalman gain K (t) =P (t|t-1) H T (t)S- 1 (t)
Wherein S (t) =h (t) P (t|t-1) H T (t) +R (t), K (t) is Kalman gain at the time of t-1, P (t|t-1) is an observation noise covariance matrix at the time of t predicted according to an observation noise covariance matrix at the time of t-1, S (t) is an intermediate variable, and H (t) is an observation matrix at the time of t;
(7) Updating covariance matrix P (t) = [ I-K (t) H (t) ] P (t|t-1)
(8) Update state X (t) =x (t|t-1) +k (t) v (t) at time K
(9) The sea state filter data Y (t) is obtained.
Preferably, the forgetting factor b takes the value of: b is more than or equal to 0.95 and less than or equal to 0.99.
Preferably, the sea state filtered data Y (t) =x (t) H (t) in (9).
The beneficial effects of the invention are as follows:
in order to solve the problems, the invention provides a time sequence-based double-frequency self-adaptive Kalman filtering processing method based on the functional characteristics of double-frequency detection of a high-frequency ground wave radar. The method is used for data processing of sea state inversion of the high-frequency ground wave radar, reduces the influence of interference, corrects inversion data and improves the accuracy of obtaining sea state data. A series of offshore activities in the ocean field, including offshore mining (petroleum, coal), offshore fishery, offshore transportation (import and export trade) and the like, have real-time ocean information as references, and can be reasonably and effectively arranged. Meanwhile, in order to prevent marine disasters such as offshore typhoons, red tides, tsunamis, we also need more comprehensive and accurate marine data. It is therefore necessary to monitor the sea surface in real time. The marine environment has many effects. Whether traffic, national defense safety or natural weather, the information is closely related to sea. The accurate ocean information plays a vital role in preventing ocean natural disasters, post-disaster reconstruction and the like.
Drawings
FIG. 1 is a schematic diagram of the method for predicting and fusing sea state data of a high-frequency ground wave radar.
FIG. 2 is a diagram showing the comparison between the dual-frequency adaptive Kalman filtering based on time sequence and the adaptive Kalman filtering based on time sequence according to the wind direction of the present invention.
FIG. 3 is a schematic diagram showing the comparison of the dual-frequency adaptive Kalman filter and the dual-frequency adaptive Kalman filter based on time sequence according to the wind direction of the present invention.
Fig. 4 is a schematic diagram showing the comparison between the dual-frequency adaptive kalman filter based on time series and the adaptive kalman filter based on time series according to the wave height of the present invention.
Fig. 5 is a schematic diagram of the comparison result of the dual-frequency adaptive kalman filter and the dual-frequency adaptive kalman filter based on time sequence in wave height according to the present invention.
Detailed Description
The first embodiment is as follows: the high-frequency ground wave radar sea state data fusion method based on the time sequence in the embodiment comprises the following specific processes:
step one: performing sea inversion processing on echoes received by a high-frequency ground wave radar (radio wave) to obtain carrier frequency 1 sea state data and carrier frequency 2 sea state data (flow speed, flow direction, wind speed, wind direction and sea wave height), and fusing the carrier frequency 1 sea state data and the carrier frequency 2 sea state data to obtain fused data;
step two: performing stability test on the fused data, and differentiating sequences which do not meet the stability until all the sequences of the fused data meet the stability;
step three: fitting a model to the sequence meeting the stability (the fitting comprises modeling, order setting and parameter estimation) to obtain a time sequence equation;
step four: subtracting the sequence fitted by the time sequence equation (giving data at 1 time and 2 time) from the fused data (the sequence meeting the stability after fusion), predicting data at 3 time by adopting a time sequence equation ARIMA (p, i, q), predicting data at 4 time by adopting a time sequence equation ARIMA (p, i, q) according to the data at 2 and 3 time, and predicting data at 5 time by adopting a time sequence equation ARIMA (p, i, q) according to the data at 3 and 4 time) to obtain a residual sequence, observing whether the residual sequence meets normal distribution, and when the residual sequence meets the normal distribution, indicating that the extracted useful information is all the time sequence ARIMA (p, i, q) is effective; when the residual sequence does not meet the normal distribution, the time sequence equation ARIMA (p, i, q) is invalid;
step five: substituting sea state data (possibly predicted and possibly inverted) into an effective time sequence equation ARIMA (p, i, q) in time sequence to obtain a predicted value of the next moment;
step six: obtaining a state transition equation and a noise equation of the adaptive Kalman filtering according to coefficients in an effective time sequence equation ARIMA (p, i, q);
step seven: giving an initial value, and performing self-adaptive Kalman filtering on sea state data (predicted by inversion) to obtain sea state filtering data;
step eight: according to the actual situation, sliding a time window with a certain length m, wherein m is more than or equal to 1 and less than or equal to n, performing sea state inversion processing on echoes received by a to-be-detected high-frequency ground wave radar (radio wave) to obtain carrier frequency 1 sea state data and carrier frequency 2 sea state data (flow velocity, flow direction, wind speed, wind direction and sea wave height), fusing the carrier frequency 1 sea state data and the carrier frequency 2 sea state data to obtain fused data, and repeating the steps two to seven to obtain sea state filtering data.
