CN115308817A - Signal learning and dynamic determination method based on reference signal characteristics - Google Patents
Signal learning and dynamic determination method based on reference signal characteristics Download PDFInfo
- Publication number
- CN115308817A CN115308817A CN202210944169.7A CN202210944169A CN115308817A CN 115308817 A CN115308817 A CN 115308817A CN 202210944169 A CN202210944169 A CN 202210944169A CN 115308817 A CN115308817 A CN 115308817A
- Authority
- CN
- China
- Prior art keywords
- reference signal
- learning
- gate
- signal
- representing
- 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.)
- Granted
Links
- 238000000034 method Methods 0.000 title claims abstract description 27
- 238000013528 artificial neural network Methods 0.000 claims description 5
- QVGXLLKOCUKJST-UHFFFAOYSA-N atomic oxygen Chemical compound [O] QVGXLLKOCUKJST-UHFFFAOYSA-N 0.000 claims description 3
- 230000006870 function Effects 0.000 claims description 3
- 239000007788 liquid Substances 0.000 claims description 3
- 229910052760 oxygen Inorganic materials 0.000 claims description 3
- 239000001301 oxygen Substances 0.000 claims description 3
- XLYOFNOQVPJJNP-UHFFFAOYSA-N water Substances O XLYOFNOQVPJJNP-UHFFFAOYSA-N 0.000 claims description 3
- 230000008859 change Effects 0.000 abstract description 11
- 210000004027 cell Anatomy 0.000 description 8
- 230000000694 effects Effects 0.000 description 2
- 238000005516 engineering process Methods 0.000 description 2
- 239000000203 mixture Substances 0.000 description 2
- 238000012986 modification Methods 0.000 description 2
- 230000004048 modification Effects 0.000 description 2
- 210000002569 neuron Anatomy 0.000 description 2
- 230000008569 process Effects 0.000 description 2
- 238000006467 substitution reaction Methods 0.000 description 2
- 238000010521 absorption reaction Methods 0.000 description 1
- 238000013459 approach Methods 0.000 description 1
- 238000004891 communication Methods 0.000 description 1
- 238000013461 design Methods 0.000 description 1
- 238000011161 development Methods 0.000 description 1
- 238000010586 diagram Methods 0.000 description 1
- 238000001914 filtration Methods 0.000 description 1
- 229910052731 fluorine Inorganic materials 0.000 description 1
- 125000001153 fluoro group Chemical group F* 0.000 description 1
- 230000007774 longterm Effects 0.000 description 1
- 238000005259 measurement Methods 0.000 description 1
- 238000010295 mobile communication Methods 0.000 description 1
- 238000012544 monitoring process Methods 0.000 description 1
- 239000013589 supplement Substances 0.000 description 1
Images
Classifications
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01W—METEOROLOGY
- G01W1/00—Meteorology
- G01W1/14—Rainfall or precipitation gauges
-
- 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
- Life Sciences & Earth Sciences (AREA)
- Environmental & Geological Engineering (AREA)
- Hydrology & Water Resources (AREA)
- Engineering & Computer Science (AREA)
- Atmospheric Sciences (AREA)
- Biodiversity & Conservation Biology (AREA)
- Ecology (AREA)
- Environmental Sciences (AREA)
- Radar Systems Or Details Thereof (AREA)
Abstract
The invention discloses a signal learning and dynamic determination method based on reference signal characteristics, which comprises the following steps: s1, constructing a reference signal learning and tracking network; s2, acquiring a reference signal in a rainless period; s3, updating parameters of the reference signal learning and tracking network based on the reference signal; s4, acquiring a reference signal of a rainfall period, and ending the updating; s5, repeating the steps S1-S3 after rainfall is finished; and (4) executing S4 again when rainfall begins, and finally finishing the signal learning and dynamic determination of the reference signal. The invention can master the change rule of the reference signal by learning and tracking the characteristics of the reference signal in the rainless period, thereby realizing the dynamic determination of the reference signal in the rainfall period.
