CN111666656A - Rainfall estimation method and rainfall monitoring system based on microwave rainfall attenuation - Google Patents
Rainfall estimation method and rainfall monitoring system based on microwave rainfall attenuation Download PDFInfo
- Publication number
- CN111666656A CN111666656A CN202010386967.3A CN202010386967A CN111666656A CN 111666656 A CN111666656 A CN 111666656A CN 202010386967 A CN202010386967 A CN 202010386967A CN 111666656 A CN111666656 A CN 111666656A
- Authority
- CN
- China
- Prior art keywords
- rainfall
- estimation
- microwave
- data
- 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.)
- Pending
Links
Images
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F30/00—Computer-aided design [CAD]
- G06F30/20—Design optimisation, verification or simulation
-
- 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
- G01S13/00—Systems using the reflection or reradiation of radio waves, e.g. radar systems; Analogous systems using reflection or reradiation of waves whose nature or wavelength is irrelevant or unspecified
- G01S13/88—Radar or analogous systems specially adapted for specific applications
- G01S13/95—Radar or analogous systems specially adapted for specific applications for meteorological use
- G01S13/958—Theoretical aspects
-
- 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
- G06N3/04—Architecture, e.g. interconnection topology
- G06N3/045—Combinations of networks
-
- Y—GENERAL TAGGING OF NEW TECHNOLOGICAL DEVELOPMENTS; GENERAL TAGGING OF CROSS-SECTIONAL TECHNOLOGIES SPANNING OVER SEVERAL SECTIONS OF THE IPC; TECHNICAL SUBJECTS COVERED BY FORMER USPC CROSS-REFERENCE ART COLLECTIONS [XRACs] AND DIGESTS
- Y02—TECHNOLOGIES OR APPLICATIONS FOR MITIGATION OR ADAPTATION AGAINST CLIMATE CHANGE
- Y02A—TECHNOLOGIES FOR ADAPTATION TO CLIMATE CHANGE
- Y02A90/00—Technologies having an indirect contribution to adaptation to climate change
- Y02A90/10—Information and communication technologies [ICT] supporting adaptation to climate change, e.g. for weather forecasting or climate simulation
Abstract
The invention provides a rainfall estimation method based on microwave rain attenuation, which comprises the following steps: acquiring microwave rainfall attenuation data and corresponding rainfall data of an estimation area, and obtaining a rainfall estimation result to be corrected through a microwave rainfall inversion model; acquiring multi-source non-rainfall data corresponding to the rainfall data, training a depth belief network by using the multi-source non-rainfall data and the microwave rainfall data set, constructing a correction preference model, and correcting and selecting the estimation data output by the microwave rainfall inversion model; and obtaining a preliminary rainfall estimation result of the estimation area through the microwave rainfall inversion model according to an estimation request parameter provided by a user, and correcting and preferring the preliminary rainfall estimation result according to the correction and preferring model to obtain a final rainfall estimation result of the estimation area. The invention also provides a rainfall monitoring system based on the microwave rainfall attenuation and a data processing device for carrying out rainfall estimation by using the rainfall estimation method.
Description
Technical Field
The invention relates to the field of earth science and meteorological science calculation, in particular to an intelligent high-space-time resolution rainfall estimation method and a rainfall monitoring system based on microwaves.
Background
Precipitation is the most important ring in describing climate and is the most important climate index. At present, under the global climate change background, the rainfall in China shows the change trend of 'extreme rainfall frequency increase' and 'uneven rainfall spatial-temporal distribution', which causes a plurality of problems of 'heavy rain disaster frequency', 'flood and drought disaster economic loss increase', 'urban flood disaster serious', 'mountain flood geological disaster increase' and the like. As a direct trigger condition of flood and drought disasters, the method is an important task in China to strengthen the research on rainfall monitoring, measures the rainfall intensity accurately and timely, has important scientific significance and application value in various aspects such as hydrological forecast, agricultural production, people's life, social operation, army guarantee, national safety and the like.
The rainfall station can intuitively and effectively obtain rainfall actual measurement data at the first time. At present, through the years of construction of national flood control command system engineering, medium and small river and mountain flood early warning, meteorological monitoring and other projects, the rainfall station observation in China has a certain quantity and scale, but the rainfall monitoring is still limited by three aspects. Firstly, at high height and high altitude river estuary region, well mountain area, desert district in china, rainfall station net density is low, and rainfall station often can not normal continuous operation moreover, and simultaneously, in remote mountain area city, traditional rainfall telemetry station needs a large amount of operation and personnel maintenance cost, and many places do not possess the station condition of building. Secondly, in urban rainfall monitoring, rainfall radar echoes are easily influenced by clutters and obstacles, results are inaccurate, urban micro-terrains and local complex terrains have large influence on measurement results, and regional radar coverage has many blind areas, so that the real-time response requirement for urban and suburban emergency treatment cannot be met. Thirdly, local rainfall monitoring means are insufficient, the space is mainly represented as low density of a western monitoring station network and limited monitoring of an eastern city, the time is mainly limited by rainfall monitoring delay, low time-space resolution and the like, and the problems of insufficient transmission distance, signal noise, terrain interference and the like exist in the monitoring technology.
Traditional rainfall monitoring is based on a rain gauge, a weather radar and a remote sensing satellite, but is limited in time and space, the rainfall on the earth surface cannot be accurately estimated, and the requirement of good real-time monitoring is difficult to guarantee.
The rainfall monitoring based on the microwave base station link can achieve real-time monitoring of real rainfall on the urban ground surface, and a large amount of networking distribution is not needed. Meanwhile, the wireless communication network in China is wide in coverage, high in signal quality and basically free of blind areas, and therefore a good building platform is provided for high-space-time-resolution precipitation monitoring.
Therefore, based on the above problems, the method for performing all-weather high-spatial-temporal-resolution real-time monitoring and analysis on local rainfall has extremely important significance for flood and drought disaster prevention, water resource development and utilization, ecological environment protection and the like, can be widely popularized and applied in the industries of weather, traffic, emergency, agriculture, environmental protection and the like, has extremely high production and research values, and has good social and economic benefits.
Disclosure of Invention
The invention provides a rainfall estimation method based on microwave rainfall attenuation, which is characterized in that a rainfall estimation result of a microwave rainfall inversion model is corrected and preferred through a depth belief network, so that a high-space-time resolution rainfall estimation result is obtained.
Specifically, the rainfall estimation method of the present invention includes: training a deep belief network by using microwave rain attenuation data and corresponding rainfall data of the estimation area and multi-source non-rainfall data corresponding to the rainfall data, and constructing a correction preferential model; obtaining a preliminary rainfall estimation result of the estimation area through a microwave rainfall inversion model according to an estimation request parameter provided by a user, and correcting and preferring the preliminary rainfall estimation result according to the correction and preferring model to obtain a final rainfall estimation result of the estimation area; generating an estimation response for the final rainfall estimation result to feed back to the user; wherein the estimation response comprises at least one of a data format file, real-time output data information, and real-time output visualization information.