The second embodiment is as follows: the first embodiment is different from the first embodiment in that in the first step, sea state inversion processing is performed on an echo received by a high-frequency ground wave radar (radio wave) to obtain carrier frequency 1 sea state data and carrier frequency 2 sea state data (flow velocity, flow direction, wind speed, wind direction and sea wave height), and the carrier frequency 1 sea state data and the carrier frequency 2 sea state data are fused to obtain fused data;
the n is the length of the data time window;
the specific process is as follows:
the sequence corresponding to the carrier frequency 1 sea state Data is Data 1 (i) I=1, 2, 3..n (a point on the position is one sea element, data at different times of one sea element corresponds to one sequence);
the sequence corresponding to the carrier frequency 2 sea state Data is Data 2 (i)，i＝1,2,3...n；
The signal to noise ratio of the carrier frequency 1 sea state data is Snr 1 (i),i＝1,2,3...n；
The signal to noise ratio of the carrier frequency 2 sea state data is Snr 2 (i),i＝1,2,3,...n；
The n is the length of the data time window;
the carrier frequency 1 sea state Data and the carrier frequency 2 sea state Data are fused to obtain fused Data (i), i=1, 2,3,..n, and the fused Data (i) has the expression:
where, is the multiplier.
Other steps and parameters are the same as in the first embodiment.
And a third specific embodiment: the first embodiment is different from the second embodiment in that in the second step, the stability of the fused data is checked, and the sequences which do not meet the stability are differentiated until all the sequences of the fused data meet the stability; the method comprises the following specific steps:
performing stability test on the fused Data (i) by adopting a time sequence diagram method;
if all sequences of the fused Data meet the stationarity (the timing diagrams of all sequences have no obvious trend item or the periodicity is the stationarity), obtaining a sequence data_stable (i), i=1, 2,3,..n, wherein data_stable (i) =data (i), i=1, 2,3,..n;
if a sequence of the fused Data does not satisfy the stationarity (the time sequence diagram of the sequence has obvious trend items or the periodicity is not satisfied), the sequence Data (i) which does not satisfy the stationarity, i=1, 2, 3..n, is differentiated, and the sequence after the differentiation is data_diff (i), i=1, 2, 3..n-1, wherein
Data_diff(i)＝Data(i+1)-Data(i),i＝1,2,3,..n-1；
Continuously performing stability test on the sequence after the difference, continuously performing the difference if the stability is not met until the stability is met to obtain a sequence data_stable (i), i=1, 2,3,..mj, and recording the number of times i of the difference;
where data_stable (i) =data_diff (i), i=1, 2,3,..m'.
i takes the value i=1, 2, 3..m' because for an unstable sequence it is not certain to do several differences, m is n-1 once and n-2 twice, so one variable is used to cover the cases.
Other steps and parameters are the same as in the first or second embodiment.
The specific embodiment IV is as follows: the difference between the present embodiment and one to three embodiments is that in the third step, fitting (fitting includes modeling, scaling and parameter estimation) of a model is performed on the sequence satisfying the stationarity, so as to obtain a time sequence equation; the method comprises the following specific steps:
step three, fitting the sequence data_stable (i) meeting stationarity, i=1, 2,3,..n to a time sequence equation ARIMA (p, i, q);
wherein p is the number of AR model parameters (the number of sea state data coefficients), and q is the number of MA model parameters (the number of noise coefficients);
step three, an AIC (automatic identification) order-determining criterion is used for determining an ARIMA (p, i, q) model (the number of p and q is known for the purpose, namely, the value is taken), and the specific process is as follows:
the AIC order criteria are as follows:
different p and q are selected, and corresponding AIC values are calculated, so that the number of coefficients corresponding to the AIC reaching the minimum value is p, and the q takes the best value;
for the sequence data_stable (i) satisfying stationarity, i=1, 2,3,..n, the residual variance is found, expressed as:
the number of actual observation values is the number of sequences meeting the stability, and the number of parameters of the model is the number of coefficients (one 5, one 3, and the total is 8);
thirdly, estimating the coefficient value of the ARIMA (p, i, q) model after p and q are determined to obtain the coefficient value of the ARIMA (p, i, q) model after p and q are determined;
and thirdly, taking the expression of the time sequence equation ARIMA (p, i, q) based on the number of the coefficients and the coefficient values, and obtaining the determined time sequence equation ARIMA (p, i, q).
Other steps and parameters are the same as in one to three embodiments.
Fifth embodiment: this embodiment differs from one to four embodiments in that the expression of the time series equation ARIMA (p, i, q) in the third step is:
x(t)＝φ 1 x(t-1)+φ 2 x(t-2)+...+φ p x(t-p)+a(t)+θ 1 a(t-1)+θ 2 a(t-2)+...θ q a(t-q)
wherein x (t) is sea state data at time t, t is time phi 1 、φ 2 、φ p Is sea state data coefficient, a is noise, theta 1 、θ 2 、θ q The noise coefficient, p is the number of AR model parameters (the number of sea state data coefficients), and q is the number of MA model parameters (the number of noise coefficients).
Other steps and parameters are the same as in one to four embodiments.
Specific embodiment six: this embodiment differs from one of the first to fifth embodiments in that in the third step
Where Q is the sum of the remaining squares of the ARIMA (p, i, Q) model, N is the length of the sequence,is the residual variance of the sequence, μ is the mean of the sequence.
Other steps and parameters are the same as in one of the first to fifth embodiments.
