CN110909863B - Rowland sky-earth wave time delay estimation method based on artificial neural network - Google Patents
Rowland sky-earth wave time delay estimation method based on artificial neural network Download PDFInfo
- Publication number
- CN110909863B CN110909863B CN201910967178.6A CN201910967178A CN110909863B CN 110909863 B CN110909863 B CN 110909863B CN 201910967178 A CN201910967178 A CN 201910967178A CN 110909863 B CN110909863 B CN 110909863B
- Authority
- CN
- China
- Prior art keywords
- neural network
- artificial neural
- layer
- lantian
- luo
- 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
Images
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
- G06N3/04—Architecture, e.g. interconnection topology
- G06N3/045—Combinations of networks
-
- 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
- G01S19/00—Satellite radio beacon positioning systems; Determining position, velocity or attitude using signals transmitted by such systems
- G01S19/01—Satellite radio beacon positioning systems transmitting time-stamped messages, e.g. GPS [Global Positioning System], GLONASS [Global Orbiting Navigation Satellite System] or GALILEO
- G01S19/13—Receivers
- G01S19/20—Integrity monitoring, fault detection or fault isolation of space segment
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
- G06N3/08—Learning methods
- G06N3/084—Backpropagation, e.g. using gradient descent
-
- 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)
- Physics & Mathematics (AREA)
- Theoretical Computer Science (AREA)
- General Physics & Mathematics (AREA)
- General Health & Medical Sciences (AREA)
- General Engineering & Computer Science (AREA)
- Biophysics (AREA)
- Computational Linguistics (AREA)
- Data Mining & Analysis (AREA)
- Evolutionary Computation (AREA)
- Artificial Intelligence (AREA)
- Molecular Biology (AREA)
- Computing Systems (AREA)
- Biomedical Technology (AREA)
- Life Sciences & Earth Sciences (AREA)
- Mathematical Physics (AREA)
- Software Systems (AREA)
- Health & Medical Sciences (AREA)
- Remote Sensing (AREA)
- Radar, Positioning & Navigation (AREA)
- Computer Security & Cryptography (AREA)
- Computer Networks & Wireless Communication (AREA)
- Geophysics And Detection Of Objects (AREA)
Abstract
The invention discloses a Rowland sky-earth wave time delay estimation method based on an artificial neural network, which specifically comprises the steps of designing an artificial neural network structure, collecting samples, preprocessing, training the artificial neural network by adopting the preprocessed samples to obtain a trained artificial neural network, and substituting Luo Lantian earth wave signals to be estimated in the trained artificial neural network to output Luo Lantian earth wave time delay. The method for estimating the Rowland sky-earth wave time delay has higher precision, and can be normally used in a low signal-to-noise ratio environment.
Description
Technical Field
The invention belongs to the technical field of digital signal processing, and particularly relates to a Rowland sky-earth wave time delay estimation method based on an artificial neural network.
Background
The Roland C system is a ground-based long-wave wireless navigation system and is used as a backup of a satellite navigation system for a long time. With the rapid development of radio technology in recent years, the positioning accuracy and the use condition of the rowland system also need to be further improved. The Roland signal is a low-frequency electromagnetic wave signal, is greatly influenced by the topography and the ionosphere, has a low signal-to-noise ratio of signals received by cities which are on inland far away from coasts, and cannot be normally used by the existing space-to-ground wave delay estimation algorithm. The artificial neural network algorithm is a machine learning algorithm, a mathematical model is established by simulating the structure of the human brain, parameters are adjusted to approximate to an actual model through a large amount of data learning, the design is reasonable, any nonlinear system can be simulated, the algorithm is used for estimating the Rowland sky-earth wave time delay, the accuracy is high, the method can be normally used in a low signal-to-noise ratio environment, and the existing Rowland sky-earth wave time delay is not effectively estimated and calculated by using the artificial neural network.
