CN115097549A - Radio link rainfall monitoring system and method based on deep learning - Google Patents
Radio link rainfall monitoring system and method based on deep learning Download PDFInfo
- Publication number
- CN115097549A CN115097549A CN202210698819.4A CN202210698819A CN115097549A CN 115097549 A CN115097549 A CN 115097549A CN 202210698819 A CN202210698819 A CN 202210698819A CN 115097549 A CN115097549 A CN 115097549A
- Authority
- CN
- China
- Prior art keywords
- rainfall
- radio
- deep learning
- monitoring
- attenuation
- 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
- 238000012544 monitoring process Methods 0.000 title claims abstract description 96
- 238000000034 method Methods 0.000 title claims abstract description 50
- 238000013135 deep learning Methods 0.000 title claims abstract description 44
- 238000004891 communication Methods 0.000 claims abstract description 29
- 230000005540 biological transmission Effects 0.000 claims abstract description 26
- 238000004422 calculation algorithm Methods 0.000 claims abstract description 10
- 238000005516 engineering process Methods 0.000 claims abstract description 10
- 238000013473 artificial intelligence Methods 0.000 claims abstract description 9
- 238000012549 training Methods 0.000 claims abstract description 8
- 238000013136 deep learning model Methods 0.000 claims description 10
- 238000007781 pre-processing Methods 0.000 claims description 10
- 238000001514 detection method Methods 0.000 claims description 9
- 230000008569 process Effects 0.000 claims description 9
- 238000013527 convolutional neural network Methods 0.000 claims description 8
- 238000003860 storage Methods 0.000 claims description 6
- 238000013528 artificial neural network Methods 0.000 claims description 5
- 238000012545 processing Methods 0.000 claims description 5
- 238000004590 computer program Methods 0.000 claims description 4
- 230000007613 environmental effect Effects 0.000 claims description 3
- 239000005433 ionosphere Substances 0.000 claims description 3
- 230000005855 radiation Effects 0.000 claims description 3
- 239000005436 troposphere Substances 0.000 claims description 3
- 238000013500 data storage Methods 0.000 claims description 2
- 230000009466 transformation Effects 0.000 claims description 2
- 230000001186 cumulative effect Effects 0.000 claims 1
- 230000004069 differentiation Effects 0.000 claims 1
- 238000004458 analytical method Methods 0.000 description 6
- 238000005259 measurement Methods 0.000 description 6
- 238000011160 research Methods 0.000 description 5
- 238000004364 calculation method Methods 0.000 description 4
- 238000012360 testing method Methods 0.000 description 4
- 238000012795 verification Methods 0.000 description 4
- 230000004913 activation Effects 0.000 description 3
- 238000010586 diagram Methods 0.000 description 3
- 230000000694 effects Effects 0.000 description 3
- 238000002474 experimental method Methods 0.000 description 3
- 230000006870 function Effects 0.000 description 3
- 239000011159 matrix material Substances 0.000 description 3
- 210000002569 neuron Anatomy 0.000 description 3
- 238000001228 spectrum Methods 0.000 description 3
- 238000003491 array Methods 0.000 description 2
- 230000001413 cellular effect Effects 0.000 description 2
- 230000007547 defect Effects 0.000 description 2
- 238000009826 distribution Methods 0.000 description 2
- 238000003973 irrigation Methods 0.000 description 2
- 230000002262 irrigation Effects 0.000 description 2
- 230000008092 positive effect Effects 0.000 description 2
- 239000000047 product Substances 0.000 description 2
- 238000012827 research and development Methods 0.000 description 2
- 238000010183 spectrum analysis Methods 0.000 description 2
- 239000013589 supplement Substances 0.000 description 2
- 230000009286 beneficial effect Effects 0.000 description 1
- 210000004027 cell Anatomy 0.000 description 1
- 230000008859 change Effects 0.000 description 1
- 238000010276 construction Methods 0.000 description 1
- 238000013480 data collection Methods 0.000 description 1
- 238000013461 design Methods 0.000 description 1
- 238000011161 development Methods 0.000 description 1
- 239000000428 dust Substances 0.000 description 1
- 239000012535 impurity Substances 0.000 description 1
- 238000009434 installation Methods 0.000 description 1
- 238000012423 maintenance Methods 0.000 description 1
- 238000004519 manufacturing process Methods 0.000 description 1
- 238000013507 mapping Methods 0.000 description 1
- 238000010295 mobile communication Methods 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 230000003287 optical effect Effects 0.000 description 1
- 230000010287 polarization Effects 0.000 description 1
- 238000002310 reflectometry Methods 0.000 description 1
- 239000004576 sand Substances 0.000 description 1
- 239000004065 semiconductor Substances 0.000 description 1
- 230000001502 supplementing effect Effects 0.000 description 1
- XLYOFNOQVPJJNP-UHFFFAOYSA-N water Substances O XLYOFNOQVPJJNP-UHFFFAOYSA-N 0.000 description 1
- 238000009736 wetting Methods 0.000 description 1
Images
Classifications
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01W—METEOROLOGY
- G01W1/00—Meteorology
- G01W1/14—Rainfall or precipitation gauges
-
- 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
-
- 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)
- Life Sciences & Earth Sciences (AREA)
- Physics & Mathematics (AREA)
- Theoretical Computer Science (AREA)
- Environmental & Geological Engineering (AREA)
- Biophysics (AREA)
- Data Mining & Analysis (AREA)
- Environmental Sciences (AREA)
- Biodiversity & Conservation Biology (AREA)
- Health & Medical Sciences (AREA)