Seventh embodiment: the difference between the present embodiment and one to six embodiments is that in the sixth step, a state transition equation and a noise equation of the adaptive kalman filter are obtained according to coefficients in an effective time sequence equation ARIMA (p, i, q); the method comprises the following specific steps:
the time series equation model is obtained by the effective time series equation ARIMA (p, i, 0) as follows:
x(t)＝φ 1 x(t-1)+φ 2 x(t-2)+...+φ p x(t-p)+a(t)
let t=t+1 to give
Let x 1 (t)＝x(t),x 2 (t)＝x(t-1),...,x p (t) =x (t-p+1), then there is
Due to x 2 (t+1)＝x 1 (t),x 3 (t+1)＝x 2 (t),...,x p (t+1)＝x p-1 (t)
Can obtain
Wherein x is 1 (t) is the value of x (t) at the time of t, and x p (t) is the value of the time x (t) of p; x is x 1 (t+1) is the sea state data value at time t+1, x 2 (t+1) is the sea state data value at time t, x p (t+1) is a sea state data value at the time t-p+2;
where Φ is the state transition matrix and Γ is the noise matrix.
Other steps and parameters are the same as in one of the first to sixth embodiments.
Eighth embodiment: the difference between the present embodiment and one of the first to seventh embodiments is that, given an initial value in the seventh step, adaptive kalman filtering is performed on sea state data (predicted by inversion) to obtain sea state filtered data; the method comprises the following specific steps:
(1) Obtaining a predicted state X (t|t-1) =ΦX (t-1) at time t from the state at time t-1
Wherein X (t|t-1) is a predicted state at time t and X (t-1) is a state at time t-1;
(2) Prediction P (t|t-1) =Φp (t-1) Φ, giving covariance matrix T +ΓQ(t)Γ T
Wherein Q is a system noise covariance matrix, P (t|t-1) is a value at the time of T-1 predicted according to the value at the time of T-1, P (T-1) is a value at the time of T-1, and T is a transposition;
(3) Introducing an adjustment update parameter for the adaptive Kalman filtering:
wherein b is a forgetting factor;
(4) Calculating the distance between a predicted value and two observed values (the two observed values are the values of one 2 inversions in the step) according to a nearest correlation rule by utilizing the double-frequency characteristic of the high-frequency ground wave radar, and selecting the observed value with the short distance as the observed value of Kalman filtering;
the observation error is as follows:
v(t)＝Z(t)-H(t)X(t|t-1)
wherein H is an observation matrix, and Z is an observation value;
the observed error v (t) is used to update the noise equation;
the predicted value is determined as: giving data at the moment 1 and the moment 2, predicting data at the moment 3 by adopting a time sequence equation ARIMA (p, i, q), predicting data at the moment 4 by adopting a time sequence equation ARIMA (p, i, q) according to the data at the moment 2 and the moment 3, and predicting data at the moment 5 by adopting a time sequence equation ARIMA (p, i, q) according to the data at the moment 3 and the moment 4;
the observed value is determined as: the observation values can only be one, and the observation values need to be selected from the two observation values, and the selection basis is that the two observation values at the moment are closer to the predicted value at the same moment; the observed value closest to the predicted value will be the observed value of the Kalman filter;
(6) Updating the observed noise covariance matrix:
R(t)＝(1-d t )R(t-1)+d t ([I-H(t)K(t-1)]v(t)v T (t)[I-H(t)K(t-1)] T +H(t)P(t-1)H T (t))
wherein I is an identity matrix, R (t-1) is an observation noise covariance matrix at the time t-1, K (t-1) is a Kalman gain at the time t-1, and v (t) is an observation error;
(6) Calculation of Kalman gain K (t) =P (t|t-1) H T (t)S- 1 (t)
Wherein S (t) =h (t) P (t|t-1) H T (t) +R (t), K (t) is Kalman gain at the time of t-1, P (t|t-1) is an observation noise covariance matrix at the time of t predicted according to an observation noise covariance matrix at the time of t-1, S (t) is an intermediate variable, and H (t) is an observation matrix at the time of t;
(7) Updating covariance matrix P (t) = [ I-K (t) H (t) ] P (t|t-1)
(8) Update state X (t) =x (t|t-1) +k (t) v (t) at time K
(9) The sea state filter data Y (t) is obtained.
Other steps and parameters are the same as those of one of the first to seventh embodiments.
Detailed description nine: this embodiment differs from one to eight embodiments in that the forgetting factor b takes the value of: b is more than or equal to 0.95 and less than or equal to 0.99.
Other steps and parameters are the same as in one to eight of the embodiments.
Detailed description ten: this embodiment differs from one of the first to ninth embodiments in that the sea state filtered data Y (t) =x (t) H (t) in (9).
Other steps and parameters are the same as in one of the first to ninth embodiments.
The following examples are used to verify the benefits of the present invention:
example 1
The data adopted in the embodiment is derived from wind direction data acquired by a radar experiment station at 3 months and 16 days of 2019, wherein the transmitting frequency of radar carrier frequency 1 is 4700khz, and the transmitting frequency of carrier frequency 2 is 5320khz.
The data were taken from the sea element corresponding to the 4 th beam and 30 th range gate, starting at 8 a.m., one batch of data every 8 minutes, for 50 batches of wind direction data. Wherein fig. 2 is a schematic diagram of the comparison result of the dual-frequency adaptive kalman filtering based on time series and the adaptive kalman filtering based on time series. FIG. 3 is a graph showing the comparison of the dual-frequency adaptive Kalman filter and the dual-frequency adaptive Kalman filter based on time sequence.