Description
Technical Field
The invention relates to the technical field of microwave rain measurement signal analysis and application, in particular to a signal learning and dynamic determination method based on reference signal characteristics.
Background
The rapid development of the mobile communication technology enables people, data and things to be closely connected, and the mature microwave communication technology provides a new bridge and a technical means for monitoring meteorological information. At present, a new method for measuring rainfall by using a microwave link is used as a supplement to traditional means such as a rain measuring cylinder, a weather radar and an meteorological satellite, and has been widely concerned by scholars at home and abroad. When a microwave signal passes through a rainfall area, attenuation of microwave energy is caused by absorption and scattering of raindrops, and a new method for measuring rainfall of a microwave link is to invert rainfall intensity based on rainfall attenuation, so that the key for measuring rainfall of the microwave link is to obtain real-time accurate rainfall attenuation.
At present, the method for acquiring the rain attenuation by the microwave link is to calculate the reference signal in the rainfall period by using signal interpolation before and after rainfall, and then to obtain the rain attenuation by differentiating the reference signal with the actually measured signal. However, this approach suffers from two significant problems: on one hand, the signal of the microwave link does not change in a simple linear way; on the other hand, the reference signal is obtained by means of interpolation, so that the rainfall inversion has obvious time delay, and the problem is more obvious particularly for long-time rainfall.
Disclosure of Invention
Aiming at the problems, the invention provides a signal learning and dynamic determination method based on reference signal characteristics, which aims to solve the technical problems in the prior art, and can grasp the change rule of a reference signal by learning and tracking the reference signal characteristics in a rainless period so as to realize the dynamic determination of the reference signal in a rainfall period.
In order to achieve the purpose, the invention provides the following scheme:
the invention provides a signal learning and dynamic determination method based on reference signal characteristics, which comprises the following steps:
s1, constructing a reference signal learning and tracking network;
s2, acquiring a reference signal in a rainless period;
s3, updating parameters of the reference signal learning and tracking network based on the reference signal;
s4, acquiring a reference signal of a rainfall period, and ending the updating;
s5, repeating the steps S1-S3 after rainfall is finished; and executing the S4 again when the rainfall begins, and finally finishing the signal learning and dynamic determination of the reference signal.
Preferably, the reference signal learning and tracking network in S1 includes a long-term neural network, a generalized regression network, and a kalman filter network.
Preferably, the reference signal in the rain-free period in S2 includes a microwave link and a satellite-ground link.
Preferably, the reference signal AC of the rain-free period in S2 sun (t) is expressed as:
AC sun (t)＝A O (t)+A V (t)+A C (t)+A S (t)+A other +C
in the formula, A o Is the oxygen decay; a. The v Moisture attenuation; a. The c Is liquid water decay in the cloud; a. The s Is the flicker attenuation; a. The other Attenuation due to other factors; c is a constant.
Preferably, the updating process of the parameters in S3 is: and inputting the reference signal into the reference signal learning and tracking network, acquiring a change characteristic by learning the reference signal, and updating the parameter of the reference signal learning and tracking network based on the change characteristic.
Preferably, the reference signal comprises a signal-to-noise ratio and a level value.
Preferably, the variation characteristic comprises a time-series variation characteristic of the reference signal during the rainless period and the rainfall period.
Preferably, the updating of the parameters in S3 is represented as:
in the formula (f) t Indicating a forgotten door; g t Represents an update gate; i.e. i t Representing an input gate; c. C t Indicating the state of the cell at time t; c. C t-1 Represents the state of the cell at time t-1; h is a total of t Represents the signal strength at time t; h is t-1 Represents the signal strength at time t-1; o. o t An output gate is shown; w is a group of xf A weight representing a visible layer input in a forgotten door; w hf Representing weights of hidden layer outputs in a forgetting gate; w xc A weight representing the visible layer input in the update gate; w is a group of hc A weight representing the output of the hidden layer in the update gate; w is a group of xi Representing the weight of the visible layer input in the input gate; w is a group of hi Weights representing hidden layer outputs in the input gate; b f A deviation indicating a forgotten door; b is a mixture of c A deviation representing a state of the cell; b i Indicating the deviation of the input gate; x is the number of t Representing the reference signal value AC sun (t), σ (·) denotes a sigmoid function.