The rainfall estimation method comprises the following steps that the microwave rainfall attenuation data comprise rainfall effective attenuation values of a microwave link; the step of obtaining the rainfall effective attenuation value specifically comprises the following steps: according to the space coordinate data of the high-frequency microwave rainfall ground monitoring station and the station microwave link signal data in the estimation area, the microwave link attenuation A obtained by actual measurement during rainfall is obtainedtMicrowave link attenuation in clear sky conditions ADAnd other attenuations AMTo obtain the effective attenuation value A of rainfallR＝At-AD-AM(ii) a And acquiring rainfall effective attenuation values of a plurality of microwave links of a plurality of microwave frequency bands between 10 and 40 GHz.
The rainfall estimation method provided by the invention comprises the following steps of: gridding the estimation region into a plurality of grid regions; obtaining a first rain attenuation relation of the microwave link i
Wherein A isiIs the effective attenuation value of rainfall, R, of the microwave link iiAverage rain intensity observed for microwave link i, diIs the length, k, of the microwave link ii、αiIs a littleThe rain attenuation conversion constant of the wave link i; obtaining a second rain attenuation relation of the microwave link i by using the propagation attenuation of each grid area of the microwave link i
Wherein r isjRain intensity of grid area j of microwave link i, lijThe length of the microwave link i in the grid area j is shown, and M is the number of the grid areas of the microwave link i; the first rain attenuation relation and the second rain attenuation relation are equal
Obtaining the rain strength r of the grid area jj(ii) a And obtaining a preliminary rainfall estimation result of the estimation area according to the rainfall intensity of all the grid areas.
The rainfall estimation method comprises the following steps: virtual temperature, potential temperature, ground dew point temperature, specific humidity, water vapor density and water vapor content.
The invention also provides a rainfall monitoring system based on microwave rain attenuation, which comprises: the model building module is used for training a deep belief network by using the microwave rain attenuation data and the corresponding rainfall data of the estimation area and the multi-source non-rainfall data corresponding to the rainfall data, and building a correction preference model; the rainfall estimation module is used for obtaining a preliminary rainfall estimation result of the estimation area through a microwave rainfall inversion model according to an estimation request parameter provided by a user, and correcting and preferring the preliminary rainfall estimation result according to the correction and preferring model to obtain a final rainfall estimation result of the estimation area; the result feedback module is used for generating an estimation response for the final rainfall estimation result and feeding back the estimation response to the user; wherein the estimation response comprises at least one of a data format file, real-time output data information, and real-time output visualization information.
The rainfall monitoring system comprises a microwave link, a microwave rain attenuation data acquisition unit and a rainfall monitoring unit, wherein the microwave rain attenuation data comprises a rainfall effective attenuation value of the microwave link; the model building module comprises: the microwave rain attenuation data acquisition module is used for acquiring microwave rain attenuation data; wherein rainfall is performed according to the high-frequency microwave in the estimation areaSpace coordinate data of ground monitoring station and station microwave link signal data, and microwave link attenuation A obtained by actual measurement in rainfalltMicrowave link attenuation in clear sky conditions ADAnd other attenuations AMTo obtain the effective attenuation value A of rainfallR＝At-AD-AM(ii) a And acquiring rainfall effective attenuation values of a plurality of microwave links of a plurality of microwave frequency bands between 10 and 40 GHz.
The rainfall monitoring system of the invention, wherein the rainfall estimation module includes: the preliminary rainfall estimation result acquisition module is used for acquiring the preliminary rainfall estimation result; the module for obtaining the preliminary rainfall estimation result specifically comprises: a region gridding module for gridding the estimation region into a plurality of grid regions;
the grid region rain intensity estimation module is used for acquiring the rain intensity of the grid region; obtaining a first rain attenuation relation of the microwave link i
Obtaining a second rain attenuation relation of the microwave link i by using the propagation attenuation of each grid area of the microwave link i
Wherein R isiAverage rain intensity (mm/h), d, observed for microwave link iiIs the length, k, of the microwave link ii、αiIs the rain attenuation conversion constant, r, of the microwave link ijRain intensity of grid area j of microwave link i, lijThe length of the microwave link i in the grid area j is shown, and M is the number of the grid areas of the microwave link i; the first rain attenuation relation and the second rain attenuation relation are equal
Obtaining the rain strength r of the grid area jj(ii) a And the regional rainfall characteristic inversion module is used for obtaining a preliminary rainfall estimation result of the estimation region according to the rainfall intensities of all the grid regions.
The rainfall monitoring system of the invention, wherein the multi-source non-rainfall data includes: virtual temperature, potential temperature, ground dew point temperature, specific humidity, water vapor density and water vapor content.
The present invention further provides a data processing apparatus, comprising: a computer readable storage medium storing executable instructions for performing the rainfall estimation method as described above; a processor for retrieving and executing executable instructions in the computer readable storage medium to perform rainfall estimation according to an estimation request parameter provided by a user.
The method for inverting and estimating the rainfall by calculating the attenuation of the high-frequency microwave signal inverts the indexes such as the shape of the raindrops, the rainfall type, the rainfall intensity and the like, the result directly acts on the real rainfall on the ground surface, the inversion result has high representativeness, and the real-time monitoring can be carried out in a larger range with higher precision; and a set of high-precision high-spatial-temporal-resolution rainfall monitoring system can be formed by combining the rainfall station, the rain measuring radar and the point scale, small scale and large scale observation of the remote sensing satellite; meanwhile, the existing microwave rainfall inversion model is fully utilized, and the inversion result is corrected and optimized by means of a depth belief network model with strong data mining capacity, so that the rainfall estimation precision and the rainfall monitoring system intelligence are improved.
Drawings
FIG. 1 is a flowchart of an embodiment of a rainfall estimation method of the present invention for establishing an estimation model;
FIG. 2 is a flowchart illustrating an embodiment of a rainfall estimation method estimation process according to the present invention;
FIG. 3 is a block diagram of a rainfall monitoring system of the present invention;
FIG. 4 is a flow chart of model construction for the rainfall monitoring system of the present invention.
FIG. 5 is a schematic diagram of a data processing apparatus of the present invention.
Detailed Description
The invention will be further described with reference to the drawings and examples to understand the objects, schemes, functions and effects of the invention.
The intelligent high-space-time resolution rainfall estimation method based on microwaves comprises two coupled processes of model construction and result estimation. The basic idea of the estimation model construction process is as follows: based on a microwave communication transmission technology, extracting a high-frequency wireless microwave signal containing rainfall information according to the characteristics of a rainfall monitoring field with high spatial-temporal resolution; according to the microwave rain attenuation characteristics, estimating correlation characteristics of different microwave frequency bands and rainfall intensity between 10 and 40Ghz to obtain a time series microwave rainfall data set matched in space, and further establishing a high-space-time resolution rainfall estimation model based on the rainfall intensity and rainfall characteristic inversion model and a deep learning algorithm.