Looking at the 26 th and 30 th batches of wind direction data in fig. 2, the wind direction of carrier frequency 1 is 35.68 and 12.78 degrees, and the wind direction of carrier frequency 2 is 90 and 110 degrees. According to the trend of the overall change of the data, the wind direction of the carrier frequency 2 can be judged to be a singular value, and then the carrier frequency fusion value is the singular value. The wind directions after the self-adaptive Kalman filtering based on the time sequence are 53.57 degrees and 60.71 degrees by taking the carrier frequency fusion value as an observation value, and the wind directions after the double-frequency self-adaptive Kalman filtering based on the time sequence are 27.72 degrees and 12.79 degrees by taking the carrier frequency 1 and the carrier frequency 2 as the observation values. It can be seen that the adaptive kalman filtering based on time series only refers to one observation value, and the result of wind direction filtering is still not ideal by selecting the observation value under the condition that the carrier frequency fusion value is a singular value. And the double-frequency self-adaptive Kalman filtering based on the time sequence can select normal wind direction data as an observation value to filter according to a nearest neighbor correlation algorithm in two carrier frequencies, and the effect is ideal.
Looking at lot 45 of fig. 3, the wind direction of carrier frequency 1 is 3.2 degrees, the wind direction of carrier frequency 2 is 12.55 degrees, the wind direction of the dual-frequency adaptive kalman filter is 17.47 degrees, and the wind direction of the dual-frequency adaptive kalman filter based on time sequence is 9.05 degrees. It can be seen that the dual-frequency adaptive kalman filtering is not capable of selecting an appropriate observed value because only the wind direction at the last moment is considered and the overall trend of data change is not considered, and the filtering result is not ideal. The time series is combined to obtain a more accurate predicted value which is closer to actual data, and the predicted value can be related to proper observation value filtering, so that the effect is ideal.
Example 2
The data adopted in the embodiment is derived from wave height data acquired by a radar experiment station in 2019, 3 and 16 days, wherein the transmitting frequency of the radar carrier frequency 1 is 4700khz, and the transmitting frequency of the carrier frequency 2 is 5320khz.
The data were taken from sea elements corresponding to the 4 th beam and 30 th range gate, starting at 8 a.m., one batch of data every 8 minutes, for 50 batches of wave height data. Wherein fig. 4 is a schematic diagram of the result of comparing the dual-frequency adaptive kalman filter based on the time series with the adaptive kalman filter based on the time series. Fig. 5 is a schematic diagram of a comparison result of a dual-frequency adaptive kalman filter and a dual-frequency adaptive kalman filter based on a time sequence.
Looking at the 30 th and 32 th batch wave height data of fig. 4, the wave height of carrier frequency 1 is 0.37 and 0.48 m, and the wave height of carrier frequency 2 is 1.1 and 0.87 m. And according to the overall change trend of the data, the wave height of the carrier frequency 2 can be judged to be a singular value, and then the carrier frequency fusion value is the singular value. The wave height of the self-adaptive Kalman filtering based on the time sequence is 0.82 and 0.71 m by taking the carrier frequency fusion value as the observation value, and the wave height of the double-frequency self-adaptive Kalman filtering based on the time sequence is 0.36 and 0.5 m by taking the carrier frequency 1 and the carrier frequency 2 as the observation value. It can be seen that the adaptive kalman filtering based on time series only refers to one observation value, and the result of wave height filtering is still selected as the observation value under the condition that the carrier frequency fusion value is a singular value, and is still not ideal. And the double-frequency self-adaptive Kalman filtering based on the time sequence can select normal wave height data as an observation value to filter according to a nearest neighbor correlation algorithm in two carrier frequencies, and the effect is ideal.
Looking at lot 49 of fig. 5, carrier frequency 1 has a wave height of 0.32 meters, carrier frequency 2 has a wave height of 0.78 meters, dual-frequency adaptive kalman filter has a wave height of 0.73 meters, and time-series-based dual-frequency adaptive kalman filter has a wave height of 0.56 meters. It can be seen that the dual-frequency adaptive kalman filtering is not capable of selecting proper observation values because only the wave height of the last moment is considered and the overall trend of data change is not considered, and the filtering result is not ideal. The time series is combined to obtain a more accurate predicted value which is closer to actual data, and the predicted value can be related to proper observation value filtering, so that the effect is ideal.
According to the data, compared with the time sequence-based self-adaptive Kalman filtering, the time sequence-based double-frequency self-adaptive Kalman filtering can be obtained, when single carrier frequency data has a larger phase difference with a predicted value, namely a singular value appears, the data of another carrier frequency can be selected as observation according to the principle of the latest association, and a better filtering effect can be obtained. Compared with the double-frequency self-adaptive Kalman filtering, the double-frequency self-adaptive Kalman filtering based on the time sequence regards ocean information as a slow-changing process, only the information of the last time point is referred to, and the time sequence is combined with the double-frequency self-adaptive Kalman filtering, so that the information of the past time point can be summarized, the trend of future data change is predicted, and the filtering effect can be better based on the predicted trend.