Disclosure of Invention
The invention aims to provide a Rowland earth wave time delay estimation method based on an artificial neural network, which can estimate Rowland earth wave time delay under the condition of low signal-to-noise ratio.
The technical scheme adopted by the invention is that the Rowland sky-earth wave time delay estimation method based on the artificial neural network is implemented according to the following steps:
step 1, designing the structure of an artificial neural network, collecting Luo Lantian ground wave signal samples, and preprocessing Luo Lantian ground wave signal samples for training the artificial neural network according to the length and the period of a Roland signal;
step 2, setting initial parameters of an artificial neural network, wherein the initial parameters comprise weight and bias of each layer of artificial neural network, learning rate of the artificial neural network and activation functions among nerve cells of each layer, and setting training times of Luo Lantian ground wave signal samples and a target threshold of a training error function;
step 3, training the artificial neural network by using the preprocessed Luo Lantian ground wave signal sample, and stopping training when the error function value is smaller than a threshold value or the learning times reach a maximum set value, so as to obtain the trained artificial neural network;
and step 4, substituting Luo Lantian ground wave signals with time delay to be estimated into a trained artificial neural network to output Luo Lantian ground wave time delay.
The present invention is also characterized in that,
the step 1 is specifically implemented according to the following steps:
step 1.1, designing an artificial neural network structure as 5 layers, wherein the artificial neural network structure comprises an input layer, a first hidden layer, a second hidden layer, a third hidden layer and an output space wave delay layer, the input vector of the input layer is 5000x1, the number of neurons of the first hidden layer is 3000 neurons, the number of neurons of the second hidden layer is 10000 neurons, the number of neurons of the third hidden layer is 3000 neurons, and the output vector of the output space wave delay layer is 2x1;
step 1.2, collecting Luo Lantian ground wave signal samples, wherein the sampling frequency is 10MHz, the single pulse repetition period of the Roland signal is 1 millisecond, and the length of a cut 5000 point is 500 microseconds;
and step 1.3, planning the sample value of the Luo Lantian ground wave signal to the [ -1,1] interval for normalization processing to obtain a Luo Lantian ground wave signal pretreatment sample.
In the step 2, the weight and the bias value of each layer of artificial neural network are set to be 0.5, the learning rate of the artificial neural network is 0.01, the activation function among the layers of neurons is a sigmoid function, the training times of Luo Lantian ground wave signal samples are 300 times, and the training error function target threshold of the Rowland sky wave signal samples, namely the error function threshold value, is 1e-4.
Step 3, inputting a Luo Lantian ground wave signal preprocessing sample by an input layer, training by a first hidden layer, a second hidden layer and a third hidden layer, and outputting by an output layer, wherein the steps are implemented specifically as follows:
step 3.1, forward output
Calculating the output vector O of each hidden layer of Luo Lantian ground wave signal samples according to the following method j ,
O j =f(∑x n w nj +θ j ) (1)
In the formula (1), j represents a hidden layer, j=1, 2,3 represent a first hidden layer and a second hidden layer, respectively, and x n Represents a sample of the Rowland sky-ground wave signal, n is a natural number other than 0, w is a weight, θ represents a bias, and f (Σx) n w nj +θ j ) Is an activation function f (a), and
step 3.2, reverse error propagation
Calculating an error value Err of the output layer according to the following formula k And error value Err of each hidden layer j ,
Err k =O k (1-O k )(T k -O k ) (3)
Err j =O j (1-O j )∑Err k w jk (4)
In the formulas (3) and (4), k represents an output layer, T k Representing the target vector of the output layer, O k An output vector representing an output layer;
step 3.3, parameter updating
Error value Err of hidden layer of each layer calculated according to step 3.2 j The weights and offsets are updated and the weights and offsets are updated,
updated weight w ij =w nj +ηErr j O j (5)
Updated offset θ j =θ j +ηErr j (6)
In the formulas (5) and (6), i represents an input layer, and η represents a learning rate;
step 3.4, training the artificial neural network again according to the updated weight and bias, namely substituting the updated weight and bias into the operation of executing the steps 3.1-3.3, and continuously and circularly training until the output layer error value Err k And (3) training times less than 1e-4 or reaching the training times set in the step (2), and stopping training.