The invention aims to solve the technical problem of providing an intelligent high-space-time resolution rainfall estimation method and a monitoring system based on microwaves, which are used for solving the problems of large observation space error of a surface rainfall foundation, radar monitoring signal interference and time lag in remote sensing monitoring.
The invention further solves the problem of high-spatial-temporal-resolution rainfall correction through a deep learning algorithm by introducing multi-source data and combining a microwave rainfall inversion model.
In order to achieve the aim, the intelligent high-space-time resolution rainfall estimation method based on the microwaves comprises the steps of establishing a high-space-time resolution rainfall estimation model based on the microwaves and estimating the high-space-time resolution rainfall based on the microwaves; the step of establishing the microwave-based high spatial-temporal resolution rainfall estimation model further comprises the following steps:
a signal extraction sub-step, which is used for extracting link microwave signal attenuation data according to the microwave measured signals of different links;
a signal characteristic correlation sub-step, which is used for estimating correlation characteristics of input signal attenuation and rain intensity of microwave links in different microwave frequency bands according to the qualitative relation between the microwave rain attenuation characteristics and the rain intensity;
and an estimation model construction sub-step, which is used for constructing an estimation model according to the microwave rainfall inversion model and the depth belief network.
The step of microwave-based high spatial and temporal resolution rainfall estimation further comprises:
an estimation request substep for the pre-processing of the estimation input data and the definition of the estimation configuration file;
and an estimation configuration step, which is used for analyzing the estimation configuration file and determining the input and the organization of different types of data sources under various estimation requirements according to the input data type, classification, quantity and the descriptive file.
And an estimation task substep, namely estimating by combining a microwave rainfall model, correcting and optimizing an estimation result by combining a depth belief network, and outputting the estimation result.
The signal extraction sub-step further comprises:
step 101, constructing a high-frequency microwave rainfall ground monitoring station space data set;
102, estimating correlation characteristics of different microwave frequency bands between 10-40 Ghz and the rain intensity according to a qualitative relation between the microwave rain attenuation characteristics and the rain intensity;
and 103, constructing time sequence microwave rainfall data matched with the model space to obtain a space-time matched path microwave rainfall data set, and taking the space-time matched path microwave rainfall data set as a training data set and an inversion model analysis data set for constructing a high space-time resolution rainfall estimation model.
In the estimation model construction substep, according to the microwave rainfall data set, a high-spatial-temporal-resolution rainfall estimation model based on microwaves is constructed by utilizing a rainfall intensity and rainfall characteristic inversion model and a depth learning algorithm.
The microwave-based high spatial-temporal resolution rainfall estimation step further comprises: and the estimation result correction and optimization sub-step is used for introducing other data such as air temperature, relative humidity and the like, correcting and optimizing the estimation result of the inversion model by constructing a depth belief network, and improving the estimation precision of the microwave rainfall.
The invention also provides a rainfall monitoring system for realizing the method, which comprises the following steps:
and the front-end data acquisition subsystem is used for transmitting and receiving the high-frequency microwave link signals of the station and transmitting the high-frequency microwave link signals with the microwave rain attenuation to the model construction subsystem and the cloud inversion estimation subsystem in real time.
The model construction subsystem is used for constructing a microwave-based high-space-time resolution rainfall estimation model according to the link microwave signal attenuation data and the space-matched time sequence microwave rainfall data;
the cloud inversion estimation subsystem is used for constructing a rainfall estimation model realized by the subsystem according to the model, estimating inverted rainfall data based on input data and estimation requirements, and performing correction and preference selection;
and the estimation result visualization subsystem is used for returning the microwave-based high-space-time resolution rainfall estimation result and visualization thereof and supporting the output and storage of the estimation result.
The rainfall monitoring system can introduce various data when an estimation model is constructed, and high-space-time resolution rainfall correction optimization is carried out by utilizing a deep belief network so as to improve the accuracy of the estimation result of the system.
The front-end data acquisition subsystem further comprises:
the microwave signal acquisition unit is used for transmitting and receiving high-frequency microwave link signals of the station;
and the signal data transmission unit is used for carrying out real-time data transmission communication among the microwave signal acquisition unit, the model construction subsystem and the cloud inversion estimation subsystem.
The model building subsystem further comprises:
a signal extraction unit, which extracts microwave signal rainfall attenuation data according to microwave actual measurement signals of different links by utilizing a signal processing technology and a filtering technology;
the signal characteristic correlation unit is used for constructing a time series microwave rainfall data set of the rainfall intensity, the rainfall characteristic inversion model and the depth belief network which are required by the space matching according to the qualitative relation between the microwave rainfall attenuation characteristic and the rainfall intensity;
and the estimation model construction unit is used for constructing a high-space-time-resolution rainfall estimation model based on microwaves by utilizing the rainfall intensity and rainfall characteristic inversion model and by utilizing the learning, training, optimization and verification of the deep belief network according to the matching data set.
The microwave-based high-spatial-temporal-resolution rainfall cloud inversion estimation subsystem further comprises:
an estimation request unit, which is used for preprocessing the estimation input data and defining the estimation configuration file, and extracting the link microwave signal and other input data;
and the estimation configuration unit determines the input and the organization of different types of data sources under various estimation requirements by analyzing the input data type, classification, quantity and the descriptive file in the estimation configuration file.
And the estimation task unit is used for estimating in combination with a microwave rainfall model, correcting and optimizing the estimation result in combination with a depth belief network, and outputting the estimation result.
The estimation result visualization subsystem further includes:
an estimation result processing unit for performing modification processing of the rainfall estimation result;
a result visualization unit for visualizing the microwave-based high temporal-spatial resolution rainfall estimation result;
a visual output unit: and performing encapsulation output on the result obtained by the estimation result processing unit according to the response result of the visualization unit.
FIG. 1 shows a specific implementation flow of an estimation model construction process in the microwave-based high spatial-temporal resolution rainfall estimation method of the present invention. As shown in fig. 1, in the method for estimating rainfall based on microwave with high spatial-temporal resolution of the present invention, the process of establishing the estimation model further includes the following steps:
and S101, constructing a high-frequency microwave rainfall ground monitoring station space data set.
The spatial data set of the high-frequency microwave rainfall ground monitoring station stores the spatial coordinates of the high-frequency microwave rainfall ground monitoring station and the microwave link signals of the station in a centralized manner, and provides a data basis for the construction of the microwave rainfall data set in the step S103.
And S102, estimating correlation characteristics of different microwave frequency bands between 10 and 40Ghz and rain intensity by combining the basic principle of microwave signal attenuation in the atmosphere according to the microwave rain attenuation characteristics.
The microwave rain attenuation characteristic is that in the propagation process of microwaves in a rain zone, attenuation loss occurs to microwave transmission energy due to the scattering, absorption and other effects of raindrops on the microwaves. The total attenuation A of microwave propagation is calculated differentially by extracting transmitting power and receiving power at a transmitting end and a receiving end of a microwave link, and the calculation formula is as follows:
A＝AW+AO＝0.182fN″，N″＝(p，T，ρ，f)
AWfor total attenuation by water vapor, AOFor the attenuation caused by drying air, f is the microwave frequency, and N' is the composite refractive index imaginary part corresponding to the frequency, and is influenced by the pressure p, the temperature T, the water vapor density rho and the frequency f.
And S102, estimating microwave link signals by using the microwave rain attenuation characteristic, and obtaining the relation between the frequency change of the microwave link in different microwave frequency bands between 10 and 40Ghz and the atmospheric attenuation according to the formula under the condition that the temperature and the atmospheric pressure are relatively stable.
And S103, constructing a time series microwave rainfall data set matched with the model space.
Obtaining an effective attenuation value A caused by rainfall according to the correlation characteristics of the microwave link signals of the high-frequency microwave rainfall ground monitoring station obtained in the step S101 and the atmospheric gas absorption attenuation model in the step S102RThe calculation formula is as follows:
AR＝At-AD-AM，
Atfor the total attenuation of the microwave link transmission measured during rainfall, ADThe attenuation of microwave link path transmission measured under clear sky conditions includes attenuation caused by atmosphere and attenuation caused by rainfall, AMThe microwave attenuation in the rainfall process is obtained by combining the data changes of temperature, air pressure and humidity measured before rainfall and during rainfall and an atmospheric gas absorption attenuation model.
Effective attenuation value A of rainfall obtainedRAnd taking a rain attenuation prediction model given in the international electric association as an inversion theoretical basis. The relation between the rain attenuation value and the rain intensity value can be further established, and the calculation formula is as follows:
AR＝kRα，
ARfor the amount of rain attenuation caused per unit distance, i.e. the effective rain attenuation value (dB/km), R is the intensity of rainfall, i.e. the rain intensity value (mm/h), k, α are frequency dependent coefficients, which can be calculated according to the standard method proposed in ITU-R:
k＝[kH+kV+(kH-kV)cos2θ cos2τ]/2，
d＝[kHαH+kVαV+(kHαH-kVαV)cos2θ cos2τ]/2k，
in the formula, kH，kV，αH，αVThe values of coefficients k and α for horizontal polarization and vertical polarization at the average path, respectively, are related to the frequency, theta is the elevation angle of the receiving wire, tau is the polarization angle of the receiving point wave, and k and α can be calculated by using the corresponding horizontal polarization coefficient and vertical polarization coefficient.
By using the electric wave rain attenuation prediction model, the obtained rainfall effective attenuation value ARThe path rain intensity R is calculated by inversion through a method of solving an equation, and the calculation formula is as follows:
and further obtaining a space-matched time sequence path microwave rainfall data set, and using the space-matched time sequence path microwave rainfall data set as a training data set and an inversion model analysis data set for constructing a high-space-time resolution rainfall estimation model.
And step S104, constructing a microwave-based high-space-time resolution rainfall estimation model according to the microwave rainfall data set.
The high-space-time resolution rainfall estimation model based on the microwaves is constructed based on a rainfall intensity and rainfall characteristic inversion model and a deep learning algorithm.
Preferably, the inversion model inverts the rainfall by using the corrected microwave signal intensity attenuation value, including rainfall characteristic analysis such as basic intensity calculation, rainfall correction, rainfall field calculation, and the like, so as to obtain real-time rainfall data. According to a singleMicrowave rain attenuation relation A on linkR＝kRαThe distribution of the two-dimensional rainfall field can be calculated by carrying out gridding processing on the area and applying the chromatography technology, and the calculation process is as follows:
according to the rain attenuation relation of the whole microwave link
AiFor the total attenuation of propagation, R, of the ith microwave link (microwave link i)iAverage rain intensity (mm/h), d, observed for microwave link iiFor the length of the microwave link i, i is 1 to N, ki、αiIs the rain attenuation conversion constant of the microwave link i. Total attenuation A of microwave link iiIt can also be expressed as the sum of the attenuations of the microwave link i within each grid:
rjfor mesh j the rain intensity to be inverted, lijFor the length of the microwave link i located in the grid area j. Rain attenuation relation of simultaneous whole microwave link i
The following can be obtained:
and obtain
After expansion, an equation set to be solved by the chromatographic model can be obtained:
this system of nonlinear equations can be rewritten in taylor expansion as:
After linearization, the system of linear equations can be solved by an iterative method:
1) initializing r (0) when the iteration step number t is equal to 0;
2) the iteration step number t is t + 1;
3) calculating corresponding P (t), Q (t);
4) calculating r (t) according to a chromatographic model;
5) when T < T, judging | | | r (T) -r (T-1) | <, if passing, outputting r (T), otherwise, returning to 2).
Thus, the rainfall intensities of all grids in the whole area are obtained, and the rainfall field characteristics of the area, namely the rainfall estimation result to be corrected of the area, are inverted.
Preferably, the deep learning algorithm overcomes the process of manually selecting features in a general machine learning method, does not need a very professional related field as prior knowledge, trains a data set with very large capacity and dimensionality through a multi-hidden-layer hierarchical structure type neural network, combines original features of data into highly abstract features, and obtains a relatively simple feature structure from the highly abstract features, so that a classification or regression model with very good robustness is constructed for forecasting rainfall. And step S104, performing auxiliary prediction and correction optimization on rainfall by establishing and training a deep belief network DBNs (deep belief networks). The DBNs are a multi-hidden-layer neural network model formed by limiting Boltzmann machine RBM (restricted Boltzmann machine) as a basic unit. Randomly allocating the space-matched time-series microwave rainfall data set of step S103 to three subsets: training data set (50%), test data set (25%), and validation data set (25%). The data set comprises matched microwave rainfall data and multi-source non-rainfall data, such as virtual temperature, potential temperature, ground dew point temperature, specific humidity, water vapor density, water vapor content and other measured data. The training set is used for training the model, the verification set is used for fine adjustment of the model, and the test set is used for testing the model result. The structure of the constructed DBNs model with multiple layers mainly comprises a Bernoulli RBM and a Gaussian RBM.
The RBM training algorithm is a contrast divergence (CD-k) algorithm, the DBNs training algorithm is a greedy learning algorithm, and the RBMs trained layer by layer are overlapped in the training process and then model weight fine adjustment is carried out. The fine tuning process of the DBNs adopts a W-S (Wake-Sleep) algorithm and a BP algorithm, wherein a "Wake" stage is used for learning and generating weights to obtain input data characteristics, and a "Sleep" stage is used for learning and identifying the weights and correctly recovering data. The model training process is as follows:
1) fully pre-training a first RBM through a CD-k algorithm;
2) fixing the parameters of the 1 st RBM, calculating the hidden layer state of the RBM according to the input vector, and taking the hidden layer state as the input vector of the 2 nd RBM;
3) fully training the 2 nd RBM, and stacking the trained 2 nd RBM above the 1 st RBM;
4) repeating the 3 steps for a plurality of times to construct the DBNs of the N layer.
The model fine tuning process is as follows:
1. except the top layer RBM, the weights of the RBMs of other layers are respectively an upward cognitive weight and a downward generation weight;
2. generating node states of each layer through external feature input and cognitive weight, and updating the generation weight between layers by using gradient descent;
3. and generating a bottom layer state through top layer representation and weight generation, and updating cognitive weight between layers.
The key of model learning lies in rainfall-related feature learning, which is represented by the ability of extracting features of input samples and the ability of fitting complex functions, and the expression of the samples in the original feature space is transformed into a new feature space through layer-by-layer feature transformation and combination, so that rainfall prediction under the support of multi-source real-time data is completed. Through hidden layer weight generation, identification and updating, the obtained network can predict rainfall in real time according to multi-source input data, and correction and optimization of rainfall estimation results in combination with the inversion model are achieved.
FIG. 2 shows a flow of an embodiment of the estimation process of the microwave-based high spatial-temporal resolution rainfall estimation method according to the present invention. As shown in fig. 2, the microwave-based high spatial-temporal resolution rainfall estimation process in the estimation method of the present invention specifically includes the following steps:
step S201, an estimation request is generated according to the estimation requirement of the user.
The user estimation requirements comprise the appointed estimation time and estimation time period, the storage position of microwave rainfall observation data of the corresponding date and the estimation result output position. The estimation request also comprises estimation parameter configuration requirements and a corresponding estimation parameter configuration file. Different estimation parameter configurations are provided for different estimation requirements such as rainage identification, rainfall field calculation and the like aiming at different data sources and different estimation requirements. In the case of the introduction of air temperature data, relative humidity data, different estimated parameter profiles may occur.
Step S202, estimating according to the user estimation requirement and the estimation request.
According to the estimation time and the estimation time period in the retrieval request, the storage position of the microwave rainfall observation data on the corresponding date, the estimation parameter configuration requirement, the estimation parameter configuration file and the estimation result output position in the step S201, the result estimation is performed.
Step S203, the estimation result is analyzed, inverted, corrected and optimized.
And for various estimation results with different estimation requirements, carrying out analysis inversion by using a rainfall characteristic analysis model and carrying out correction optimization by combining a deep learning algorithm to obtain the final estimation results of the attention area, the attention time period and the attention requirement of the user.
And step S204, generating an estimation response and returning an estimation result.
Generating an estimation response to encapsulate the estimation result, and performing a comprehensive output process according to the estimation request, the estimation result output position and the final estimation result of S203 in step S201. For example, various optional data format files such as DBF format, DAT format or TIFF format are output or real-time estimation results are directly output and returned to the user on a system interaction panel.
Step S205, visualizing the estimation result and packaging and outputting the visualization execution result.
Visualization of the estimation result is realized based on an ArcGIS software platform, a database technology and a visualization analysis technology, and the estimation result is visually and visually displayed to a user in real time or is packaged and output.
As shown in FIG. 3, the present invention illustrates a microwave-based high spatial-temporal resolution rainfall monitoring system, which comprises four modules: the system comprises a front-end data acquisition subsystem, a model construction subsystem, a cloud inversion estimation subsystem and a visualization subsystem.
The front-end data acquisition subsystem 301 includes: microwave signal acquisition unit 3011, signal data transmission unit 3012.
The microwave signal acquisition unit 3011 transmits and receives high-frequency microwave link signals of a site, and the signal data transmission unit 3012 performs data communication between the microwave signal acquisition unit 3011 and the model construction subsystem 302 and the cloud inversion estimation subsystem 303, and transmits the high-frequency microwave link signals with microwave rain attenuation to the signal extraction unit 3021 and the cloud inversion estimation subsystem 303 in real time.
The model building subsystem 302 includes: a signal extraction unit 3021, a signal feature association unit 3022, and a model construction unit 3023.
The signal extraction unit 3021 extracts link microwave signal attenuation data by combining a microwave rain attenuation characteristic and a filtering operator according to different link microwave actual measurement signals provided by the microwave signal acquisition unit 3011.
The signal feature correlation unit 3022 relates to a time series microwave rainfall dataset task of constructing a spatial matching required by the rainfall intensity and rainfall feature inversion model and the depth belief network. The principle is that according to the microwave rain attenuation characteristic, the correlation characteristic of the attenuation and the rain intensity of microwave link input signals of different microwave frequency bands is estimated.
The signal extraction unit 3021 and the signal feature association unit 3022 exist in the model construction unit 3023 in a static manner, and provide a microwave rainfall data set in the cloud inversion estimation subsystem 303, thereby avoiding repeated construction of the signal extraction unit and the signal feature association unit.
The model construction unit 3023 relates to learning, training, optimizing, and verifying a rainfall intensity and rainfall feature inversion model and a deep belief network, provided that the signal feature association unit 3022 provides a microwave rainfall data set.
The estimation model construction process is shown in fig. 4, and the implementation process of the model construction subsystem specifically includes:
step S401, constructing a high-frequency microwave rainfall ground monitoring station space data set, wherein the high-frequency microwave rainfall ground monitoring station space data set comprises high-frequency microwave rainfall ground monitoring station space coordinates and station microwave link signals;
step S402, estimating correlation characteristics of different microwave frequency bands between 10 and 40Ghz and rain intensity according to the microwave rain attenuation characteristics;
step S403, obtaining a space-matched time series path microwave rainfall data set according to the correlation characteristics of the microwave rainfall;
step S404, analyzing and inverting the microwave rainfall data set according to different estimation requirements by combining a microwave rainfall inversion model;
step S405, distributing the microwave rainfall data set into a training data set, a testing data set and a verification data set at random according to a certain proportion, and enabling the training data set, the testing data set and the verification data set to enter a deep belief network for learning, training, optimizing and verifying;
and S406, performing auxiliary prediction and correction optimization on the estimation result of the microwave rainfall inversion model by using a depth belief network, and constructing an intelligent high-space-time resolution rainfall estimation model based on microwaves.
Estimation subsystem 303 includes: an evaluation request unit 3031, an evaluation configuration unit 3022, and an evaluation task unit 3033.
The evaluation request unit 3031 relates to an evaluation configuration file, which may be in TXT format or XML format, consisting of three elements, input location, evaluation mode, output location. The estimation mode relates to the definition of the type, classification and quantity of input data, and the selection and definition of the estimation region, the estimation period and the estimation method.
The estimation configuration unit 3022 provides a modular abstraction method for different data sources required by different estimation requirements, supports the input and organization of different types of data sources under multiple estimation requirements, and determines which data source types should be used for access and invocation in different estimation processes by parsing the input data types, classifications, quantities, and descriptive files in the estimation configuration file.
The estimation task unit 3033 is the core of the estimation subsystem. The evaluation task unit 3033 is implemented based on the signal extraction unit 3021 and the signal feature association unit 3022. According to the rainfall intensity and rainfall characteristic inversion model and the depth belief network model provided by the model construction unit 3023, the estimation request unit 3031 reads the estimation data source content preprocessed by the signal extraction unit 3021 and the signal characteristic association unit 3022, and estimates the corresponding region, the corresponding time period, and the corresponding requirement by combining the estimation configuration file information analyzed by the estimation configuration unit 3022. After the estimation is completed, a corresponding estimation result file is returned according to the output position defined by the estimation configuration unit 3022, or the input estimation result processing unit 3041 directly displays the estimation result file to the user on the system interaction panel through the visualization unit 3042.
The visualization subsystem 304 includes: an estimation result processing unit 3041, a visualization unit 3042, and a visualization output unit 3043.
The estimation result processing unit 3041: based on the database technology and the visual analysis technology, the real-time microwave rainfall estimation result obtained by the estimation subsystem 303 is visually processed.
The visualization unit 3042: visualization is performed according to the result of the estimation result processing unit 3041, and an arcgissenserver or an ArcGIS Engine is encapsulated to perform visual display of the estimation result.
The visual output unit 3043: the result obtained by the estimation result processing unit 3041 is encapsulated and output according to the response result of the visualization unit 3042.
FIG. 5 is a schematic diagram of a data processing apparatus of the present invention. As shown in fig. 5, an embodiment of the present invention further provides a data processing apparatus, including: a computer-readable storage medium, and a processor. The computer readable storage medium of the present invention stores executable instructions, which when executed by a processor of a data processing apparatus, implement the rainfall estimation method described above. It will be understood by those skilled in the art that all or part of the steps of the above method may be implemented by instructing relevant hardware (e.g., processor, FPGA, ASIC, etc.) through a program, and the program may be stored in a readable storage medium, such as a read-only memory, a magnetic or optical disk, etc. All or some of the steps of the above embodiments may also be implemented using one or more integrated circuits. Accordingly, the modules in the above embodiments may be implemented in hardware, for example, by an integrated circuit, or in software, for example, by a processor executing programs/instructions stored in a memory. Embodiments of the invention are not limited to any specific form of hardware or software combination.
Compared with the prior art, the invention has the beneficial technical effects that:
(1) indexes such as raindrop shape, rainfall type and rain intensity are inverted by calculating attenuation of high-frequency microwave signals, the result directly acts on real surface rainfall, the inversion result is high in representativeness, and real-time monitoring can be achieved in a large range with high precision.
(2) A set of high-precision rainfall monitoring system with high spatial and temporal resolution can be formed by combining the point scale, small scale and large scale observation of a rainfall station, a rain measuring radar and a remote sensing satellite.
(3) The existing microwave rainfall inversion model is fully utilized, and meanwhile, the inversion result is corrected and optimized by means of a depth belief network model with strong data mining capacity, so that the rainfall estimation precision and the rainfall monitoring system intelligence are improved.
The above embodiments are only for illustrating the invention and are not to be construed as limiting the invention, and those skilled in the art can make various changes and modifications without departing from the spirit and scope of the invention, therefore, all equivalent technical solutions also fall into the scope of the invention, and the scope of the invention is defined by the claims.
Claims (10)
1. A rainfall estimation method based on microwave rain attenuation is characterized by comprising the following steps:
training a deep belief network by using microwave rain attenuation data and corresponding rainfall data of the estimation area and multi-source non-rainfall data corresponding to the rainfall data, and constructing a correction preferential model;
and according to an estimation request parameter provided by a user, obtaining a preliminary rainfall estimation result of the estimation area through a microwave rainfall inversion model, and correcting and preferring the preliminary rainfall estimation result according to the correction and preferring model to obtain a final rainfall estimation result of the estimation area.
2. The rainfall estimation method of claim 1, wherein said microwave rain attenuation data comprises rainfall effective attenuation values of a microwave link; the step of obtaining the rainfall effective attenuation value specifically comprises the following steps:
according to the space coordinate data of the high-frequency microwave rainfall ground monitoring station and the station microwave link signal data in the estimation area, the microwave link attenuation A obtained by actual measurement during rainfall is obtainedtMicrowave link attenuation in clear sky conditions ADAnd other attenuations AMTo obtain the effective attenuation value A of rainfallR＝At-AD-AM；
And acquiring rainfall effective attenuation values of a plurality of microwave links of a plurality of microwave frequency bands between 10 and 40 GHz.
3. The rainfall estimation method of claim 2, wherein the step of obtaining preliminary rainfall estimation results for the estimation region comprises:
gridding the estimation region into a plurality of grid regions;
obtaining a first rain attenuation relation of the microwave link i
Wherein A isiIs the effective attenuation value of rainfall, R, of the microwave link iiAverage rain intensity observed for microwave link i, diIs the length, k, of the microwave link ii、αiBeing microwave links iRain attenuation conversion constant;
obtaining a second rain attenuation relation of the microwave link i by using the propagation attenuation of each grid area of the microwave link i
Wherein r isjRain intensity of grid area j of microwave link i, lijThe length of the microwave link i in the grid area j is shown, and M is the number of the grid areas of the microwave link i;
the first rain attenuation relation and the second rain attenuation relation are equal
Obtaining the rain strength r of the grid area jj；
And obtaining a preliminary rainfall estimation result of the estimation area according to the rainfall intensity of all the grid areas.
4. The rainfall estimation method of claim 1, wherein said multi-source non-rainfall data comprises: virtual temperature, potential temperature, ground dew point temperature, specific humidity, water vapor density and water vapor content.
5. The rainfall estimation method of claim 1, further comprising:
generating an estimation response for the final rainfall estimation result to feed back to the user;
wherein the estimation response comprises at least one of a data format file, real-time output data information, and real-time output visualization information.
6. The utility model provides a rainfall monitoring system based on microwave rain decay which characterized in that includes:
the model building module is used for training a deep belief network by using the microwave rain attenuation data and the corresponding rainfall data of the estimation area and the multi-source non-rainfall data corresponding to the rainfall data, and building a correction preference model;
the rainfall estimation module is used for obtaining a preliminary rainfall estimation result of the estimation area through a microwave rainfall inversion model according to an estimation request parameter provided by a user, and correcting and preferring the preliminary rainfall estimation result according to the correction and preferring model to obtain a final rainfall estimation result of the estimation area;
the result feedback module is used for generating an estimation response for the final rainfall estimation result and feeding back the estimation response to the user; wherein the estimation response comprises at least one of a data format file, real-time output data information, and real-time output visualization information.
7. The rainfall monitoring system of claim 6, wherein the microwave rain attenuation data comprises rainfall effective attenuation values for the microwave link; the model building module comprises:
the microwave rain attenuation data acquisition module is used for acquiring microwave rain attenuation data; according to space coordinate data of a high-frequency microwave rainfall ground monitoring station and station microwave link signal data in the estimation area, microwave link attenuation A obtained by actual measurement during rainfall is obtainedtMicrowave link attenuation in clear sky conditions ADAnd other attenuations AMTo obtain the effective attenuation value A of rainfallR＝At-AD-AM(ii) a And acquiring rainfall effective attenuation values of a plurality of microwave links of a plurality of microwave frequency bands between 10 and 40 GHz.
8. The rainfall monitoring system of claim 7, wherein the rainfall estimation module comprises: the preliminary rainfall estimation result acquisition module is used for acquiring the preliminary rainfall estimation result; the module for obtaining the preliminary rainfall estimation result specifically comprises:
a region gridding module for gridding the estimation region into a plurality of grid regions;
the grid region rain intensity estimation module is used for acquiring the rain intensity of the grid region; obtaining a first rain attenuation relation of the microwave link i
Propagation per grid area with microwave link iAttenuating to obtain a second rain attenuation relation of the microwave link i
Wherein R isiAverage rain intensity (mm/h), d, observed for microwave link iiIs the length, k, of the microwave link ii、αiIs the rain attenuation conversion constant, r, of the microwave link ijRain intensity of grid area j of microwave link i, lijThe length of the microwave link i in the grid area j is shown, and M is the number of the grid areas of the microwave link i; the first rain attenuation relation and the second rain attenuation relation are equal
Obtaining the rain strength r of the grid area jj；
And the regional rainfall characteristic inversion module is used for obtaining a preliminary rainfall estimation result of the estimation region according to the rainfall intensities of all the grid regions.
9. The rainfall monitoring system of claim 6, wherein the multi-source non-rainfall data comprises: virtual temperature, potential temperature, ground dew point temperature, specific humidity, water vapor density and water vapor content.
10. A data processing apparatus, comprising:
computer-readable storage medium storing executable instructions for performing the method of rain estimation according to any of claims 1 to 5
A processor for retrieving and executing executable instructions in the computer readable storage medium to perform rainfall estimation according to an estimation request parameter provided by a user.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202010386967.3A CN111666656A (en) | 2020-05-09 | 2020-05-09 | Rainfall estimation method and rainfall monitoring system based on microwave rainfall attenuation |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202010386967.3A CN111666656A (en) | 2020-05-09 | 2020-05-09 | Rainfall estimation method and rainfall monitoring system based on microwave rainfall attenuation |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN111666656A true CN111666656A (en) | 2020-09-15 |
Family
ID=72383177
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202010386967.3A Pending CN111666656A (en) | 2020-05-09 | 2020-05-09 | Rainfall estimation method and rainfall monitoring system based on microwave rainfall attenuation |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN111666656A (en) |
Cited By (16)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN112213727A (en) * | 2020-10-15 | 2021-01-12 | 国家卫星气象中心(国家空间天气监测预警中心) | Precipitation correction method of satellite-borne radar based on active and passive microwave combined detection |
| CN113009595A (en) * | 2021-02-09 | 2021-06-22 | 海南省气象科学研究所 | Rainfall monitoring method based on vehicle-mounted sensor and related device |
| 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 |
| CN113114388A (en) * | 2021-04-13 | 2021-07-13 | 中国人民解放军国防科技大学 | Antenna wetting loss determination method based on level time sequence dynamic memory |
| CN113177360A (en) * | 2021-05-06 | 2021-07-27 | 中国人民解放军国防科技大学 | Rain intensity self-adaptive estimation system and method based on clustering rain attenuation relation |
| CN113466969A (en) * | 2021-05-11 | 2021-10-01 | 深圳捷豹电波科技有限公司 | Rainfall monitoring method, receiving device, rainfall monitoring system and storage medium |
| CN113589405A (en) * | 2021-07-06 | 2021-11-02 | 中国科学院国家空间科学中心 | GNSS polarized signal forward modeling method and system suitable for small rainfall magnitude |
| CN113672864A (en) * | 2021-07-15 | 2021-11-19 | 中国电波传播研究所（中国电子科技集团公司第二十二研究所） | Weighted algorithm applied to rain attenuation statistical prediction for annual average rain top height |
| CN113962083A (en) * | 2021-10-21 | 2022-01-21 | 中国人民解放军国防科技大学 | Microwave lean information rainfall field construction method based on sparse representation |
| CN114509765A (en) * | 2021-12-31 | 2022-05-17 | 江苏微之润智能技术有限公司 | Method for automatically adjusting counter wave mechanical rotating shaft by utilizing deep reinforcement learning |
| 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 |
| CN114814993A (en) * | 2022-03-25 | 2022-07-29 | 河海大学 | Microwave attenuation snowfall intensity monitoring method based on DCGAN and 2D-CNN |
| CN114970340A (en) * | 2022-05-18 | 2022-08-30 | 河海大学 | Commercial microwave inversion rainfall-driven urban road ponding simulation prediction substitution method under incomplete information |
| 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 |
| CN116383701A (en) * | 2023-04-13 | 2023-07-04 | 中国人民解放军国防科技大学 | Microwave link rainy period detection method based on learning reconstruction |
| GB2621908A (en) * | 2022-05-13 | 2024-02-28 | Ho Hai Univ | Hail weather identification and classification method based on multi-channel deep residual shrinkage network |
Citations (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN106547036A (en) * | 2016-10-21 | 2017-03-29 | 中国人民解放军理工大学 | A kind of united Regional Precipitation measuring method of multiband microwave link |
| CN109272146A (en) * | 2018-08-23 | 2019-01-25 | 河海大学 | A kind of Forecasting Flood method corrected based on deep learning model and BP neural network |
-
2020
- 2020-05-09 CN CN202010386967.3A patent/CN111666656A/en active Pending
Patent Citations (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN106547036A (en) * | 2016-10-21 | 2017-03-29 | 中国人民解放军理工大学 | A kind of united Regional Precipitation measuring method of multiband microwave link |
| CN109272146A (en) * | 2018-08-23 | 2019-01-25 | 河海大学 | A kind of Forecasting Flood method corrected based on deep learning model and BP neural network |
Cited By (26)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN112213727A (en) * | 2020-10-15 | 2021-01-12 | 国家卫星气象中心(国家空间天气监测预警中心) | Precipitation correction method of satellite-borne radar based on active and passive microwave combined detection |
| CN112213727B (en) * | 2020-10-15 | 2024-01-02 | 国家卫星气象中心(国家空间天气监测预警中心) | Precipitation correction method of satellite-borne radar based on active and passive microwave combined detection |
| CN113009595A (en) * | 2021-02-09 | 2021-06-22 | 海南省气象科学研究所 | Rainfall monitoring method based on vehicle-mounted sensor and related device |
| CN113114388B (en) * | 2021-04-13 | 2022-02-15 | 中国人民解放军国防科技大学 | Antenna wetting loss determination method based on level time sequence dynamic memory |
| CN113114388A (en) * | 2021-04-13 | 2021-07-13 | 中国人民解放军国防科技大学 | Antenna wetting loss determination method based on level time sequence dynamic memory |
| CN113095590B (en) * | 2021-04-29 | 2022-04-29 | 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 |
| 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 |
| CN113177360A (en) * | 2021-05-06 | 2021-07-27 | 中国人民解放军国防科技大学 | Rain intensity self-adaptive estimation system and method based on clustering rain attenuation relation |
| CN113177360B (en) * | 2021-05-06 | 2022-04-29 | 中国人民解放军国防科技大学 | Rain intensity self-adaptive estimation system and method based on clustering rain attenuation relation |
| CN113466969A (en) * | 2021-05-11 | 2021-10-01 | 深圳捷豹电波科技有限公司 | Rainfall monitoring method, receiving device, rainfall monitoring system and storage medium |
| CN113466969B (en) * | 2021-05-11 | 2023-10-27 | 深圳捷豹电波科技有限公司 | Rainfall monitoring method, receiving equipment, rainfall monitoring system and storage medium |
| CN113589405A (en) * | 2021-07-06 | 2021-11-02 | 中国科学院国家空间科学中心 | GNSS polarized signal forward modeling method and system suitable for small rainfall magnitude |
| CN113672864A (en) * | 2021-07-15 | 2021-11-19 | 中国电波传播研究所（中国电子科技集团公司第二十二研究所） | Weighted algorithm applied to rain attenuation statistical prediction for annual average rain top height |
| CN113672864B (en) * | 2021-07-15 | 2023-08-15 | 中国电波传播研究所（中国电子科技集团公司第二十二研究所） | Annual average roof height weighting algorithm applied to statistical prediction of rain attenuation |
| CN113962083A (en) * | 2021-10-21 | 2022-01-21 | 中国人民解放军国防科技大学 | Microwave lean information rainfall field construction method based on sparse representation |
| CN114509765A (en) * | 2021-12-31 | 2022-05-17 | 江苏微之润智能技术有限公司 | Method for automatically adjusting counter wave mechanical rotating shaft by utilizing deep reinforcement learning |
| CN114814993B (en) * | 2022-03-25 | 2023-04-18 | 河海大学 | Microwave attenuation snowfall intensity monitoring method based on DCGAN and 2D-CNN |
| CN114814993A (en) * | 2022-03-25 | 2022-07-29 | 河海大学 | Microwave attenuation snowfall intensity monitoring method based on DCGAN and 2D-CNN |
| 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 |
| GB2621908A (en) * | 2022-05-13 | 2024-02-28 | Ho Hai Univ | Hail weather identification and classification method based on multi-channel deep residual shrinkage network |
| CN114970340B (en) * | 2022-05-18 | 2023-05-05 | 河海大学 | Urban road ponding simulation prediction substitution method driven by commercial microwave inversion rainfall under incomplete information |
| CN114970340A (en) * | 2022-05-18 | 2022-08-30 | 河海大学 | Commercial microwave inversion rainfall-driven urban road ponding simulation prediction substitution method under incomplete information |
| 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 |
| CN116383701A (en) * | 2023-04-13 | 2023-07-04 | 中国人民解放军国防科技大学 | Microwave link rainy period detection method based on learning reconstruction |
| CN116383701B (en) * | 2023-04-13 | 2024-01-26 | 中国人民解放军国防科技大学 | Microwave link rainy period detection method based on learning reconstruction |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN111666656A (en) | Rainfall estimation method and rainfall monitoring system based on microwave rainfall attenuation | |
| US20220043182A1 (en) | Spatial autocorrelation machine learning-based downscaling method and system of satellite precipitation data | |
| CN105243435B (en) | A kind of soil moisture content prediction technique based on deep learning cellular Automation Model | |
| CN111651885A (en) | Intelligent sponge urban flood forecasting method | |
| Zhang et al. | Assimilation of precipitation-affected radiances in a cloud-resolving WRF ensemble data assimilation system | |
| CN110232471B (en) | Rainfall sensor network node layout optimization method and device | |
| CN109871637B (en) | Near-ground air temperature estimation method under cloud-sky condition | |
| CN105988146A (en) | Application data processing method of spaceborne microwave radiometer | |
| CN111210483B (en) | Simulated satellite cloud picture generation method based on generation of countermeasure network and numerical mode product | |
| Kidd et al. | Digital mapping of a soil drainage index for irrigated enterprise suitability in Tasmania, Australia | |
| Eingrüber et al. | Climate change simulation and trend analysis of extreme precipitation and floods in the mesoscale Rur catchment in western Germany until 2099 using Statistical Downscaling Model (SDSM) and the Soil & Water Assessment Tool (SWAT model) | |
| Baik et al. | Agricultural drought assessment based on multiple soil moisture products | |
| CN105069295A (en) | Assimilation method for satellite and ground rainfall measured values based on Kalman filtering | |
| CN112329265A (en) | Satellite remote sensing rainfall refinement space estimation method and system | |
| Fan et al. | A comparative study of four merging approaches for regional precipitation estimation | |
| CN114048944A (en) | Estimation method for people to be evacuated and houses to be damaged under rainstorm induced geological disaster | |
| Zhang et al. | Assimilation of MWHS and MWTS radiance data from the FY-3A satellite with the POD-3DEnVar method for forecasting heavy rainfall | |
| Thiha et al. | Assessment of the Myitnge River flow responses in Myanmar under changes in land use and climate | |
| CN112285808B (en) | Method for reducing scale of APHRODITE precipitation data | |
| Haddjeri et al. | Exploring the sensitivity to precipitation, blowing snow, and horizontal resolution of the spatial distribution of simulated snow cover | |
| CN110596017B (en) | Hyperspectral image soil heavy metal concentration assessment method based on space weight constraint and variational self-coding feature extraction | |
| CN109948175B (en) | Satellite remote sensing albedo missing value inversion method based on meteorological data | |
| CN114819737B (en) | Method, system and storage medium for estimating carbon reserves of highway road vegetation | |
| CN114912075A (en) | Soil humidity data missing filling method fusing spatial-temporal three-dimensional information of site and satellite remote sensing observation | |
| CN114239274A (en) | Method for calculating root layer soil water by multi-source remote sensing data driven index filtering model |
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 |