In conclusion, the method can effectively improve singular values of the sea state inversion of the high-frequency ground wave radar, is beneficial to offshore information monitoring, and has the characteristics of simplicity and convenience in implementation, self-adaptive fusion filtering and the like.
The present invention is capable of other and further embodiments and its several details are capable of modification and variation in light of the present invention, as will be apparent to those skilled in the art, without departing from the spirit and scope of the invention as defined in the appended claims.
Claims (1)
1. The high-frequency ground wave radar sea state data fusion method based on the time sequence is characterized by comprising the following steps of: the method comprises the following specific processes:
step one: performing sea inversion processing on echoes received by the high-frequency ground wave radar to obtain carrier frequency 1 sea state data and carrier frequency 2 sea state data, and fusing the carrier frequency 1 sea state data and the carrier frequency 2 sea state data to obtain fused data;
the sea state data are the flow speed, the flow direction, the wind speed, the wind direction and the sea wave height;
step two: performing stability test on the fused data, and differentiating sequences which do not meet the stability until all the sequences of the fused data meet the stability;
step three: fitting a model to the sequence meeting the stability to obtain a time sequence equation;
step four: subtracting the sequence meeting the stability after the fusion of the step two from the sequence fitted by the time sequence equation to obtain a residual sequence, observing whether the residual sequence meets normal distribution, and when the residual sequence meets the normal distribution, enabling the time sequence equation ARIMA (p, i, q) to be effective; when the residual sequence does not meet the normal distribution, the time sequence equation ARIMA (p, i, q) is invalid;
step five: substituting the sea state data into an effective time sequence equation ARIMA (p, i, q) according to the time sequence, and obtaining a predicted value of the next moment;
step six: obtaining a state transition equation and a noise equation of the adaptive Kalman filtering according to coefficients in an effective time sequence equation ARIMA (p, i, q);
step seven: giving an initial value, and performing self-adaptive Kalman filtering on the sea state data to obtain sea state filtering data;
step eight: sliding a time window with a certain length m according to the actual situation, wherein m is more than or equal to 1 and less than or equal to n, performing sea state inversion processing on echoes received by the to-be-detected high-frequency ground wave radar to obtain carrier frequency 1 sea state data and carrier frequency 2 sea state data, fusing the carrier frequency 1 sea state data and the carrier frequency 2 sea state data to obtain fused data, and repeating the second step to the seventh step to obtain sea state filtering data;
in the first step, sea state inversion processing is carried out on an echo received by a high-frequency ground wave radar to obtain carrier frequency 1 sea state data and carrier frequency 2 sea state data, and the carrier frequency 1 sea state data and the carrier frequency 2 sea state data are fused to obtain fused data;
the n is the length of the data time window;
the specific process is as follows:
the sequence corresponding to the carrier frequency 1 sea state Data is Data 1 (i)，i＝1,2,3...n；
The sequence corresponding to the carrier frequency 2 sea state Data is Data 2 (i)，i＝1,2,3...n；
The signal to noise ratio of the carrier frequency 1 sea state data is Snr 1 (i),i＝1,2,3...n；
The signal to noise ratio of the carrier frequency 2 sea state data is Snr 2 (i),i＝1,2,3,...n；
The n is the length of the data time window;
the carrier frequency 1 sea state Data and the carrier frequency 2 sea state Data are fused to obtain fused Data (i), i=1, 2,3,..n, and the fused Data (i) has the expression:
wherein, is the multiplication;
in the second step, the stability of the fused data is checked, and the sequences which do not meet the stability are differentiated until all the sequences of the fused data meet the stability; the method comprises the following specific steps:
performing stability test on the fused Data (i) by adopting a time sequence diagram method;
if all sequences of the fused Data meet the stability, obtaining a sequence data_stable (i), i=1, 2,3,..n, wherein data_stable (i) =data (i), i=1, 2,3,..n;
if a certain sequence of the fused Data does not meet the stationarity, differentiating the sequence Data (i) which does not meet the stationarity, i=1, 2, 3..n, wherein the sequence after the differentiation is data_diff (i), i=1, 2, 3..n-1, and the like, wherein the Data is the Data
Data_diff(i)＝Data(i+1)-Data(i),i＝1,2,3,..n-1；