- Artificial Intelligence (AREA)
- Biomedical Technology (AREA)
- Atmospheric Sciences (AREA)
- Computational Linguistics (AREA)
- Ecology (AREA)
- Evolutionary Computation (AREA)
- General Health & Medical Sciences (AREA)
- Molecular Biology (AREA)
- Computing Systems (AREA)
- General Engineering & Computer Science (AREA)
- General Physics & Mathematics (AREA)
- Mathematical Physics (AREA)
- Software Systems (AREA)
- Hydrology & Water Resources (AREA)
- Radio Relay Systems (AREA)
Abstract
The invention belongs to the technical field of rainfall monitoring, and discloses a radio link rainfall monitoring system and method based on deep learning, wherein a ground communication link and a satellite signal are used as signal sources, a radio monitoring receiver is used for receiving radio signals passing through different rainfall environments, and rainfall during collection is recorded; constructing a deep learning algorithm model suitable for rainfall monitoring of a radio transmission link, and training a rainfall attenuation model suitable for distinguishing radio transmission by using the inquired rainfall as a signal data label; and analyzing rainfall attenuation caused by rainfall in different rainfall environments by using artificial intelligence and a radio technology, and exporting the trained model for radio rainfall monitoring. The invention provides a method for detecting rainfall of a radio transmission link based on deep learning by combining with the means of deep learning, so that the coverage of rainfall monitoring is wider, and the rainfall forecast is timely and accurately carried out.
Description
Technical Field
The invention belongs to the technical field of rainfall monitoring, and particularly relates to a radio link rainfall monitoring system and method based on deep learning.
Background
At present, a lot of researches are made by many scholars at home and abroad for accurately detecting rainfall, and the existing means for measuring rainfall mainly comprise a rain gauge, a weather radar, a weather satellite, a raindrop spectrometer and the like, but the existing means for measuring rainfall respectively have different defects:
(1) the rain gauge is accurate in rainfall monitoring, but due to insufficient spatial distribution precision and monitoring range, the obtained meteorological information is only rainfall data of a small local area near a monitoring station, and cannot represent rainfall of the whole area;
(2) the meteorological radar cannot monitor rainfall in a large area due to high manufacturing cost and limited transverse scanning radius; the accuracy of radar inversion for the rainfall intensity is often influenced by the atmospheric pressure, the temperature and the humidity and the rainfall phase state, so that a large error is generated when the rainfall is calculated by using the radar reflectivity, and the radar is easily influenced by the terrain and ground clutter when measuring the rainfall on the near ground, so that the error of a measurement result is large;
(3) the weather satellite is through the trend of analysis cloud and fog, use visible light and infrared wave to survey cloud top temperature and the water content etc. of cloud and fog to obtain the rainfall probability in current region, nevertheless because the change of relief and shape potential field can lead to the testing result inaccurate, simultaneously, the penetrability that is used for infrared wave and the visible light wave band of satellite remote sensing meteorological monitoring is general, can't obtain the rainfall information of inside and the nearly ground of cloud and fog, consequently its measurement accuracy is not good enough.
(4) The raindrop spectrum instrument can not ensure the accuracy of rainfall monitoring because the raindrop spectrum changes greatly due to the variability of the type and rainfall intensity of rainfall and the difference of the altitude, atmospheric pressure and the like of the area.
In the course of researching rainfall, many scholars put forward the idea of measuring rainfall by using attenuation characteristics of radio electromagnetic waves, and in the last eighties, the relation between rainfall attenuation and rainfall rate, antenna parameter frequency and the like, namely a DAH model, is put forward for the first time:
the rainfall attenuation can be calculated through the DAH model, but in actual situations, the antenna can be erected outdoors, the antenna can be wetted when rainfall occurs, meanwhile, the free space attenuation can occur in the propagation process of the radio, attenuation caused by the fact that the antenna is wetted by rainwater and other attenuation factors in the rainfall process are not considered in the rain attenuation calculating mode, the attenuation and the atmospheric gas attenuation are mainly included, multipath propagation attenuation, clutter attenuation and the like are mainly included, and the calculation of the rain attenuation is deviated. In the prior art, the attenuation on a link is further decomposed into rainfall attenuation, free space attenuation and wet antenna attenuation by using an artificial rainfall method. Namely:
A L ＝A rain +A free +A wa
to address this problem, the International Telecommunications Union (ITU), proposed a specific model of rainfall attenuation, the ITU-R p.838-3 rain attenuation recommendation. The recommendation gives the relationship between the rainfall attenuation and the used frequency, the polarization mode of the antenna and the rainfall in detail, and the specific attenuation can be calculated from the power relation of the rainfall intensity.