The invention discloses the following technical effects:
compared with the existing reference signal determination method, the reference signal characteristic-based signal learning and dynamic determination method provided by the invention has the advantages of high precision, small time delay and the like compared with the existing method by learning the time sequence change rule of the reference signal, mastering the change rule of the reference signal and only utilizing the signal value before rainfall to realize the dynamic accurate determination of the reference signal in the rainfall period.
Drawings
In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings required in the embodiments will be briefly described below, and it is obvious that the drawings in the following description are only some embodiments of the present invention, and it is obvious for those skilled in the art that other drawings can be obtained according to these drawings without creative efforts.
FIG. 1 is a flow chart of a method of the present invention;
FIG. 2 is a technical roadmap in an embodiment of the present invention;
fig. 3 is a learning network design diagram in an embodiment of the invention.
Detailed Description
The technical solutions in the embodiments of the present invention will be clearly and completely described below with reference to the drawings in the embodiments of the present invention, and it is obvious that the described embodiments are only a part of the embodiments of the present invention, and not all of the embodiments. All other embodiments, which can be derived by a person skilled in the art from the embodiments given herein without making any creative effort, shall fall within the protection scope of the present invention.
In order to make the aforementioned objects, features and advantages of the present invention more comprehensible, the present invention is described in detail with reference to the accompanying drawings and the detailed description thereof.
Referring to fig. 1-2, the present embodiment provides a method for learning and dynamically determining a signal based on reference signal characteristics, comprising the following steps:
s1, constructing a reference signal learning and tracking network Net, as shown in fig. 3. The reference signal learning and tracking network Net comprises a long-time neural network, a short-time neural network, a generalized regression network and a Kalman filtering network.
Taking a long-time neural network as an example, a plurality of LSTM neurons are established in sequence, and each neuron comprises a forgetting gate, an updating gate and an output gate. When the rain-making device works, the forgetting gate deletes useless information in the unit state c (t-1) at the previous moment according to the actually-measured attenuation baseline value received at the current moment and the previous signal data h (t-1), and only h (t-1) is considered during rainfall. Based on the data received by the current cell, the update gate decides which useful information to add to c (t). Through the first two gates, the output gate generates the current cell state c (t) and signal information h (t) while predicting the attenuation baseline value.
And S2, acquiring a reference signal in a rainless period from the signal receiver. The reference signal in the rainless period comprises a microwave link and a satellite-ground link. The reference signals in the rainless period comprise a microwave link and a satellite-ground link.
Wherein the reference signal AC during the rainless period sun (t) is expressed as:
AC sun (t)＝A O (t)+A V (t)+A C (t)+A S (t)+A other +C
in the formula, A o Is the oxygen decay; a. The v Moisture attenuation; a. The c Is liquid water decay in the cloud; a. The s Is the scintillation decay; a. The other Is other factorsThe resulting attenuation; c is a constant.
And S3, updating the parameters of the reference signal learning and tracking network Net based on the reference signal.
The updating process of the parameters is as follows: firstly, inputting reference signals such as signal-to-noise ratio and level value of a microwave link or a satellite-to-ground link into a reference signal learning and tracking network Net; determining time sequence change characteristics of the reference signal in a rainless period and a rainfall period by learning the reference signal; and updating the parameters of the reference signal learning and tracking network Net based on the time sequence change characteristics.
Wherein the updating of the parameters is represented as:
in the formula (f) t Indicating a forgotten door; g t Represents an update gate; i.e. i t Representing an input gate; c. C t Indicating the cell state at time t; c. C t-1 Represents the state of the cell at time t-1; h is a total of t Represents the signal strength at time t; h is t-1 Represents the signal strength at time t-1; o t An output gate is shown; w xf A weight representing a visible layer input in a forgotten door; w hf A weight representing the output of a hidden layer in the forgetting gate; w xc A weight representing the visible layer input in the update gate; w hc Representing weights for hidden layer outputs in the update gate; w xi Representing the weight of the visible layer input in the input gate; w hi Weights representing hidden layer outputs in the input gate; b is a mixture of f A deviation indicating a forgotten door; b c A deviation representing a state of the cell; b i Indicating the deviation of the input gate; x is a radical of a fluorine atom t Representing the reference signal value AC sun (t) of (d). σ (-) denotes the sigmoid function, as follows:
s4, dynamically determining a reference signal AC according to time sequence through a reference signal learning and tracking network Net in a rainfall period refer (t) when the measured signal AC is not inputted any more rain (t)。
S5, repeating S1-S3 after rainfall is finished; and (4) when the rainfall starts, executing S4 again, and finally finishing the signal learning and dynamic determination of the reference signal.
The invention discloses the following technical effects:
compared with the existing reference signal determination method, the reference signal characteristic-based signal learning and dynamic determination method provided by the invention has the advantages of high precision, small time delay and the like compared with the existing method by learning the time sequence change rule of the reference signal, mastering the change rule of the reference signal and only utilizing the signal value before rainfall to realize the dynamic accurate determination of the reference signal in the rainfall period.
Finally, it should be noted that: the above-mentioned embodiments are only specific embodiments of the present invention, which are used for illustrating the technical solutions of the present invention and not for limiting the same, and the protection scope of the present invention is not limited thereto, although the present invention is described in detail with reference to the foregoing embodiments, those skilled in the art should understand that: those skilled in the art can still make modifications or changes to the embodiments described in the foregoing embodiments, or make equivalent substitutions for some features, within the scope of the disclosure; such modifications, changes or substitutions do not depart from the spirit and scope of the present invention in its spirit and scope. Are intended to be covered by the scope of the present invention. Therefore, the protection scope of the present invention shall be subject to the protection scope of the claims.
Claims (8)
1. The signal learning and dynamic determination method based on the reference signal characteristics is characterized by comprising the following steps of:
s1, constructing a reference signal learning and tracking network;
s2, acquiring a reference signal in a rainless period;
s3, updating parameters of the reference signal learning and tracking network based on the reference signal;
s4, acquiring a reference signal of a rainfall period, and ending the updating;
s5, repeating the steps S1-S3 after rainfall is finished; and executing the S4 again when the rainfall begins, and finally finishing the signal learning and dynamic determination of the reference signal.
2. The method according to claim 1, wherein the reference signal learning and tracking network in S1 comprises a long-and-short neural network, a generalized regression network, and a kalman filter network.
3. The method of claim 1, wherein the reference signal during the no rain period in S2 comprises a microwave link and a satellite-to-ground link.
4. The method of claim 1, wherein the reference signal AC in the no rain period in S2 is used as the reference signal characteristic sun (t) is expressed as:
AC sun (t)＝A O (t)+A V (t)+A C (t)+A S (t)+A other +C
in the formula, A o Is the oxygen decay; a. The v Moisture attenuation; a. The c Is liquid water decay in the cloud; a. The s Is the scintillation decay; a. The other Attenuation due to other factors; c is a constant.
5. The method of claim 1, wherein the updating of the parameters in S3 comprises: and inputting the reference signal into the reference signal learning and tracking network, acquiring a variation characteristic by learning the reference signal, and updating the parameter of the reference signal learning and tracking network based on the variation characteristic.
6. The method of claim 5, wherein the reference signal comprises a signal-to-noise ratio and a level value.
7. The method of claim 5, wherein the variation characteristics comprise time-series variation characteristics of the reference signal during the rainless period and the rainfall period.
8. The method of claim 1, wherein the updating of the parameters in S3 is represented by:
in the formula (f) t Indicating a forgetting gate; g t Represents an update gate; i.e. i t Representing an input gate; c. C t Indicating the state of the cell at time t; c. C t-1 Represents the state of the cell at time t-1; h is t Represents the signal strength at time t; h is a total of t-1 Represents the signal strength at time t-1; o. o t An output gate is shown; w is a group of xf A weight representing a visible layer input in a forgotten door; w is a group of hf Representing weights of hidden layer outputs in a forgetting gate; w is a group of xc A weight representing the visible layer input in the update gate; w is a group of hc A weight representing the output of the hidden layer in the update gate; w xi Representing the weight of the visible layer input in the input gate; w is a group of hi Representing weights of hidden layer outputs in the input gate; b f A deviation indicating a forgotten door; b c A deviation indicative of a state of the cell; b i Indicating the deviation of the input gate; x is the number of t Representing the reference signal value AC sun (t), σ (·) denotes a sigmoid function.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202210944169.7A CN115308817B (en) | 2022-08-05 | 2022-08-05 | Signal learning and dynamic determination method based on reference signal characteristics |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202210944169.7A CN115308817B (en) | 2022-08-05 | 2022-08-05 | Signal learning and dynamic determination method based on reference signal characteristics |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN115308817A true CN115308817A (en) | 2022-11-08 |
| CN115308817B CN115308817B (en) | 2024-04-19 |
Family
ID=83860380
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202210944169.7A Active CN115308817B (en) | 2022-08-05 | 2022-08-05 | Signal learning and dynamic determination method based on reference signal characteristics |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN115308817B (en) |
Citations (7)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN111666656A (en) * | 2020-05-09 | 2020-09-15 | 江苏微之润智能技术有限公司 | Rainfall estimation method and rainfall monitoring system based on microwave rainfall attenuation |
| CN112651172A (en) * | 2020-12-17 | 2021-04-13 | 杭州鲁尔物联科技有限公司 | Rainfall peak type dividing method, device, equipment and storage medium |
| CN113095590A (en) * | 2021-04-29 | 2021-07-09 | 中国人民解放军国防科技大学 | High spatial-temporal resolution reconstruction analysis and short-term prediction method for microwave horizontal rainfall field |
| WO2022023887A1 (en) * | 2020-07-30 | 2022-02-03 | Università Di Pisa | Method for the estimating the presence of rain |
| CN114692692A (en) * | 2022-04-02 | 2022-07-01 | 河海大学 | Snowfall identification method based on microwave attenuation signal fusion kernel extreme learning machine |
| US20220244426A1 (en) * | 2019-05-15 | 2022-08-04 | Hd Rain | Precipitation measurement method and device |
| CN115097549A (en) * | 2022-06-20 | 2022-09-23 | 丝路梵天(甘肃)通信技术有限公司 | Radio link rainfall monitoring system and method based on deep learning |
-
2022
- 2022-08-05 CN CN202210944169.7A patent/CN115308817B/en active Active
Patent Citations (7)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20220244426A1 (en) * | 2019-05-15 | 2022-08-04 | Hd Rain | Precipitation measurement method and device |
| CN111666656A (en) * | 2020-05-09 | 2020-09-15 | 江苏微之润智能技术有限公司 | Rainfall estimation method and rainfall monitoring system based on microwave rainfall attenuation |
| WO2022023887A1 (en) * | 2020-07-30 | 2022-02-03 | Università Di Pisa | Method for the estimating the presence of rain |
| CN112651172A (en) * | 2020-12-17 | 2021-04-13 | 杭州鲁尔物联科技有限公司 | Rainfall peak type dividing method, device, equipment and storage medium |
| CN113095590A (en) * | 2021-04-29 | 2021-07-09 | 中国人民解放军国防科技大学 | High spatial-temporal resolution reconstruction analysis and short-term prediction method for microwave horizontal rainfall field |
| CN114692692A (en) * | 2022-04-02 | 2022-07-01 | 河海大学 | Snowfall identification method based on microwave attenuation signal fusion kernel extreme learning machine |
| CN115097549A (en) * | 2022-06-20 | 2022-09-23 | 丝路梵天(甘肃)通信技术有限公司 | Radio link rainfall monitoring system and method based on deep learning |
Non-Patent Citations (3)
| Title |
|---|
| MINGHAO XIAN ET AL.: "Rainfall Monitoring Based on Machine Learning by Earth-Space Link in the Ku Band", APPLIED EARTH OBSERVATIONS AND REMOTE SENSING, vol. 13, pages 3656, XP011797447, DOI: 10.1109/JSTARS.2020.3004375 * |
| 杨瑞科;李磊;仲普;赵振维;: "我国典型地区动态降雨衰减时间序列仿真研究", 电波科学学报, no. 05, pages 53 - 58 * |
| 林淑鲜;朱立东;: "Ka波段雨衰减时间序列的合成及性能仿真", 计算机仿真, no. 03, 15 March 2013 (2013-03-15) * |
Also Published As
| Publication number | Publication date |
|---|---|
| CN115308817B (en) | 2024-04-19 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN111310968B (en) | LSTM neural network circulating hydrologic forecasting method based on mutual information | |
| CN109711617B (en) | Medium-and-long-term runoff prediction method based on BLSTM deep learning | |
| CN103606006B (en) | Sludge volume index (SVI) soft measuring method based on self-organized T-S fuzzy nerve network | |
| CN111767517B (en) | BiGRU multi-step prediction method, system and storage medium applied to flood prediction | |
| CN113687433B (en) | Bi-LSTM-based magnetotelluric signal denoising method and system | |
| CN110401978B (en) | Indoor positioning method based on neural network and particle filter multi-source fusion | |
| CN107229823B (en) | Method for analyzing uncertainty of wind load effect extreme value | |
| CN113536665B (en) | Road surface temperature short-term prediction method and system based on characteristic engineering and LSTM | |
| CN112330065A (en) | Runoff forecasting method based on basic flow segmentation and artificial neural network model | |
| CN105678417A (en) | Prediction method and device for tunnel face water inflow of construction tunnel | |
| CN113361690A (en) | Water quality prediction model training method, water quality prediction device, water quality prediction equipment and medium | |
| CN112434890A (en) | Prediction method of tunnel settlement time sequence based on CEEMDAN-BilSTM | |
| CN111428419A (en) | Suspended sediment concentration prediction method and device, computer equipment and storage medium | |
| CN116562414A (en) | Land water reserve change prediction method based on multisource hydrographic meteorological data | |
| CN110197704B (en) | Self-organizing BP neural network water output total phosphorus prediction method based on limited Boltzmann machine | |
| CN116930609A (en) | Electric energy metering error analysis method based on ResNet-LSTM model | |
| CN111369078A (en) | Water supply quality prediction method based on long-term and short-term memory neural network | |
| CN114706144A (en) | Marine environment forecasting method, device, system and storage medium | |
| CN115308817A (en) | Signal learning and dynamic determination method based on reference signal characteristics | |
| CN111368461B (en) | Improved grey model tunnel settlement monitoring method based on SVD denoising processing | |
| CN114386672B (en) | Environment big data Internet of things intelligent detection system | |
| CN117332898A (en) | New energy small time scale power time sequence rolling prediction method based on machine learning | |
| CN112363140A (en) | Thermodynamic constraint extrapolation objective correction method based on cyclic neural network | |
| CN115840908A (en) | Method for constructing PM2.5 three-dimensional dynamic monitoring field by microwave link based on LSTM model | |
| CN113487069B (en) | Regional flood disaster risk assessment method based on GRACE daily degradation scale and novel DWSDI index |
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 |