Continuously performing stability test on the sequence after the difference, if the stability is not met, continuously performing the difference until the stability is met to obtain a sequence data_stable (i), i=1, 2,3,..m ', and recording the times i', i '=n-m';
wherein data_stable (i) =data_diff (i), i=1, 2,3,..m';
in the third step, fitting a model to the sequence meeting the stability to obtain a time sequence equation; the method comprises the following specific steps:
step three, fitting the sequence data_stable (i) meeting stationarity, i=1, 2,3,..n to a time sequence equation ARIMA (p, i, q);
wherein p is the number of AR model parameters, and q is the number of MA model parameters;
step three, an AIC (automatic identification) order determining criterion is used for determining an ARIMA (p, i, q) model, and the specific process is as follows:
the AIC order criteria are as follows:
different p and q are selected, and corresponding AIC values are calculated, so that the number of coefficients corresponding to the AIC reaching the minimum value is p, and the q takes the best value;
thirdly, estimating the coefficient value of the ARIMA (p, i, q) model after p and q are determined to obtain the coefficient value of the ARIMA (p, i, q) model after p and q are determined;
step three, based on the number of coefficients and coefficient values, carrying out an expression of a time sequence equation ARIMA (p, i, q) to obtain a determined time sequence equation ARIMA (p, i, q);
the expression of the time sequence equation ARIMA (p, i, q) in the step III is as follows:
x(t)＝φ 1 x(t-1)+φ 2 x(t-2)+...+φ p x(t-p)+a(t)+θ 1 a(t-1)+θ 2 a(t-2)+...θ q a(t-q)
wherein x (t) is sea state data at time t, t is time phi 1 、φ 2 、φ p Is sea state data coefficient, a (t), a (t-1), a (t-2), a (t-q) are noise, θ 1 、θ 2 、θ q For the noise coefficient, p is the number of AR model parameters, and q is the number of MA model parameters;
in the third step
Wherein Q (μ, φ) 1 ,...φ p ,θ 1 ,...θ q ) For the sum of the remaining squares of the ARIMA (p, i, q) model, N is the length of the sequence,the residual variance of the sequence, μ is the mean of the sequence;
in the step six, a state transition equation and a noise equation of the adaptive Kalman filtering are obtained according to coefficients in an effective time sequence equation ARIMA (p, i, q); the method comprises the following specific steps:
the time series equation model is obtained by the effective time series equation ARIMA (p, i, 0) as follows:
x(t)＝φ 1 x(t-1)+φ 2 x(t-2)+...+φ p x(t-p)+a(t)
let t=t+1 to give
Let x 1 (t)＝x(t),x 2 (t)＝x(t-1),...,x p (t) =x (t-p+1), then there is
Due to x 2 (t+1)＝x 1 (t),x 3 (t+1)＝x 2 (t),...,x p (t+1)＝x p-1 (t)
Can obtain
Wherein x is 1 (t) is the value of x (t) at the time of t, and x p (t) is the value of the time x (t) of p; x is x 1 (t+1) is the sea state data value at time t+1, x 2 (t+1) is the sea state data value at time t, x p (t+1) is a sea state data value at the time t-p+2;
wherein phi is a state transition matrix, and Γ is a noise matrix;
setting an initial value in the step seven, and performing self-adaptive Kalman filtering on the sea state data to obtain sea state filtering data; the method comprises the following specific steps:
(1) Obtaining a predicted state X (t|t-1) =ΦX (t-1) at time t from the state at time t-1
Wherein X (t|t-1) is a predicted state at time t and X (t-1) is a state at time t-1;
(2) Prediction P (t|t-1) =Φp (t-1) Φ, giving covariance matrix T +ΓQ(t)Γ T
Wherein Q (T) is a system noise covariance matrix, P (t|t-1) is a value of a predicted T moment according to the value of the T-1 moment, P (T-1) is the value of the T-1 moment, and T is a transposition;
(3) Introducing an adjustment update parameter for the adaptive Kalman filtering:
wherein b is a forgetting factor;
(4) Calculating the distance between the predicted value and the two observed values, and selecting the observed value with the short distance as the observed value of Kalman filtering;
the observation error is as follows:
v(t)＝Z(t)-H(t)X(t|t-1)
wherein H (t) is an observation matrix, and Z (t) is an observation value;
(5) Updating the observed noise covariance matrix:
R(t)＝(1-d t )R(t-1)+
d t ([I-H(t)K(t-1)]v(t)v T (t)[I-H(t)K(t-1)] T +H(t)P(t-1)H T (t))
wherein I is an identity matrix, R (t-1) is an observation noise covariance matrix at the time t-1, K (t-1) is a Kalman gain at the time t-1, and v (t) is an observation error;
(6) Calculation of Kalman gain K (t) =P (t|t-1) H T (t)S- 1 (t)
Wherein S (t) =h (t) P (t|t-1) H T (t) +R (t), K (t) is Kalman gain at the time of t-1, P (t|t-1) is an observation noise covariance matrix at the time of t predicted according to an observation noise covariance matrix at the time of t-1, S (t) is an intermediate variable, and H (t) is an observation matrix at the time of t;
(7) Updating covariance matrix P (t) = [ I-K (t) H (t) ] P (t|t-1)
(8) Update state X (t) =x (t|t-1) +k (t) v (t) at time K
(9) Obtaining sea state filtering data Y (t);
the forgetting factor b takes the value of: b is more than or equal to 0.95 and less than or equal to 0.99;
the sea state filtered data Y (t) =x (t) H (t) in (9).