In 2006, some scholars proposed the idea of using commercial cellular links for detection of the environment. The method is characterized in that the existing wireless communication network is proposed to be used for environment monitoring for the first time in the world so as to achieve the purpose of supplementing the existing meteorological monitoring means such as a rain gauge, a meteorological radar, a meteorological satellite and the like, and the rainfall estimation can be possible in local areas by using the wireless communication network.
In 2007, a scholars estimates the rainfall intensity by using the microwave attenuation of 13.9GHz-24.1GHz and distinguishes the start-stop time of rainfall, and shows that the detection of urban rainfall by using double-frequency electromagnetic waves has better precision than that by using single-frequency electromagnetic waves.
Since 2009, some scholars have performed a test of inverting the rainfall field in urban areas by using a commercial microwave network, and a test of monitoring rainfall in real time by using a cellular wireless network with 2400 links has reached the full environment of the netherlands.
In 2020, a scholars uses a convolutional neural network to measure rainfall attenuation data in a microwave link, and then rainfall is calculated. They used 10-40GHz signals as signal sources to calculate, and successfully distinguished the rainy period from the arid period.
The research of using a microwave link to calculate rainfall attenuation in China is in the starting stage, and the microwave transmission characteristics are researched in 2010 on the factors of raindrop spectrum distribution, rainfall intensity, frequency, temperature and the like, so that the influences of the rainfall intensity on the microwave transmission characteristics are the largest, and the influences of the rainfall intensity on the frequency are the second and the temperature are the smallest.
A method for inverting the rain intensity by using a microwave link rain attenuation measurement value is proposed in 2015. The signal of the frequency band of 15 GHZ-20 GHz is used, an OTT raindrop spectrometer is arranged at a transmitting end to measure rainfall attenuation, the signal indicates that the actual rainfall attenuation is the most value, the maximum value of the rainfall attenuation is 35dB by using the frequency band signal of 15 GHz-20 GHz, and the rainfall attenuation is closely related to data such as link length, rain cells, transmitting frequency and the like. In 2020, the rainfall is calculated by using data provided by Hainan Limited company of China Mobile communication group and a calculation model provided by International Telecommunication Union (ITU), and the result shows that the rainfall measured by a communication base station signal attenuation method is wholly slightly larger than the rainfall station measurement result, the deviation of the rainfall station and the rainfall station can be controlled within 2mm/h, which shows that the rainfall station and the communication base station signal attenuation method have better consistency, and shows that the method for calculating rainfall by using the attenuation of the base station signal can be used as a supplement of the existing rainfall monitoring method.
Through the above analysis, the problems and defects of the prior art are as follows:
(1) the experimental link is single in design: in the current stage of research on rainfall attenuation and rainfall, the used link is mostly high-frequency signals of 18-24GHz, and the experimental place is also fixed, so that the generalization capability of the model is weak, and the model cannot be used in multiple scenes.
(2) Time of rainfall and its discontinuity: since rainfall requires a plurality of meteorological factors to act together, the time of rainfall is not fixed and is random. And the rainfall is interrupted and does not last for a period of time because the rainfall is disturbed by temperature, wind field and the like.
(3) There is no suitable sample database: at present, most experiments related to rainfall attenuation calculation for rainfall capacity are experiments related to rainfall attenuation calculation, wherein the rainfall attenuation is calculated by using level changes of a transmitting end and a receiving end, signal intensity changes on a communication link are not used, and a rainfall detection related experiment is performed by combining a deep learning means.
(4) The algorithmic model is not a deep learning model.
Disclosure of Invention
Aiming at the problems in the prior art, the invention provides a radio link rainfall monitoring system and method based on deep learning.
The invention is realized in such a way that a radio link rainfall monitoring method based on deep learning comprises the following steps:
taking a ground communication link and a satellite signal as a signal source, receiving radio signals passing through different rainfall environments by using a radio monitoring receiver, and recording rainfall during acquisition;
step two, constructing a deep learning algorithm model suitable for rainfall monitoring of a radio propagation link, and using a method of an artificial neural network and a convolutional neural network as the deep learning algorithm model provided by the invention, wherein the specific model is a deep learning module in fig. 4, and the following relation exists between the output and the input of a full connection layer in the known artificial neural network:
where ω represents the weight, x represents the input, y represents the output, f is the activation function, and n is the number of neurons.
The deep convolutional neural network algorithm is provided by AlexNet, and becomes the mainstream research direction. One neuron of a convolutional neural network can be described as:
y (i) ＝f|Cov (i) (W,x)|
in the formula: i represents different neurons; x represents an input matrix; y represents a feature image; w represents a convolution kernel; b is an offset. The convolution operation formula can be described as:
in the formula: r is the number of rows of the matrix x; c is the number of columns of matrix x. In the convolutional neural network, each filter of the convolutional layer sequentially convolves the input data in the whole receptive field, and the obtained result constitutes a feature map of the input data, so that the local features of the image are extracted. After the convolution operation, the mapping is completed by using the bias and the activation function on the output result:
in the above formula: l represents the number of layers representing the convolutional layer; k represents a convolution kernel; b represents a bias term; m j An input feature map representing a previous layer of convolution; f represents an activation function.
After the model is determined, the invention firstly takes the accurate rainfall information obtained by query as a signal data label to train a rainfall attenuation model suitable for distinguishing the radio during transmission.
And step three, analyzing rainfall attenuation caused by rainfall in different rainfall environments by using artificial intelligence and a radio technology, and exporting the trained model for radio rainfall monitoring.
Further, the specific process of the first step is as follows:
and selecting a proper radio communication link frequency band, a satellite downlink frequency band and a receiving antenna to receive equipment data, and selecting a monitoring receiver with consistent frequency band to receive radio signal data.
The device data of the invention comprises: receiver average signal received power, antenna gain, feeder loss, device self loss, environmental electromagnetic interference parameters, link signal gain, selected equipment antenna gain and radiation pattern, connected device loss, free space attenuation loss, wetted antenna attenuation loss, ground vegetation attenuation loss, ionosphere, troposphere, atmospheric attenuation loss present in satellite communications when receiving satellite signals, and connected device loss.
Further, in the rainfall information when the recording signal data is acquired in the first step, the recording is performed by accumulating rainfall per minute.
Furthermore, the second step is suitable for distinguishing the rainfall attenuation model of the radio during transmission, and the rainfall attenuation is distinguished by using a method of combining a convolutional neural network and an artificial neural network, so that the purpose of monitoring the rainfall is achieved.
Furthermore, the trained deep learning model is used for learning rainfall attenuation characteristics to distinguish different rainfall capacities, so that the purpose of monitoring the rainfall capacity is achieved.
Another object of the present invention is to provide a deep learning-based radio link rainfall monitoring system, which includes:
the signal source transmitting module is used for transmitting a signal source by using a signal station, the signal source comprises a satellite signal and a ground link signal, and the signal passes through a rainfall environment in the transmission process;
the signal receiving module receives signals by using a special frequency band antenna and a radio detection receiver;
the data storage module is used for storing the acquired original signals under different rainfall environments into the storage device through the server;
and the data preprocessing module is used for preprocessing the data, selecting a proper deep learning model and obtaining different rainfall data.
Further, the method for preprocessing the data by the data preprocessing module comprises the following steps: the collected original time domain signal data in different rainfall periods are converted into frequency domain signals through Fourier transformation, a proper deep learning model is built to learn rainfall attenuation characteristics caused by rainfall, and then rainfall information in different collection periods is distinguished.
In combination with the technical solutions and the technical problems to be solved, please analyze the advantages and positive effects of the technical solutions to be protected in the present invention from the following aspects:
first, aiming at the technical problems existing in the prior art and the difficulty in solving the problems, the technical problems to be solved by the technical scheme of the present invention are closely combined with the technical scheme to be protected and the results and data in the research and development process, and some creative technical effects brought after the problems are solved are analyzed in detail and deeply. The specific description is as follows:
the invention relates to a method for identifying and analyzing the signal difference of communication link signals collected at different time under the rainfall environment, thereby obtaining the rainfall at the corresponding time. Different rainfall attenuation characteristics caused by rainfall are learned by collecting communication link signals in different rainfall periods and using a deep learning means, and different rainfall is obtained by analyzing the rainfall attenuation characteristics in different periods, so that rainfall detection is realized.
The present invention uses spectral analysis as the primary means of analysis. The spectrum analysis is a technology for decomposing a complex signal into simpler signals, can well reflect the characteristics of the signals which cannot be reflected in a time domain, and is beneficial to the accuracy of the signal analysis.
The invention provides a method for detecting rainfall of a radio transmission link based on deep learning by combining with the means of deep learning, so that the coverage of rainfall monitoring can be wider, and the rainfall monitoring can be more timely. Thereby ensuring that the rainfall forecast can be carried out timely and accurately.
The rainfall monitoring method is different from the existing rainfall monitoring method by using a rain gauge, and the rainfall is monitored by using a radio and artificial intelligence method, so that the coverage area of rainfall monitoring is wider and more accurate:
1) the measurement range is wide:
the rainfall monitoring method at the present stage is mostly a rain gauge built by a weather monitoring station, rainfall data obtained by the rain gauge only represents rainfall in local areas, and cannot represent rainfall in the whole area.
2) The measurement precision is accurate:
when the rain gauge is used for monitoring rainfall, the rain gauge needs to be placed outdoors, and due to outdoor strong wind, sand storm and the like, other impurities such as dust and the like can be mixed in the rain gauge, so that the rainfall monitoring is inaccurate.
Secondly, considering the technical scheme as a whole or from the perspective of products, the technical effect and advantages of the technical scheme to be protected by the invention are specifically described as follows:
the invention can be applied to multiple purposes and can be widely applied to the fields of agriculture, transportation, industry and the like. In the aspect of agriculture, the invention can monitor the irrigation condition of crops by radio; in the aspect of transportation, the invention can monitor the areas such as highways, airports and the like, is used for accurately acquiring rainfall conditions in real time and ensures the traffic safety; in the industrial field, rainfall monitoring can avoid some economic losses due to rain.
Third, as inventive supplementary proof of the claims of the present invention, there are several important aspects as follows:
(1) the technical scheme of the invention fills the technical blank in the industry at home and abroad:
at present, scientific researchers at home and abroad carry out a series of researches on calculating rainfall by using radio rainfall attenuation, but no people use commercial communication links and satellite signals as signal sources and then carry out analysis by combining a deep learning means. Therefore, the invention has certain innovativeness.
(2) The technical scheme of the invention solves the technical problem that people are eager to solve but can not succeed all the time:
at present, a method related to rainfall monitoring in China is a tipping bucket type rain gauge, rainfall data acquired by the tipping bucket type rain gauge when rainfall is monitored is data of a small local area (0-100 m), the data cannot represent rainfall in a large range (1-5 km), if rainfall in a large range is acquired, more tipping bucket type rain gauges need to be installed, the installation cost and the later maintenance cost of the tipping bucket type rain gauge are too high, the rainfall monitoring is carried out on the basis of signals of commercial radio links, the maximum monitoring range can reach 5km, and the method is superior to the tipping bucket type rain gauge. The rainfall detection method is novel and has wider coverage area. (4) The technical scheme of the invention overcomes the technical prejudice whether:
drawings
Fig. 1 is a flowchart of a deep learning-based radio link rainfall monitoring method according to an embodiment of the present invention.
Fig. 2 is a schematic diagram of a deep learning-based radio link rainfall monitoring system according to an embodiment of the present invention.
Fig. 3 is a data collection flow chart of a deep learning-based radio link rainfall monitoring method according to an embodiment of the present invention.
Fig. 4 is a diagram of processing and analyzing radio signal data for different rainfall according to an embodiment of the present invention.
Detailed Description
In order to make the objects, technical solutions and advantages of the present invention more apparent, the present invention is further described in detail with reference to the following embodiments. It should be understood that the specific embodiments described herein are merely illustrative of the invention and are not intended to limit the invention.
First, an embodiment is explained. This section is an explanatory embodiment expanding on the claims so as to fully understand how the present invention is embodied by those skilled in the art.
With the development of communication business in China, communication stations in China are more and more densely built, compared with the traditional weather monitoring station, a communication link has a wider coverage area, the coverage area is generally 1-3km, and the rainfall near a base station is calculated by utilizing the attenuation of signals of the communication link, so that the existing rainfall monitoring means such as a rain gauge and a weather radar can be supplemented. In some remote areas or areas where no weather monitoring station is built, the attenuation of the communication link can be used to try to calculate the rainfall, making it a new rainfall monitoring method. And in some places with weather monitoring stations, rainfall monitoring means such as rainfall supplement rainfall meters can be calculated by utilizing the attenuation of communication link signals.
Meanwhile, in the transmission process, satellite signals can directly penetrate through a rainfall environment and reach a ground monitoring receiver, so that the satellite signals can be used as signal sources for analysis during rainfall monitoring.
As shown in fig. 1, a deep learning-based radio link rainfall monitoring method provided in an embodiment of the present invention includes:
s101, taking a ground communication link and a satellite signal as a signal source, receiving radio signals passing through different rainfall environments by using a radio monitoring receiver, and recording rainfall during collection;
s102, constructing a deep learning algorithm model suitable for rainfall monitoring of a radio transmission link, and training a rainfall attenuation model suitable for distinguishing radio transmission by using the inquired rainfall as a signal data label;
and S103, analyzing rainfall attenuation caused by rainfall in different rainfall environments by using artificial intelligence and a radio technology, and exporting the trained model for radio rainfall monitoring.
The specific process of step S101 in the embodiment of the present invention is:
and selecting a proper radio communication link frequency band, a satellite downlink frequency band and a receiving antenna to receive equipment data, and selecting a monitoring receiver with consistent frequency band to receive radio signal data. The device data includes: receiver average signal received power, antenna gain, feeder loss, device self loss, environmental electromagnetic interference parameters, link signal gain, selected equipment antenna gain and radiation pattern, connected device loss, free space attenuation loss, wetted antenna attenuation loss, ground vegetation attenuation loss, ionosphere, troposphere, atmospheric attenuation loss in satellite communications, and connected device loss.
In the step S101 in the embodiment of the present invention, in recording the rainfall information during collection, the rainfall is accumulated every minute.
In step S102 in the embodiment of the present invention, the rainfall attenuation model for radio transmission is adapted to differentiate rainfall attenuation based on the convolutional neural network and the artificial neural network, so as to achieve the purpose of monitoring rainfall. The trained deep learning model is used for learning rainfall attenuation characteristics to distinguish different rainfall capacities, and the purpose of rainfall capacity monitoring is achieved.
As shown in fig. 2, a deep learning-based radio link rainfall detection system according to an embodiment of the present invention includes:
the signal source comprises a satellite signal and a ground link signal; the rainfall environment through which the signal passes during transmission; a special band antenna for receiving signals, a radio detection receiver, and a terminal memory device.
As shown in fig. 3, the present invention provides a detailed flow chart for signal acquisition, which mainly includes three modules, which respectively introduce acquisition of signal data, storage of data, and preprocessing of data, and after acquiring original signals in different rainfall environments, the original signals are stored in a storage device through a server, and then a suitable deep learning model is selected through preprocessing of data, so as to obtain different rainfall data.
As shown in fig. 4, the present invention provides a detailed flow chart of data processing, which includes that acquired original time domain signal data in different rainfall periods are converted into frequency domain signals through fourier transform, and a suitable deep learning model is built to learn rainfall attenuation characteristics caused by rainfall, so as to distinguish rainfall information in different rainfall periods.
And II, application embodiment. In order to prove the creativity and the technical value of the technical scheme of the invention, the part is an application example of the technical scheme of the claims to a specific product or related technology.
The radio link rainfall monitoring method based on deep learning provided by the embodiment of the invention can be widely applied to the fields of agriculture, transportation, industry and the like. In the aspect of agriculture, the invention can monitor the irrigation condition of crops by radio; in the aspect of transportation, the invention can monitor the areas such as highways, airports and the like, is used for accurately acquiring rainfall conditions in real time and ensures the traffic safety; in the industrial field, rainfall monitoring can avoid some economic losses caused by rainwater.
A computer device, characterized in that the computer device comprises a memory and a processor, the memory storing a computer program which, when executed by the processor, causes the processor to carry out the steps of:
taking a ground communication link and a satellite signal as a signal source, receiving radio signals passing through different rainfall environments by using a radio monitoring receiver, and recording rainfall during collection;
constructing a deep learning algorithm model suitable for rainfall monitoring of a radio transmission link, and training a rainfall attenuation model suitable for distinguishing radio transmission by using the inquired rainfall as a signal data label;
and step three, analyzing rainfall attenuation caused by rainfall in different rainfall environments by using artificial intelligence and a radio technology, and exporting the trained model for radio rainfall monitoring.
A computer-readable storage medium storing a computer program which, when executed by a processor, causes the processor to perform the steps of:
taking a ground communication link and a satellite signal as a signal source, receiving radio signals passing through different rainfall environments by using a radio monitoring receiver, and recording rainfall during collection;
constructing a deep learning algorithm model suitable for rainfall monitoring of a radio transmission link, and training a rainfall attenuation model suitable for distinguishing radio transmission by using the inquired rainfall as a signal data label;
and step three, analyzing rainfall attenuation caused by rainfall in different rainfall environments by using artificial intelligence and a radio technology, and exporting the trained model for radio rainfall monitoring.
The information data processing terminal is characterized by being used for realizing the deep learning-based radio link rainfall monitoring method.
And thirdly, evidence of relevant effects of the embodiment. The embodiment of the invention achieves some positive effects in the process of research and development or use, and has great advantages compared with the prior art, and the following contents are described by combining data, diagrams and the like in the test process.
The signal frequency band for rainfall monitoring is determined firstly and is a commercial communication link and a Beidou Feng cloud 4A satellite downlink frequency band.
Selecting proper radio monitoring equipment, a receiver antenna and a satellite antenna for different rainy periods; and a signal data acquisition plan is made by combining a rainfall period, and the construction of an acquisition platform is completed.
When the acquisition platform is set up, the antenna is erected outdoors in open space, so that the radio signal can be ensured to pass through a rainfall environment. Ensuring that the radio signals can carry rainfall attenuation characteristics caused by rainfall. And the influence of clutter in the building can be reduced.
Meanwhile, when the antenna is collected in a rainy period, the antenna is wrapped, so that wet antenna attenuation caused by antenna wetting is removed, radio signals are guaranteed to only carry rainfall attenuation characteristics, other interference factors are avoided, and the rainfall is calculated more accurately.
The invention collects three types of commercial radio link signals in different rainfall periods, finds that rainfall attenuation characteristics of the signals can be learned by deep learning, firstly uses known rainfall signal data to train a model, and then carries out verification, finds that the verification accuracy rate can reach 94 percent, and the training and verification processes are shown in a table 1.
TABLE 1
The trained model is exported, signal data acquired in the rest rainfall periods are imported to serve as a verification set, rainfall information given by a weather monitoring station serves as a data tag, the generalization capability of the model is verified, the highest accuracy can reach 93.3%, and as shown in fig. 2, the model is proved to have better generalization capability. Therefore, the invention builds a radio-deep learning monitoring platform for monitoring rainfall in real time, so as to monitor rainfall information in real time.
TABLE 2
It should be noted that the embodiments of the present invention can be realized by hardware, software, or a combination of software and hardware. The hardware portion may be implemented using dedicated logic; the software portions may be stored in a memory and executed by a suitable instruction execution system, such as a microprocessor or specially designed hardware. It will be appreciated by those skilled in the art that the apparatus and methods described above may be implemented using computer executable instructions and/or embodied in processor control code, for example such code provided on a carrier medium such as a diskette, CD-or DVD-ROM, a programmable memory such as read-only memory (firmware) or a data carrier such as an optical or electronic signal carrier. The apparatus and its modules of the present invention may be implemented by hardware circuits such as very large scale integrated circuits or gate arrays, semiconductors such as logic chips, transistors, or programmable hardware devices such as field programmable gate arrays, programmable logic devices, etc., or by software executed by various types of processors, or by a combination of hardware circuits and software, e.g., firmware.
The above description is only for the purpose of illustrating the present invention and the appended claims are not to be construed as limiting the scope of the invention, which is intended to cover all modifications, equivalents and improvements that are within the spirit and scope of the invention as defined by the appended claims.
Claims (10)
1. A deep learning-based radio link rainfall monitoring method is characterized in that the deep learning-based radio link rainfall monitoring system and method comprises the following steps:
taking a ground communication link and a satellite signal as a signal source, receiving radio signals passing through different rainfall environments by using a radio monitoring receiver, and recording rainfall during acquisition;
constructing a deep learning algorithm model suitable for rainfall monitoring of a radio transmission link, and training a rainfall attenuation model suitable for distinguishing radio transmission by using the inquired rainfall as a signal data label;
and step three, analyzing rainfall attenuation caused by rainfall in different rainfall environments by using artificial intelligence and a radio technology, and exporting the trained model for radio rainfall monitoring.
2. The deep learning-based radio link rainfall monitoring method according to claim 1, wherein the specific process of the first step is as follows:
and selecting a proper radio communication link frequency band, a proper satellite downlink frequency band and a proper receiving antenna to receive equipment data, and selecting a monitoring receiver with consistent frequency band to receive radio signal data.
3. The deep learning-based radio link rainfall monitoring method of claim 2, wherein the device data comprises: receiver average signal received power, antenna gain, feeder loss, device self loss, environmental electromagnetic interference parameters, link signal gain, selected equipment antenna gain and radiation pattern, connected device loss, free space attenuation loss, wetted antenna attenuation loss, ground vegetation attenuation loss, ionosphere, troposphere, atmospheric attenuation loss in satellite communications, and connected device loss.
4. The deep learning-based radio link rainfall monitoring method of claim 1 wherein the first step records cumulative rainfall per minute of rainfall in the rainfall information as collected.
5. The radio link rainfall monitoring method based on deep learning of claim 1, wherein the second step is adapted to distinguish the rainfall attenuation model of the radio during transmission based on the convolutional neural network and the artificial neural network for rainfall attenuation differentiation, so as to achieve the purpose of rainfall monitoring;
the trained deep learning model is used for learning rainfall attenuation characteristics to distinguish different rainfall capacities, and the purpose of rainfall capacity monitoring is achieved.
6. A deep learning based radio link rainfall monitoring system for implementing the deep learning based radio link rainfall monitoring method of any one of claims 1 to 5, the deep learning based radio link rainfall monitoring system comprising:
the signal source transmitting module is used for transmitting a signal source by using a signal station, the signal source comprises a satellite signal and a ground link signal, and the signal passes through a rainfall environment in the transmission process;
the signal receiving module receives signals by using a special frequency band antenna and a radio detection receiver;
the data storage module is used for storing the acquired original signals under different rainfall environments into the storage device through the server;
and the data preprocessing module is used for preprocessing the data, selecting a proper deep learning model and obtaining different rainfall data.
7. The deep learning-based radio link rainfall monitoring system of claim 6, wherein the method of pre-processing data by the data pre-processing module comprises: the collected original time domain signal data in different rainfall periods are converted into frequency domain signals through Fourier transformation, a proper deep learning model is built to learn rainfall attenuation characteristics caused by rainfall, and then rainfall information in different collection periods is distinguished.
8. A computer device, characterized in that the computer device comprises a memory and a processor, the memory storing a computer program which, when executed by the processor, causes the processor to carry out the steps of:
taking a ground communication link and a satellite signal as a signal source, receiving radio signals passing through different rainfall environments by using a radio monitoring receiver, and recording rainfall during acquisition;
constructing a deep learning algorithm model suitable for rainfall monitoring of a radio transmission link, and training a rainfall attenuation model suitable for distinguishing radio transmission by using the inquired rainfall as a signal data label;
and step three, analyzing rainfall attenuation caused by rainfall in different rainfall environments by using artificial intelligence and a radio technology, and exporting the trained model for radio rainfall monitoring.
9. A computer-readable storage medium storing a computer program which, when executed by a processor, causes the processor to perform the steps of:
taking a ground communication link and a satellite signal as a signal source, receiving radio signals passing through different rainfall environments by using a radio monitoring receiver, and recording rainfall during collection;
constructing a deep learning algorithm model suitable for rainfall monitoring of a radio transmission link, and training a rainfall attenuation model suitable for distinguishing radio transmission by using the inquired rainfall as a signal data label;
and step three, analyzing rainfall attenuation caused by rainfall in different rainfall environments by using artificial intelligence and a radio technology, and exporting the trained model for radio rainfall monitoring.
10. An information data processing terminal, characterized in that the information data processing terminal is used for implementing the deep learning-based radio link rainfall monitoring method according to any one of claims 1 to 5.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202210698819.4A CN115097549B (en) | 2022-06-20 | 2022-06-20 | Radio link rainfall monitoring system and method based on deep learning |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202210698819.4A CN115097549B (en) | 2022-06-20 | 2022-06-20 | Radio link rainfall monitoring system and method based on deep learning |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN115097549A true CN115097549A (en) | 2022-09-23 |
| CN115097549B CN115097549B (en) | 2024-08-20 |
Family
ID=83290619
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202210698819.4A Active CN115097549B (en) | 2022-06-20 | 2022-06-20 | Radio link rainfall monitoring system and method based on deep learning |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN115097549B (en) |
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN115308817A (en) * | 2022-08-05 | 2022-11-08 | 中国人民解放军国防科技大学 | Signal learning and dynamic determination method based on reference signal characteristics |
Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN110031916A (en) * | 2019-03-07 | 2019-07-19 | 中国人民解放军国防科技大学 | Rainfall intensity measurement method based on satellite-ground link attenuation effect |
| CN110826810A (en) * | 2019-11-13 | 2020-02-21 | 吉林农业大学 | Regional rainfall prediction method combining spatial reasoning and machine learning |
| CN111582546A (en) * | 2020-04-09 | 2020-08-25 | 中国科学院国家空间科学中心 | Method and system for acquiring global rainfall information at different places by utilizing rainfall data set |
| US20210056461A1 (en) * | 2019-08-22 | 2021-02-25 | International Business Machines Corporation | Precipitation detection using mobile devices |
| US20220244426A1 (en) * | 2019-05-15 | 2022-08-04 | Hd Rain | Precipitation measurement method and device |
-
2022
- 2022-06-20 CN CN202210698819.4A patent/CN115097549B/en active Active
Patent Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN110031916A (en) * | 2019-03-07 | 2019-07-19 | 中国人民解放军国防科技大学 | Rainfall intensity measurement method based on satellite-ground link attenuation effect |
| US20220244426A1 (en) * | 2019-05-15 | 2022-08-04 | Hd Rain | Precipitation measurement method and device |
| US20210056461A1 (en) * | 2019-08-22 | 2021-02-25 | International Business Machines Corporation | Precipitation detection using mobile devices |
| CN110826810A (en) * | 2019-11-13 | 2020-02-21 | 吉林农业大学 | Regional rainfall prediction method combining spatial reasoning and machine learning |
| CN111582546A (en) * | 2020-04-09 | 2020-08-25 | 中国科学院国家空间科学中心 | Method and system for acquiring global rainfall information at different places by utilizing rainfall data set |
Non-Patent Citations (1)
| Title |
|---|
| 孙一丹: "GNSS 信号估算大气可降雨系统原理及应用进展", <中国农业资源与区划>, vol. 43, no. 9, 30 September 2022 (2022-09-30), pages 53 - 59 * |
Cited By (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN115308817A (en) * | 2022-08-05 | 2022-11-08 | 中国人民解放军国防科技大学 | Signal learning and dynamic determination method based on reference signal characteristics |
| CN115308817B (en) * | 2022-08-05 | 2024-04-19 | 中国人民解放军国防科技大学 | Signal learning and dynamic determination method based on reference signal characteristics |
Also Published As
| Publication number | Publication date |
|---|---|
| CN115097549B (en) | 2024-08-20 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN109697323B (en) | Rainfall observation method integrating satellite remote sensing and mobile communication base station signals | |
| CN111666656A (en) | Rainfall estimation method and rainfall monitoring system based on microwave rainfall attenuation | |
| CN109709577A (en) | A kind of Three-dimensional Variational Data Assimilation method of the aerosol LIDAR inverting PM2.5 based on WRF-Chem mode | |
| CN109581546A (en) | Rainfall type identification method based on microwave link attenuation and polarization information | |
| KR102432397B1 (en) | A system for providing rainfall probability information using meteorological image information provided by the Meteorological Agency based on machine learning and a method using the same | |
| CN111079835B (en) | Himapari-8 atmospheric aerosol inversion method based on deep full-connection network | |
| Bianchi et al. | Quality control of rain gauge measurements using telecommunication microwave links | |
| CN116363601B (en) | Data acquisition and analysis method and system for pollution monitoring equipment | |
| Gharanjik et al. | Centralized rainfall estimation using carrier to noise of satellite communication links | |
| CN110930048A (en) | Crop drought risk assessment system and method based on disaster mechanism process | |
| CN111337549A (en) | GPS multi-satellite fusion soil humidity monitoring method based on fuzzy entropy | |
| CN115097549B (en) | Radio link rainfall monitoring system and method based on deep learning | |
| US20230161071A1 (en) | Method for distinguishing sunny-rainy weather based on time division long-term evolution network | |
| CN112329265A (en) | Satellite remote sensing rainfall refinement space estimation method and system | |
| CN112051576A (en) | Intelligent multi-frequency microwave rainfall monitoring method | |
| Beritelli et al. | A neural network pattern recognition approach to automatic rainfall classification by using signal strength in LTE/4G networks | |
| CN115639625A (en) | High-spatial-resolution automatic refined atmospheric horizontal visibility forecasting system | |
| Djibo et al. | High-resolution rainfall maps from commercial microwave links for a data-scarce region in West Africa | |
| CN118153802A (en) | Remote sensing and multi-environment factor coupled wheat key waiting period prediction method and device | |
| CN110441830B (en) | Small-area weather prediction method based on automatic weather station and storage medium | |
| KR20220111578A (en) | A system for providing rainfall probability information using meteorological image information provided by the Meteorological Agency based on machine learning and a method using the same | |
| CN114828077B (en) | Regional rainfall field reconstruction method based on cellular mobile communication network | |
| CN111597692A (en) | Surface net radiation estimation method, system, electronic equipment and storage medium | |
| CN115421220A (en) | Multi-factor local precipitation indication method and system based on deep learning | |
| CN114414446A (en) | Near-ground particulate matter concentration estimation method, device and equipment based on satellite remote sensing |
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 |