The time delay of the ground wave and the time delay of the sky wave in the Luo Lantian ground wave signal samples acquired in the step 1 are between 30 and 200 microseconds, the signal to noise ratio is-15 dB to 15dB, and the number of Luo Lantian ground wave signal samples is not less than 10000.
The method has the advantages that the method for estimating the Rowland sky-earth wave time delay based on the artificial neural network applies the WRELAX algorithm and the invariance principle, converts the multidimensional optimization problem into the repeated one-dimensional optimization problem, is simple and clear, and is convenient to operate; the time delay error estimation of the Rowland system is accurate, and the positioning accuracy of the Rowland system is improved.
Drawings
FIG. 1 is a block diagram of an artificial neural network in a Rowland earth delay estimation method based on the artificial neural network;
FIG. 2 is a basic algorithm block diagram of an artificial neural network in a Rowland earth wave delay estimation method based on the artificial neural network;
fig. 3 is an error histogram of the performance of the artificial neural network after training under different signal-to-noise ratios, and fig. 3 (a), (b), and (c) are the rowland earth wave time delay estimation error histograms of the artificial neural network after training under environments with signal-to-noise ratios of 10dB, 0dB, and-10 dB, respectively.
Detailed Description
The present invention will be described in detail with reference to the accompanying drawings and detailed description.
The invention discloses a Rowland sky-earth wave time delay estimation method based on an artificial neural network, which is implemented according to the following steps:
step 1, designing the structure of an artificial neural network, collecting Luo Lantian ground wave signal samples, and preprocessing Luo Lantian ground wave signal samples for training the artificial neural network according to the length and the period of a Roland signal:
step 1.1, as shown in fig. 1, designing an artificial neural network structure as 5 layers, wherein the artificial neural network structure comprises an input layer, a first hidden layer, a second hidden layer, a third hidden layer and an output space wave delay layer, the input vector of the input layer is 5000x1, the neuron number of the first hidden layer is 3000 neurons, the neuron number of the second hidden layer is 10000 neurons, the neuron number of the third hidden layer is 3000 neurons, and the output vector of the output space wave delay layer is 2x1;
step 1.2, collecting Luo Lantian ground wave signal samples, wherein the ground wave time delay and the sky wave time delay are between 30 and 200 microseconds, the signal to noise ratio is-15 dB to 15dB, the number of Luo Lantian ground wave signal samples is not less than 10000, the sampling frequency is 10MHz, the single pulse repetition period of the Roland signal is 1 millisecond, and the length of a cut 5000 point is 500 microseconds;
and step 1.3, planning the sample value of the Luo Lantian ground wave signal to the [ -1,1] interval for normalization processing to obtain a Luo Lantian ground wave signal pretreatment sample.
Step 2, setting initial parameters of the artificial neural network, namely setting the weight and the bias weight of each layer of the artificial neural network to be 0.5, the learning rate of the artificial neural network to be 0.01, and the activation function among the neurons of each layer to be a sigmoid function, and setting the training times of Luo Lantian ground wave signal samples to be 300 times and the target threshold of the training error function, namely the error function threshold to be 1e-4.
Step 3, training the artificial neural network by using the rowland sky-ground wave signal sample, as shown in fig. 2, inputting a Luo Lantian ground wave signal pretreatment sample by an input layer, training by a first hidden layer, a second hidden layer and a third hidden layer, outputting by an output layer, and stopping training when an error function is smaller than a threshold value or the learning times reaches a maximum set value, thereby obtaining the trained artificial neural network:
step 3.1, forward output
Calculating the output vector O of each hidden layer of Luo Lantian ground wave signal samples according to the following method j ,
O j =f(∑x n w nj +θ j ) (1)
In the formula (1), j represents a hidden layer, j=1, 2,3 represent a first hidden layer and a second hidden layer, respectively, and x n Represents a sample of the Rowland sky-ground wave signal, n is a natural number other than 0, w is a weight, θ represents a bias, and f (Σx) n w nj +θ j ) Is an activation function f (a), and
step 3.2, reverse error propagation
Calculating an error value Err of the output layer according to the following formula k And error value Err of each hidden layer j ,
Err k =O k (1-O k )(T k -O k ) (3)
Err j =O j (1-O j )∑Err k w jk (4)
In the formulas (3) and (4), k represents an output layer, O k Output vector representing output layer, T k A target vector representing an output layer;
step 3.3, parameter updating
Error value Err of hidden layer of each layer calculated according to step 3.2 j The weights and offsets are updated and the weights and offsets are updated,
updated weight w ij =w nj +ηErr j O j (5)
Updated offset θ j =θ j +ηErr j (6)
In the formulas (5) and (6), i represents an input layer, and η represents a learning rate;
step 3.4, training the artificial neural network again according to the updated weight and bias, namely substituting the updated weight and bias into the operation of executing the steps 3.1-3.3, and continuously and circularly training until the output layer error value Err k And (3) training times less than 1e-4 or reaching the training times set in the step (2), and stopping training.
And step 4, substituting Luo Lantian ground wave signals with time delay to be estimated into a trained artificial neural network to output Luo Lantian ground wave time delay.
In order to ensure that the performance of the trained artificial neural network is good, 90% of the Luo Lantian ground wave signal pretreatment samples are marked in the step 1.3 to serve as training samples, the rest 10% are served as test samples, the step 2-step 3 is executed through the training samples to obtain the trained artificial neural network, the test samples are input into the trained artificial neural network, and Luo Lantian ground wave time delay of the test samples is output. Evaluating the performance of the trained artificial neural network according to the output result, wherein 90% of the time delay estimation error of the Rowland sky-earth wave is concentrated in an interval less than 0.1 microsecond in an environment with signal-to-noise ratio SNR=10 dB as shown in FIG. 3 (a); as shown in fig. 3 (b), in the snr=0 dB environment, 75% of the rowland earth wave delay estimation error is concentrated in the interval of less than 1 microsecond, and 90% is concentrated in the interval of less than 2 microseconds; as shown in fig. 3 (c), in the snr= -10dB environment, more than 50% of the rowland earth wave delay estimation error is within 5 microseconds, so that it can be found that the artificial neural network trained in the rowland earth wave delay estimation method based on the artificial neural network can still be used in the environment with low signal-to-noise ratio, and the accuracy of the Luo Lantian earth wave delay estimation value is high. And the Luo Lantian ground wave signal of the time delay to be estimated is substituted into the trained and estimated artificial neural network, and the Luo Lantian ground wave time delay can be obtained.
Claims (5)
1. The method for estimating the Rowland earth wave time delay based on the artificial neural network is characterized by comprising the following steps of:
step 1, designing the structure of an artificial neural network, collecting Luo Lantian ground wave signal samples, and preprocessing Luo Lantian ground wave signal samples for training the artificial neural network according to the length and the period of a Roland signal;
step 2, setting initial parameters of an artificial neural network, wherein the initial parameters comprise weight and bias of each layer of artificial neural network, learning rate of the artificial neural network and activation functions among nerve cells of each layer, and setting training times of Luo Lantian ground wave signal samples and a target threshold of a training error function;
step 3, training the artificial neural network by using the Roland sky wave signal sample, and stopping training when the error function is smaller than the threshold value or the learning times reach the maximum set value, so as to obtain the trained artificial neural network;
and step 4, substituting Luo Lantian ground wave signals with time delay to be estimated into a trained artificial neural network to output Luo Lantian ground wave time delay.
2. The method for estimating the time delay of the rowland wave based on the artificial neural network according to claim 1, wherein the step 1 is specifically implemented according to the following steps:
step 1.1, designing an artificial neural network structure as 5 layers, wherein the artificial neural network structure comprises an input layer, a first hidden layer, a second hidden layer, a third hidden layer and an output space wave delay layer, the input vector of the input layer is 5000x1, the number of neurons of the first hidden layer is 3000 neurons, the number of neurons of the second hidden layer is 10000 neurons, the number of neurons of the third hidden layer is 3000 neurons, and the output vector of the output space wave delay layer is 2x1;
step 1.2, collecting Luo Lantian ground wave signal samples, wherein the sampling frequency is 10MHz, the single pulse repetition period of the Roland signal is 1 millisecond, and the length of a cut 5000 point is 500 microseconds;
and step 1.3, planning the sample value of the Luo Lantian ground wave signal to the [ -1,1] interval for normalization processing to obtain a Luo Lantian ground wave signal pretreatment sample.
3. The method for estimating the time delay of the rowland wave based on the artificial neural network according to claim 1, wherein in the step 2, the weight and the bias value of each layer of the artificial neural network are set to be 0.5, the learning rate of the artificial neural network is 0.01, the activation function among the neurons of each layer is a sigmoid function, the training times of Luo Lantian ground wave signal samples are 300 times, and the training error function target threshold of the rowland wave signal samples, namely the error function threshold value is 1e-4.
4. The method for estimating the time delay of the rowland wave based on the artificial neural network according to claim 2, wherein the step 3 is to input Luo Lantian ground wave signal preprocessing samples from an input layer, train the samples through a first hidden layer, a second hidden layer and a third hidden layer, and output the samples from an output layer, and is specifically implemented according to the following steps:
step 3.1, forward output
Calculating the output vector O of each hidden layer of Luo Lantian ground wave signal samples according to the following method j ,
O j =f(∑x n w nj +θ j ) (1)
In the formula (1), j represents a hidden layer, j=1, 2,3 represent a first hidden layer and a second hidden layer, respectively, and x n Represents a sample of the Rowland sky-ground wave signal, n is a natural number other than 0, w is a weight, θ represents a bias, and f (Σx) n w nj +θ j ) Is an activation function f (a), and
step 3.2, reverse error propagation
Calculating an error value Err of the output layer according to the following formula k And error value Err of each hidden layer j ,
Err k =O k (1-O k )(T k -O k ) (3)
Err j =O j (1-O j )∑Err k w jk (4)
In the formulas (3) and (4), k represents an output layer, T k Representing the target vector of the output layer, O k An output vector representing an output layer;
step 3.3, parameter updating
Error value Err of hidden layer of each layer calculated according to step 3.2 j The weights and offsets are updated and the weights and offsets are updated,
updated weight w ij =w nj +ηErr j O j (5)
Updated offset θ j =θ j +ηErr j (6)
In the formulas (5) and (6), i represents an input layer, and η represents a learning rate;
step 3.4, training the artificial neural network again according to the updated weight and bias, namely substituting the updated weight and bias into the operation of executing the steps 3.1-3.3, and continuously and circularly training until the output layer error value Err k And (3) training times less than 1e-4 or reaching the training times set in the step (2), and stopping training.
5. The method for estimating the time delay of the rowland wave based on the artificial neural network according to claim 1, wherein the time delay of the land wave and the time delay of the sky wave in the Luo Lantian land wave signal samples collected in the step 1 are between 30 and 200 microseconds, the signal-to-noise ratio is-15 dB to 15dB, and the number of the Luo Lantian land wave signal samples is not less than 10000.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201910967178.6A CN110909863B (en) | 2019-10-12 | 2019-10-12 | Rowland sky-earth wave time delay estimation method based on artificial neural network |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201910967178.6A CN110909863B (en) | 2019-10-12 | 2019-10-12 | Rowland sky-earth wave time delay estimation method based on artificial neural network |
Publications (2)
Publication Number | Publication Date |
---|---|
CN110909863A CN110909863A (en) | 2020-03-24 |
CN110909863B true CN110909863B (en) | 2023-04-21 |
Family
ID=69815513
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN201910967178.6A Active CN110909863B (en) | 2019-10-12 | 2019-10-12 | Rowland sky-earth wave time delay estimation method based on artificial neural network |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN110909863B (en) |
Families Citing this family (3)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN113433536A (en) * | 2021-05-27 | 2021-09-24 | 中国船舶重工集团公司第七0七研究所 | Method for identifying Loran C sky and ground waves based on MEDLL |
CN113644999B (en) * | 2021-06-18 | 2024-01-30 | 西安理工大学 | Luo Lantian ground wave separation method based on Levenberg-Marquart algorithm |
CN116232282B (en) * | 2023-01-12 | 2023-12-19 | 湖南大学无锡智能控制研究院 | Time-varying time delay estimation method, device and system based on adaptive all-pass filter |
Family Cites Families (2)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
EP2122542B1 (en) * | 2006-12-08 | 2017-11-01 | Medhat Moussa | Architecture, system and method for artificial neural network implementation |
CN103581188B (en) * | 2013-11-05 | 2016-08-03 | 中国科学院计算技术研究所 | A kind of network security situation prediction method and system |
-
2019
- 2019-10-12 CN CN201910967178.6A patent/CN110909863B/en active Active
Also Published As
Publication number | Publication date |
---|---|
CN110909863A (en) | 2020-03-24 |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN110909863B (en) | Rowland sky-earth wave time delay estimation method based on artificial neural network | |
CN109993280B (en) | Underwater sound source positioning method based on deep learning | |
Reynolds et al. | Wind speed estimation from CYGNSS using artificial neural networks | |
CN108318862B (en) | Sound source positioning method based on neural network | |
CN104820993B (en) | It is a kind of to combine particle filter and track the underwater weak signal target tracking for putting preceding detection | |
CN111639747B (en) | GNSS-R sea surface wind speed inversion method and system based on BP neural network | |
CN109782231A (en) | A kind of end-to-end sound localization method and system based on multi-task learning | |
CN110213003B (en) | Wireless channel large-scale fading modeling method and device | |
CN104950296B (en) | Based on the remaining sane non-homogeneous detection method of weight Weighted adaptive power | |
CN109658948B (en) | Migratory bird migration activity-oriented acoustic monitoring method | |
CN103679199B (en) | Independent component analysis method of making an uproar based on invasive weeds algorithm | |
CN106291607B (en) | GNSS multipath signal simulation generation method and system | |
CN111639746A (en) | GNSS-R sea surface wind speed inversion method and system based on CNN neural network | |
CN116148863A (en) | Multi-mode deep learning method for satellite-borne GNSS-R sea surface wind speed inversion based on CNN | |
CN109190640A (en) | A kind of the intercept type acquisition method and acquisition system of the planktonic organism based on big data | |
Tafazoli et al. | Performance improvement of GPS GDOP approximation using recurrent wavelet neural network | |
CN113093225B (en) | Wide-area and local-area fused high-precision ionospheric scintillation model establishment method | |
CN115097378A (en) | Incoherent scattering source detection and positioning method based on convolutional neural network | |
CN115659138A (en) | Underwater measurement signal noise reduction method based on dynamic water interference environment electromagnetic method | |
Morid et al. | Solar radiation estimation using temperature-based, stochastic and artificial neural networks approaches | |
CN111736127B (en) | Source domain generation and distribution parameter generalization method for unknown sea clutter | |
CN114936570A (en) | Interference signal intelligent identification method based on lightweight CNN network | |
Gao et al. | Retrieval of ocean wind speed using airborne reflected GNSS signals | |
Xie et al. | MF-ANN: A Novel Artificial Neural Network based Method for Ocean Wind Speed Retrieval on Spaceborne GNSS-R Signal | |
CN113534085B (en) | Sea surface wind speed and effective wave height joint inversion method of interference imaging altimeter |
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 |