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202110649446.7A CN113391287B (en) | 2021-06-10 | 2021-06-10 | High-frequency ground wave radar sea state data fusion method based on time sequence |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202110649446.7A CN113391287B (en) | 2021-06-10 | 2021-06-10 | High-frequency ground wave radar sea state data fusion method based on time sequence |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN113391287A CN113391287A (en) | 2021-09-14 |
| CN113391287B true CN113391287B (en) | 2023-09-01 |
Family
ID=77620328
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202110649446.7A Active CN113391287B (en) | 2021-06-10 | 2021-06-10 | High-frequency ground wave radar sea state data fusion method based on time sequence |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN113391287B (en) |
Families Citing this family (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN116303786B (en) * | 2023-03-18 | 2023-10-27 | 上海圈讯科技股份有限公司 | Block chain financial big data management system based on multidimensional data fusion algorithm |
| CN116502478B (en) * | 2023-06-29 | 2023-09-01 | 中国电建集团华东勘测设计研究院有限公司 | Submarine topography monitoring-based pile-off auxiliary decision-making method for self-lifting platform |
Citations (21)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| WO2005062984A2 (en) * | 2003-12-24 | 2005-07-14 | Automotive Systems Laboratory, Inc. | Road curvature estimation system |
| CN102650527A (en) * | 2012-05-25 | 2012-08-29 | 北京航空航天大学 | Temperature compensation method for denoising fiber-optic gyroscope on basis of time series analysis |
| EP2793046A1 (en) * | 2011-12-12 | 2014-10-22 | Mitsubishi Electric Corporation | Radar device |
| JP2015052491A (en) * | 2013-09-06 | 2015-03-19 | 沖電気工業株式会社 | Signal processing apparatus, signal processing method and computer program |
| CN105301593A (en) * | 2015-06-12 | 2016-02-03 | 中国电子科技集团公司第四十一研究所 | Flight target altitude estimation method based on dual-frequency RCS information fusion |
| CN105549049A (en) * | 2015-12-04 | 2016-05-04 | 西北农林科技大学 | Adaptive Kalman filtering algorithm applied to GPS navigation |
| CN105807323A (en) * | 2016-03-09 | 2016-07-27 | 哈尔滨工程大学 | Method for determining magnetic target position by using small sub-region identification algorithm |
| CN106707235A (en) * | 2017-03-08 | 2017-05-24 | 南京信息工程大学 | Indoor range finding positioning method based on improved traceless Kalman filtering |
| CN107832259A (en) * | 2017-11-02 | 2018-03-23 | 国网四川雅安电力（集团）股份有限公司 | A kind of load forecasting method based on time series and Kalman filtering |
| CN108169744A (en) * | 2017-12-08 | 2018-06-15 | 中国船舶重工集团公司第七二四研究所 | A kind of ground wave radar and satellite ocean dynamic inversion use processing method |
| CN108983300A (en) * | 2018-06-19 | 2018-12-11 | 长安大学 | A kind of transient electromagnetic Tunnel prediction method under rock tunnel(ling) machine execution conditions |
| CN109884337A (en) * | 2019-03-05 | 2019-06-14 | 哈尔滨工业大学 | A method of wind direction of ocean surface is detected using high-frequency ground wave radar |
| CN110221304A (en) * | 2019-06-04 | 2019-09-10 | 哈尔滨工程大学 | A kind of more sonar data fusion methods of underwater robot |
| CN110954921A (en) * | 2019-12-03 | 2020-04-03 | 浙江大学 | Laser radar echo signal-to-noise ratio improving method based on block matching 3D collaborative filtering |
| CN111896984A (en) * | 2020-07-10 | 2020-11-06 | 自然资源部第一海洋研究所 | GNSS-based real-time high-precision wave measurement method and device |
| CN112270122A (en) * | 2020-10-10 | 2021-01-26 | 清华大学 | Inversion evaluation method for fire source parameters of building fire |
| EP3770628A1 (en) * | 2019-07-24 | 2021-01-27 | Veoneer Sweden AB | A method for radar interference mitigation |
| CN112363122A (en) * | 2020-11-23 | 2021-02-12 | 哈尔滨工业大学（威海） | Extraction method and application of weak harmonic signals in high-frequency ground wave radar ionosphere noise |
| WO2021034784A1 (en) * | 2019-08-16 | 2021-02-25 | Poltorak Technologies, LLC | Device and method for medical diagnostics |
| CN112488427A (en) * | 2020-12-21 | 2021-03-12 | 新疆工程学院 | ARIMA model-based medium-and-long-term photovoltaic available electric quantity prediction method |
| CN112578458A (en) * | 2020-10-19 | 2021-03-30 | 中国石油大学(华东) | Pre-stack elastic impedance random inversion method and device, storage medium and processor |
Family Cites Families (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US8483432B2 (en) * | 2009-12-23 | 2013-07-09 | General Electric Company | Methods for automatic segmentation and temporal tracking |
| US20190204433A1 (en) * | 2017-12-29 | 2019-07-04 | Viettel Group | Method of tracking target by using 2d radar with sensor |
-
2021
- 2021-06-10 CN CN202110649446.7A patent/CN113391287B/en active Active
Patent Citations (21)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| WO2005062984A2 (en) * | 2003-12-24 | 2005-07-14 | Automotive Systems Laboratory, Inc. | Road curvature estimation system |
| EP2793046A1 (en) * | 2011-12-12 | 2014-10-22 | Mitsubishi Electric Corporation | Radar device |
| CN102650527A (en) * | 2012-05-25 | 2012-08-29 | 北京航空航天大学 | Temperature compensation method for denoising fiber-optic gyroscope on basis of time series analysis |
| JP2015052491A (en) * | 2013-09-06 | 2015-03-19 | 沖電気工業株式会社 | Signal processing apparatus, signal processing method and computer program |
| CN105301593A (en) * | 2015-06-12 | 2016-02-03 | 中国电子科技集团公司第四十一研究所 | Flight target altitude estimation method based on dual-frequency RCS information fusion |
| CN105549049A (en) * | 2015-12-04 | 2016-05-04 | 西北农林科技大学 | Adaptive Kalman filtering algorithm applied to GPS navigation |
| CN105807323A (en) * | 2016-03-09 | 2016-07-27 | 哈尔滨工程大学 | Method for determining magnetic target position by using small sub-region identification algorithm |
| CN106707235A (en) * | 2017-03-08 | 2017-05-24 | 南京信息工程大学 | Indoor range finding positioning method based on improved traceless Kalman filtering |
| CN107832259A (en) * | 2017-11-02 | 2018-03-23 | 国网四川雅安电力（集团）股份有限公司 | A kind of load forecasting method based on time series and Kalman filtering |
| CN108169744A (en) * | 2017-12-08 | 2018-06-15 | 中国船舶重工集团公司第七二四研究所 | A kind of ground wave radar and satellite ocean dynamic inversion use processing method |
| CN108983300A (en) * | 2018-06-19 | 2018-12-11 | 长安大学 | A kind of transient electromagnetic Tunnel prediction method under rock tunnel(ling) machine execution conditions |
| CN109884337A (en) * | 2019-03-05 | 2019-06-14 | 哈尔滨工业大学 | A method of wind direction of ocean surface is detected using high-frequency ground wave radar |
| CN110221304A (en) * | 2019-06-04 | 2019-09-10 | 哈尔滨工程大学 | A kind of more sonar data fusion methods of underwater robot |
| EP3770628A1 (en) * | 2019-07-24 | 2021-01-27 | Veoneer Sweden AB | A method for radar interference mitigation |
| WO2021034784A1 (en) * | 2019-08-16 | 2021-02-25 | Poltorak Technologies, LLC | Device and method for medical diagnostics |
| CN110954921A (en) * | 2019-12-03 | 2020-04-03 | 浙江大学 | Laser radar echo signal-to-noise ratio improving method based on block matching 3D collaborative filtering |
| CN111896984A (en) * | 2020-07-10 | 2020-11-06 | 自然资源部第一海洋研究所 | GNSS-based real-time high-precision wave measurement method and device |
| CN112270122A (en) * | 2020-10-10 | 2021-01-26 | 清华大学 | Inversion evaluation method for fire source parameters of building fire |
| CN112578458A (en) * | 2020-10-19 | 2021-03-30 | 中国石油大学(华东) | Pre-stack elastic impedance random inversion method and device, storage medium and processor |
| CN112363122A (en) * | 2020-11-23 | 2021-02-12 | 哈尔滨工业大学（威海） | Extraction method and application of weak harmonic signals in high-frequency ground wave radar ionosphere noise |
| CN112488427A (en) * | 2020-12-21 | 2021-03-12 | 新疆工程学院 | ARIMA model-based medium-and-long-term photovoltaic available electric quantity prediction method |
Non-Patent Citations (1)
| Title |
|---|
| 双频高频地波雷达海态参数反演及数据处理;李晓楠;《《中国优秀硕士学位论文全文数据库（电子期刊）信息科技辑》》;20210115(第1期);第I136-1109页正文第54-61页 * |
Also Published As
| Publication number | Publication date |
|---|---|
| CN113391287A (en) | 2021-09-14 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN113391287B (en) | High-frequency ground wave radar sea state data fusion method based on time sequence | |
| Cornman et al. | A fuzzy logic method for improved moment estimation from Doppler spectra | |
| CN103809173B (en) | Frame CFAR target detection Tracking Integrative method | |
| CN110221256B (en) | SAR interference suppression method based on deep residual error network | |
| CN110749871B (en) | Parameter estimation method of dual-polarization weather radar | |
| US8279113B2 (en) | Method for filtering a radar signal after it has been reflected by a target | |
| CN111856406B (en) | Target detection method and device based on FMCW radar echo | |
| CN111624573A (en) | Time domain self-adaptive target detection method under sea clutter background | |
| CN115657080A (en) | Self-adaptive adjustment method for wind measuring distance gate of laser radar | |
| CN115356702A (en) | Evaporation waveguide deep learning joint inversion method based on marine multi-source radar echo | |
| CN108196238B (en) | Clutter map detection method based on adaptive matched filtering under Gaussian background | |
| CN103616680B (en) | Mobile dim target tracking-before-detecting method based on discrete variable rate sampling | |
| CN114355298A (en) | Radar composite modulation pulse signal identification method | |
| CN112946653A (en) | Dual-polarization meteorological radar signal recovery method and system and storage medium | |
| US8842037B2 (en) | High frequency surfacewave radar | |
| CN113933915B (en) | Short-term and temporary extrapolation forecasting method based on space-time disturbance information interaction integration nesting | |
| CN113848532A (en) | FMCW radar signal noise reduction system and method based on noise reduction model | |
| Chen et al. | CFAR Detection in Nonhomogeneous Weibull Sea Clutter for Skywave OTHR | |
| Zhe et al. | Comparative study on chaos identification of ionospheric clutter from HFSWR | |
| CN111611678A (en) | Typhoon wind direction fuzzy solution optimization method and device, electronic equipment and storage medium | |
| CN115825961A (en) | Self-adaptive selection method for sectional differential propagation phase shift rate of dual-polarization radar | |
| McDonald et al. | Chaotic behaviour and non-linear prediction of airborne radar sea clutter data | |
| CN117233745B (en) | Sea maneuvering target tracking method on non-stationary platform | |
| Yanovsky et al. | Advanced algorithm of velocity measurement for modern meteorological radar | |
| Qiu et al. | CPS attack detection of ships using particle filter |
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 | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |