CN117369026A - Real-time high-precision cloud cluster residence time prediction method - Google Patents
Real-time high-precision cloud cluster residence time prediction method Download PDFInfo
- Publication number
- CN117369026A CN117369026A CN202311656804.2A CN202311656804A CN117369026A CN 117369026 A CN117369026 A CN 117369026A CN 202311656804 A CN202311656804 A CN 202311656804A CN 117369026 A CN117369026 A CN 117369026A
- Authority
- CN
- China
- Prior art keywords
- cloud
- cloud cluster
- network
- cluster
- calculation
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Granted
Links
- 238000000034 method Methods 0.000 title claims abstract description 63
- 230000033001 locomotion Effects 0.000 claims abstract description 86
- 238000004364 calculation method Methods 0.000 claims abstract description 71
- 238000004422 calculation algorithm Methods 0.000 claims abstract description 52
- 238000005516 engineering process Methods 0.000 claims abstract description 19
- 238000013473 artificial intelligence Methods 0.000 claims abstract description 8
- 239000000284 extract Substances 0.000 claims abstract description 4
- 238000012545 processing Methods 0.000 claims description 22
- 238000004891 communication Methods 0.000 claims description 15
- 238000010801 machine learning Methods 0.000 claims description 15
- 108091092878 Microsatellite Proteins 0.000 claims description 12
- 230000007246 mechanism Effects 0.000 claims description 12
- 238000012544 monitoring process Methods 0.000 claims description 12
- 230000008569 process Effects 0.000 claims description 12
- 230000008859 change Effects 0.000 claims description 11
- 238000013135 deep learning Methods 0.000 claims description 11
- 230000001133 acceleration Effects 0.000 claims description 9
- 239000000443 aerosol Substances 0.000 claims description 9
- 238000004088 simulation Methods 0.000 claims description 8
- 238000012549 training Methods 0.000 claims description 8
- 239000002775 capsule Substances 0.000 claims description 7
- 239000008186 active pharmaceutical agent Substances 0.000 claims description 6
- 238000000605 extraction Methods 0.000 claims description 6
- 230000006870 function Effects 0.000 claims description 6
- 238000010191 image analysis Methods 0.000 claims description 6
- 238000007781 pre-processing Methods 0.000 claims description 6
- 230000004044 response Effects 0.000 claims description 5
- 230000005540 biological transmission Effects 0.000 claims description 4
- 238000013461 design Methods 0.000 claims description 4
- 230000008093 supporting effect Effects 0.000 claims description 4
- 238000010521 absorption reaction Methods 0.000 claims description 3
- 230000032683 aging Effects 0.000 claims description 3
- 238000004458 analytical method Methods 0.000 claims description 3
- 238000013527 convolutional neural network Methods 0.000 claims description 3
- 238000013136 deep learning model Methods 0.000 claims description 3
- 230000002068 genetic effect Effects 0.000 claims description 3
- 230000003993 interaction Effects 0.000 claims description 3
- 230000003287 optical effect Effects 0.000 claims description 3
- 238000005457 optimization Methods 0.000 claims description 3
- 239000002245 particle Substances 0.000 claims description 3
- 238000006116 polymerization reaction Methods 0.000 claims description 3
- 238000005070 sampling Methods 0.000 claims description 3
- 230000011218 segmentation Effects 0.000 claims description 3
- 230000002269 spontaneous effect Effects 0.000 claims description 3
- 238000003860 storage Methods 0.000 claims description 3
- 230000002194 synthesizing effect Effects 0.000 claims description 3
- 230000009286 beneficial effect Effects 0.000 abstract description 3
- 239000000203 mixture Substances 0.000 description 2
- 230000002265 prevention Effects 0.000 description 2
- 230000015572 biosynthetic process Effects 0.000 description 1
- 230000007547 defect Effects 0.000 description 1
- 238000001514 detection method Methods 0.000 description 1
- 238000010586 diagram Methods 0.000 description 1
- 238000009826 distribution Methods 0.000 description 1
- 230000007480 spreading Effects 0.000 description 1
- 238000003892 spreading Methods 0.000 description 1
- 239000013589 supplement Substances 0.000 description 1
Classifications
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01W—METEOROLOGY
- G01W1/00—Meteorology
- G01W1/10—Devices for predicting weather conditions
-
- 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
-
- 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
- G01S17/00—Systems using the reflection or reradiation of electromagnetic waves other than radio waves, e.g. lidar systems
- G01S17/88—Lidar systems specially adapted for specific applications
- G01S17/95—Lidar systems specially adapted for specific applications for meteorological use
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01W—METEOROLOGY
- G01W1/00—Meteorology
- G01W1/02—Instruments for indicating weather conditions by measuring two or more variables, e.g. humidity, pressure, temperature, cloud cover or wind speed
-
- 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
- G06N3/044—Recurrent networks, e.g. Hopfield networks
- G06N3/0442—Recurrent networks, e.g. Hopfield networks characterised by memory or gating, e.g. long short-term memory [LSTM] or gated recurrent units [GRU]
-
- 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
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V10/00—Arrangements for image or video recognition or understanding
- G06V10/20—Image preprocessing
- G06V10/26—Segmentation of patterns in the image field; Cutting or merging of image elements to establish the pattern region, e.g. clustering-based techniques; Detection of occlusion
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V10/00—Arrangements for image or video recognition or understanding
- G06V10/40—Extraction of image or video features
- G06V10/54—Extraction of image or video features relating to texture
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V10/00—Arrangements for image or video recognition or understanding
- G06V10/70—Arrangements for image or video recognition or understanding using pattern recognition or machine learning
- G06V10/82—Arrangements for image or video recognition or understanding using pattern recognition or machine learning using neural networks
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V20/00—Scenes; Scene-specific elements
- G06V20/10—Terrestrial scenes
- G06V20/17—Terrestrial scenes taken from planes or by drones
Landscapes
- Engineering & Computer Science (AREA)
- Physics & Mathematics (AREA)
- Theoretical Computer Science (AREA)
- General Physics & Mathematics (AREA)
- Remote Sensing (AREA)
- Environmental & Geological Engineering (AREA)
- Life Sciences & Earth Sciences (AREA)
- Evolutionary Computation (AREA)
- Multimedia (AREA)
- Radar, Positioning & Navigation (AREA)
- Software Systems (AREA)
- Health & Medical Sciences (AREA)
- Artificial Intelligence (AREA)
- Computing Systems (AREA)
- General Health & Medical Sciences (AREA)
- Electromagnetism (AREA)
- Molecular Biology (AREA)
- Environmental Sciences (AREA)
- Mathematical Physics (AREA)
- Ecology (AREA)
- General Engineering & Computer Science (AREA)
- Biodiversity & Conservation Biology (AREA)
- Computer Networks & Wireless Communication (AREA)
- Atmospheric Sciences (AREA)
- Biomedical Technology (AREA)
- Biophysics (AREA)
- Computational Linguistics (AREA)
- Data Mining & Analysis (AREA)
- Medical Informatics (AREA)
- Databases & Information Systems (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Management, Administration, Business Operations System, And Electronic Commerce (AREA)
Abstract
The invention relates to the technical field of radio wave prediction meteorological, and discloses a method for predicting cloud cluster residence time in real time with high precision, which comprises the following steps: acquiring the position and motion information of the cloud cluster through a satellite cloud image remote sensing technology; predicting the motion trail of the cloud cluster by using an artificial intelligent algorithm; calculating the residence time of the cloud cluster according to the prediction result and the actual observation data; and releasing the calculation result in real time. The real-time high-precision cloud cluster residence time prediction method accurately predicts the residence time of the cloud cluster in real time, provides more accurate data support for weather forecast and disaster early warning, utilizes satellite remote sensing to acquire cloud image information and an artificial intelligence technology to track, predict and calculate the cloud cluster, can realize rapid release of a calculation result in short time, extracts the movement pattern characteristics of the cloud cluster, can improve the prediction accuracy, is beneficial to improving the precision of residence time calculation, and can also cover a large-range area for prediction calculation, thereby meeting the requirements of different application details.
Description
Technical Field
The invention relates to the technical field of radio wave prediction meteorological, in particular to a method for predicting cloud cluster residence time in real time with high precision.
Background
Radio wave prediction of weather is a method of predicting an approaching weather using radio wave propagation characteristics in the atmosphere. Different temperature, humidity composition within the atmosphere can affect the propagation velocity and attenuation of radio waves. By utilizing this feature, the weather elements can be predicted by detecting parameters of radio waves from different directions and estimating the atmospheric composition.
The formation and movement of the cloud layer have decisive influence on weather forecast, the residence time of the cloud layer is mainly predicted through an empirical algorithm or numerical simulation in the prior art, but the accuracy is limited, the monitoring area is limited, and the prediction early warning level is difficult to improve. For this purpose, a corresponding technical solution needs to be designed to solve.
Disclosure of Invention
Aiming at the defects of the prior art, the invention provides a method for predicting cloud residence time in real time with high precision, which solves the technical problems that the accuracy of predicting cloud residence time is limited, the monitoring area is limited, and the prediction early warning level is difficult to be improved in the prior art.
In order to achieve the above purpose, the invention is realized by the following technical scheme: a real-time high-precision cloud cluster residence time prediction method comprises the following steps:
s1, acquiring position and motion information of a cloud cluster through a satellite cloud image remote sensing technology;
s2, predicting the motion trail of the cloud cluster by using an artificial intelligent algorithm;
s3, calculating the residence time of the cloud cluster according to the prediction result and the actual observation data;
and S4, publishing the calculation result in real time.
Preferably, in step S1, the acquiring of the position and motion information of the cloud cluster further includes a multiband camera unmanned plane, a distributed laser radar network, a multi-parameter weather sensor, a millimeter wave Yun Pu instrument, a cloud image intelligent analysis algorithm and a cloud cluster simulation system;
capturing cloud cluster images by utilizing a multiband camera unmanned aerial vehicle, and acquiring cloud cluster boundary and motion vector information by an image analysis technology;
the distributed laser radar network is used for reconstructing a three-dimensional cloud cluster structure in real time by analyzing the reflected signals of airborne particles, and capturing the movement and deformation of the cloud cluster;
using a multi-parameter meteorological sensor, using a balloon or an unmanned aerial vehicle to carry out three-dimensional arrangement, obtaining parameters of temperature, humidity and pressure in the cloud cluster, and presuming the movement of the cloud cluster through change;
by utilizing a millimeter wave Yun Pu instrument, the motion state and the evolution process of the cloud cluster are reversely deduced by measuring the millimeter wave absorption characteristics of the cloud cluster at different positions;
developing an intelligent cloud image analysis algorithm, and performing image processing on cloud video shot by a ground monitoring network to obtain cloud movement information;
and establishing a cloud cluster simulation system with a turbulence model and a cloud micro-physical model coupled, inputting meteorological elements, and simulating cloud cluster evolution and movement processes.
Preferably, in step S1, the cloud cluster position and motion information obtained by satellite cloud image remote sensing specifically includes the following method steps:
a satellite cloud image sensor with double wave bands is used, one wave band is in a visible light range, and the other wave band is in a middle infrared range, and a visible light image and an infrared enhanced image of a cloud cluster are obtained at the same time;
a brand new cloud cluster boundary segmentation algorithm is applied, a convolutional neural network is used for analyzing cloud cluster texture characteristics, and cloud cluster boundaries are accurately extracted;
analyzing the change of the cloud cluster in the continuous phase images by utilizing a spatial relationship graph network, and determining the motion trail of the cloud cluster key points;
establishing an optical flow deep learning model, inputting a two-phase cloud picture, and outputting a motion vector field of each pixel in the cloud cluster;
designing a cloud cluster motion predictor, and based on an LSTM network, synthesizing a cloud cluster boundary, a key point track and a motion vector field to predict a cloud cluster state at a future moment;
using cloud computing and a distributed computing framework to realize rapid processing of mass satellite cloud image data and give out real-time cloud positioning and movement results;
constructing a cloud-end cooperative cloud image processing system, compressing and preprocessing satellite data by a cloud end, and finishing key algorithm calculation by side end equipment;
the system has the functions of autonomous learning and model enhancement, and the algorithm is continuously optimized during satellite operation, so that the extraction precision is improved.
Preferably, in step S2, the method for predicting the cloud cluster motion trail by using the artificial intelligence algorithm specifically includes the following steps:
using a convolution LSTM network as a prediction model to simulate a space-time relationship in the cloud cluster movement process;
the network input comprises a historical multi-time phase cloud chart and meteorological field data, wherein the meteorological field data comprises a wind field and a pressure field, and a cloud cluster motion law is learned;
the network structure adopts a coding-decoding architecture, a coder extracts cloud group motion related characteristics, and a decoder completes future motion prediction;
an attention module is led into the decoder to pay attention to different parts of input data, so that self-adaptive learning is realized;
the network output comprises cloud cluster boundary box sequences of N time periods in the future, and the cloud cluster boundary box sequences represent predicted cloud cluster positions;
li Yongmeng Teslaloy random sampling method enhances data and improves model generalization capability;
designing a genetic algorithm-based super-parameter optimization strategy, and searching out an optimal network structure;
an incremental learning algorithm is applied, and a model is updated by using newly acquired data, so that the prediction accuracy is continuously improved;
and model parallel training is realized on the cloud platform by utilizing multiple GPUs, and the operation process is accelerated.
Preferably, in step S3, the method step of calculating the residence time of the cloud cluster according to the prediction result and the actual observation data comprises the following steps:
based on radar data and weather models: combining radar echo data and a meteorological model, predicting a movement track of the cloud cluster by utilizing the intensity, speed and direction information of the radar echo, and calculating the residence time of the cloud cluster by comparing the movement track with actual observation data;
based on a weather database and a machine learning algorithm: utilizing a large number of weather databases, combining a machine learning algorithm, establishing a correlation model between cloud cluster motion and residence time, and outputting the residence time of the cloud cluster by inputting a prediction result and actual observation data;
based on satellite cloud image texture characteristics and a deep learning algorithm: and predicting the movement track of the cloud cluster by extracting the texture features of the cloud cluster in the satellite cloud picture and calculating the residence time of the cloud cluster by utilizing a deep learning algorithm and combining actual observation data.
Preferably, in step S4, the specific step method for issuing the calculation result in real time includes the following steps:
the unmanned aerial vehicle is used for carrying and deploying a small satellite network, a plurality of microsatellites are deployed at important observation points and prediction points, and the microsatellites are carried with machine learning chips with strong calculation processing capacity, receive cloud image data in real time and perform prediction calculation;
after the calculation is completed, each micro satellite directly transmits the result back to the ground control center through a satellite communication network, and meanwhile, the micro satellite can directly communicate with a nearby unmanned aerial vehicle through a beacon;
the control center pushes the calculation result to emergency centers and user terminals in various places in real time, and the unmanned aerial vehicle nearby directly receives the calculation result and propagates the result to a farther place by utilizing the maneuverability of the unmanned aerial vehicle nearby.
Preferably, the machine learning chip includes, but is not limited to, an AI acceleration processor, an ARM high performance CPU, an FPGA programmable logic device, a high capacity Flash or SSD memory, a general PCIe/M.2 interface, a low power consumption, an API, a bit manipulation and communication interface;
the AI acceleration processor comprises an NVIDIA Jetson series singlechip and is used for preprocessing cloud picture data, deep learning feature extraction and model prediction calculation;
the ARM high-performance central processing unit is matched with the SDRAM with larger capacity to perform online training of models and data;
the FPGA programmable logic device realizes a customized AI acceleration processing function and module, and improves the calculation efficiency;
the high-capacity Flash or SSD memory is provided with a pre-trained machine learning model and an instance database;
the universal PCIe/M.2 interface supports updating and replacing an AI processing module and an algorithm at any time;
the low-power design is suitable for a satellite miniaturized operation environment;
the API is used for integrating a third-party AI algorithm framework and a model;
the bit manipulation and communication interface enables data transmission and result output.
Preferably, the weather forecast and disaster early warning provide accurate data support including aerosol sensing, and specifically include the following steps:
a miniature aerosol sensor that releases a specific component in an atmosphere within a cloud target area;
the micro aerosol sensor capsule is spread along with wind, and the cloud form change information of the passing region is perceived and recorded;
when the number of the sensor capsules reaches a threshold value, a temporary climate sensing network is formed through spontaneous polymerization;
the sensing network completes real-time summarizing and transmitting of cloud information in the area through gas signal interaction;
the weather prediction mechanism collects signals in the atmosphere through specific equipment and decodes the signals into a prediction result;
thus, the instant distributed application in the physical space is realized, and a new supporting channel is provided for decision making.
Preferably, the weather forecast and disaster early warning provide accurate data support to calculate and save cloud cluster prediction results by using a distributed computing network based on a blockchain, and specifically comprise the following steps:
each satellite/unmanned aerial vehicle node participating in calculation and storage is used as a block in a calculation network, and an encryption technology is used for generating a unique ID;
the calculation results are not collected to the central server, but written into each participating node at the same time, and a tamper-proof result chain is formed through a consensus mechanism;
the meteorological and disaster early warning mechanism is directly connected to the distributed network, and the latest snapshot state of the prediction result is obtained in real time;
cloud images and motion data are shared and uploaded to a network in a decentralization mode, so that the safety and auditability of the data are ensured;
and the calculation result directly triggers the release of the early warning event through an intelligent contract mechanism, so that the decision response time is shortened.
Preferably, the weather forecast and disaster early warning provide accurate data support to transmit the calculation result in real time by using a wired network of the unmanned aerial vehicle and a ground base station, and specifically comprise the following steps:
arranging a plurality of ground base stations and a low-altitude spiral unmanned aerial vehicle network in a cloud picture monitoring area;
each base station and a plurality of unmanned aerial vehicles nearby form a small-sized wired communication network;
each unmanned plane is provided with a calculation and sensing module which participates in predictive calculation and information exchange;
after the calculation is finished, the result is immediately uploaded to a base station through a wired channel and then is sent to an emergency command center;
meanwhile, the wired network between the bases transmits the original data in real time for calculation reference;
by increasing the coverage area of the base station, the low-altitude affinity network can rapidly reach large-scale coverage, and the early warning aging requirement is met.
Compared with the prior art, the invention has the beneficial effects that: the residence time of the cloud cluster is accurately predicted in real time, and more accurate data support is provided for weather forecast and disaster early warning; the cloud image information is obtained by satellite remote sensing, and the cloud cluster is tracked, predicted and calculated by an artificial intelligence technology, so that the calculation result can be rapidly released in short time, and the real-time requirements of weather forecast and disaster prevention early warning are met; the artificial intelligence algorithm is used for learning and training a large amount of historical cloud image data, cloud cluster motion mode characteristics are extracted, prediction accuracy can be improved, and the accuracy of residence time calculation is improved by comparing and correcting actual observation data with a prediction result; the satellite remote sensing technology can realize the cloud image monitoring of a wide monitoring area, and the prediction calculation can cover a wide area as well, so that decision support is provided for regional weather forecast and regional disaster early warning; single or multiple cloud clusters can be tracked and predicted independently, and personalized residence time calculation results are given according to the cloud cluster characteristics of different types, sizes and the like, so that the requirements of different application fineness are met; the method has the advantages that not only is the residence time predicted, but also more related cloud cluster information such as movement trend, change rule and the like can be predicted by the reserved space of the method, and the prediction and judgment capability is improved; along with the continuous iterative upgrade of the artificial intelligent algorithm and the sample data, the prediction model and the result can be improved in real time, and the prediction early warning level is kept.
Drawings
FIG. 1 is a schematic diagram of the method steps of the present invention.
Detailed Description
The following description of the embodiments of the present invention will be made clearly and completely with reference to the accompanying drawings, in which it is apparent that the embodiments described are only some embodiments of the present invention, but not all embodiments. All other embodiments, which can be made by those skilled in the art based on the embodiments of the invention without making any inventive effort, are intended to be within the scope of the invention.
Referring to fig. 1, the embodiment of the invention provides a technical scheme: a real-time high-precision cloud cluster residence time prediction method comprises the following steps:
s1, acquiring position and motion information of a cloud cluster through a satellite cloud image remote sensing technology;
s2, predicting the motion trail of the cloud cluster by using an artificial intelligent algorithm;
s3, calculating the residence time of the cloud cluster according to the prediction result and the actual observation data;
and S4, publishing the calculation result in real time.
Further improved, in step S1, the obtaining of the position and motion information of the cloud cluster further includes a multiband camera unmanned plane, a distributed laser radar network, a multi-parameter weather sensor, a millimeter wave Yun Pu instrument, a cloud image intelligent analysis algorithm and a cloud cluster simulation system;
cloud image capturing is carried out by utilizing a multiband camera unmanned aerial vehicle, cloud boundary and motion vector information is obtained through an image analysis technology, and data with higher time and spatial resolution can be obtained compared with a satellite cloud image;
the distributed laser radar network is used, the three-dimensional cloud cluster structure is reconstructed in real time by analyzing the reflected signals of the airborne particles, and the movement and deformation of the cloud cluster are captured, so that the radar has more pertinence compared with the traditional radar;
three-dimensional distribution is carried out by using a multiparameter meteorological sensor and a balloon or an unmanned aerial vehicle, so that parameters of temperature, humidity and pressure in the cloud cluster are obtained, cloud cluster movement is estimated through change, and detailed data in the cloud cluster can be obtained;
by utilizing a millimeter wave Yun Pu instrument, the millimeter wave can be used for perspective cloud layers by measuring the millimeter wave absorption characteristics of cloud clusters at different positions and reversely pushing the motion state and the evolution process of the cloud clusters;
developing an intelligent cloud image analysis algorithm, performing image processing on cloud video shot by a ground monitoring network to obtain cloud movement information, and having low cost;
and a cloud cluster simulation system with a turbulence model and a cloud micro-physical model coupled is established, meteorological elements are input, cloud cluster evolution and movement processes are simulated, and the cloud cluster simulation system is used as supplement of positioning data.
The method can obtain richer and novel positioning and movement information of the cloud cluster, and is beneficial to improving the prediction accuracy of the residence time of the cloud cluster.
Compared with satellite cloud image remote sensing, the method has the advantages that higher time resolution data are obtained, and information updating of a few minutes can be achieved through unmanned aerial vehicles and ground images; acquiring finer spatial resolution data, such as cloud cluster structures with smaller dimensions that can be observed by unmanned aerial vehicles and radar networks; the cloud internal parameters can be obtained, modeling is facilitated, and parameters such as temperature, humidity and the like in the cloud can be observed through multi-parameter detection and millimeter wave technology; lower cost, such as ground camera shooting and analog system is more economical and practical than satellite; the system has better customizability and expansibility, and can design an observation scheme according to requirements.
Further improved, in step S1, the cloud cluster position and motion information obtained by satellite cloud image remote sensing specifically includes the following steps:
a satellite cloud image sensor with double wave bands is used, one wave band is in a visible light range, and the other wave band is in a middle infrared range, and a visible light image and an infrared enhanced image of a cloud cluster are obtained at the same time;
a brand new cloud cluster boundary segmentation algorithm is applied, a convolutional neural network is used for analyzing cloud cluster texture characteristics, and cloud cluster boundaries are accurately extracted;
analyzing the change of the cloud cluster in the continuous phase images by utilizing a spatial relationship graph network, and determining the motion trail of the cloud cluster key points;
establishing an optical flow deep learning model, inputting a two-phase cloud picture, and outputting a motion vector field of each pixel in the cloud cluster;
designing a cloud cluster motion predictor, and based on an LSTM network, synthesizing a cloud cluster boundary, a key point track and a motion vector field to predict a cloud cluster state at a future moment;
using cloud computing and a distributed computing framework to realize rapid processing of mass satellite cloud image data and give out real-time cloud positioning and movement results;
constructing a cloud-end cooperative cloud image processing system, compressing and preprocessing satellite data by a cloud end, and finishing key algorithm calculation by side end equipment;
the system has the functions of autonomous learning and model enhancement, and the algorithm is continuously optimized during satellite operation, so that the extraction precision is improved.
Compared with the traditional method, the scheme utilizes richer satellite data sources and applies the front-edge deep learning technology, so that more accurate cloud cluster motion information can be obtained, and an important supporting effect on prediction of cloud cluster residence time is achieved.
Further improved, in step S2, the method for predicting cloud movement track by artificial intelligence algorithm specifically includes the following steps:
using a convolution LSTM network as a prediction model to simulate a space-time relationship in the cloud cluster movement process;
the network input comprises a historical multi-time phase cloud chart and meteorological field data, wherein the meteorological field data comprises a wind field and a pressure field, and a cloud cluster motion law is learned;
the network structure adopts a coding-decoding architecture, a coder extracts cloud group motion related characteristics, and a decoder completes future motion prediction;
an attention module is led into the decoder to pay attention to different parts of input data, so that self-adaptive learning is realized;
the network output comprises cloud cluster boundary box sequences of N time periods in the future, and the cloud cluster boundary box sequences represent predicted cloud cluster positions;
li Yongmeng Teslaloy random sampling method enhances data and improves model generalization capability;
designing a genetic algorithm-based super-parameter optimization strategy, and searching out an optimal network structure;
an incremental learning algorithm is applied, and a model is updated by using newly acquired data, so that the prediction accuracy is continuously improved;
and model parallel training is realized on the cloud platform by utilizing multiple GPUs, and the operation process is accelerated.
Compared with the existing method, the technical route fully utilizes the advantages of the deep learning technology for modeling the time data, can realize end-to-end high-precision cloud cluster motion prediction, and provides support for subsequent residence time calculation.
Further improved, in step S3, the method steps of calculating the residence time of the cloud according to the prediction result and the actual observation data comprise the following steps:
based on radar data and weather models: combining radar echo data and a meteorological model, predicting the movement track of the cloud cluster by utilizing the intensity, speed and direction information of the radar echo, and calculating the residence time of the cloud cluster by comparing the movement track with actual observation data;
based on a weather database and a machine learning algorithm: utilizing a large number of weather databases, combining a machine learning algorithm, establishing a correlation model between cloud cluster motion and residence time, and outputting the residence time of the cloud cluster by inputting a prediction result and actual observation data;
based on satellite cloud image texture characteristics and a deep learning algorithm: and predicting the movement track of the cloud cluster by extracting the texture features of the cloud cluster in the satellite cloud picture and calculating the residence time of the cloud cluster by utilizing a deep learning algorithm and combining actual observation data.
And different data sources and algorithm technologies are utilized, and the residence time of the cloud cluster is calculated through comparison of a prediction result and actual observation data.
Further improved, in step S4, the specific step method for publishing the calculation result in real time includes the following steps:
the unmanned aerial vehicle is used for carrying and deploying a small satellite network, a plurality of microsatellites are deployed at important observation points and prediction points, and the microsatellites are carried with machine learning chips with strong calculation processing capacity, receive cloud image data in real time and perform prediction calculation;
after the calculation is completed, each micro satellite directly transmits the result back to the ground control center through a satellite communication network, and meanwhile, the micro satellite can directly communicate with a nearby unmanned aerial vehicle through a beacon;
the control center pushes the calculation result to emergency centers and user terminals in various places in real time, and the unmanned aerial vehicle nearby directly receives the calculation result and propagates the result to a farther place by utilizing the maneuverability of the unmanned aerial vehicle nearby. Through unmanned aerial vehicle and miniature satellite network, the calculation result can directly cover the wide area around the predicted point under the condition of no delay at all, solves the problem of delay of the traditional ground-based communication network, and better serves for emergency decision.
Further refinements, machine learning chips include, but are not limited to, AI acceleration processors, ARM high performance central processing units, FPGA programmable logic devices, high capacity Flash or SSD memory, general purpose PCIe/m.2 interfaces, low power consumption, APIs, bit manipulation and communication interfaces;
the AI acceleration processor comprises an NVIDIA Jetson series singlechip and is used for preprocessing cloud picture data, deep learning feature extraction and model prediction calculation;
the ARM high-performance central processing unit is matched with the SDRAM with larger capacity to perform online training of models and data;
the FPGA programmable logic device realizes a customized AI acceleration processing function and module, and improves the calculation efficiency;
the high-capacity Flash or SSD memory is provided with a pre-trained machine learning model and an instance database;
the universal PCIe/M.2 interface supports updating and replacing an AI processing module and an algorithm at any time;
the low-power design is suitable for a satellite miniaturized operation environment;
the API is used for integrating a third-party AI algorithm framework and a model;
the bit manipulation and communication interface enables data transmission and result output.
Further improved, the weather forecast and disaster early warning provides accurate data support including aerosol sensing, and specifically includes the following steps:
a miniature aerosol sensor that releases a specific component in an atmosphere within a cloud target area;
the micro aerosol sensor capsule is spread along with wind, and the cloud form change information of the passing region is perceived and recorded;
when the number of the sensor capsules reaches a threshold value, a temporary climate sensing network is formed through spontaneous polymerization;
the sensing network completes real-time summarizing and transmitting of cloud information in the area through gas signal interaction;
the weather prediction mechanism collects signals in the atmosphere through specific equipment and decodes the signals into a prediction result;
thus, the instant distributed application in the physical space is realized, and a new supporting channel is provided for decision making.
The idea that the micro sensing capsule spreading along with the wind forms a temporary physical network is utilized, so that the communication limit can be broken through, and timely and accurate data support can be provided for a specific area.
Further improved, the weather forecast and disaster warning provide accurate data support for calculating and storing cloud cluster prediction results by using a distributed computing network based on a block chain, and specifically comprise the following steps:
each satellite/unmanned aerial vehicle node participating in calculation and storage is used as a block in a calculation network, and an encryption technology is used for generating a unique ID;
the calculation results are not collected to the central server, but written into each participating node at the same time, and a tamper-proof result chain is formed through a consensus mechanism;
the meteorological and disaster early warning mechanism is directly connected to the distributed network, and the latest snapshot state of the prediction result is obtained in real time;
cloud images and motion data are shared and uploaded to a network in a decentralization mode, so that the safety and auditability of the data are ensured;
through an intelligent contract mechanism, the calculation result can directly trigger the release of the early warning event, and the decision response time is shortened.
The efficient distributed computing architecture based on the block chain and the decentralised computing power network can furthest improve prediction precision and early warning response capability, monitor and trust data and results, and provide a brand-new technical support for weather and public safety.
Specifically, the weather forecast and disaster early warning provide accurate data support to transmit calculation results in real time by utilizing a wired network of the unmanned aerial vehicle and a ground base station, and specifically comprise the following steps:
arranging a plurality of ground base stations and a low-altitude spiral unmanned aerial vehicle network in a cloud picture monitoring area;
each base station and a plurality of unmanned aerial vehicles nearby form a small-sized wired communication network;
each unmanned plane is provided with a calculation and sensing module which participates in predictive calculation and information exchange;
after the calculation is finished, the result is immediately uploaded to a base station through a wired channel and then is sent to an emergency command center;
meanwhile, the wired network between the bases transmits the original data in real time for calculation reference;
by increasing the coverage area of the base station, the low-altitude affinity network can rapidly reach large-scale coverage, and the early warning aging requirement is met.
Compared with the satellite or wireless communication, the method is realized on the basis of the prior art, a new transmission channel is provided, prediction accuracy and response efficiency are hopefully improved, and necessary support is provided for decision making.
In summary, the residence time of the cloud cluster is accurately predicted in real time, and more accurate data support is provided for weather forecast and disaster early warning; the cloud image information is obtained by satellite remote sensing, and the cloud cluster is tracked, predicted and calculated by an artificial intelligence technology, so that the calculation result can be rapidly released in short time, and the real-time requirements of weather forecast and disaster prevention early warning are met; the artificial intelligence algorithm is used for learning and training a large amount of historical cloud image data, cloud cluster motion mode characteristics are extracted, prediction accuracy can be improved, and the accuracy of residence time calculation is improved by comparing and correcting actual observation data with a prediction result; the satellite remote sensing technology can realize the cloud image monitoring of a wide monitoring area, and the prediction calculation can cover a wide area as well, so that decision support is provided for regional weather forecast and regional disaster early warning; single or multiple cloud clusters can be tracked and predicted independently, and personalized residence time calculation results are given according to the cloud cluster characteristics of different types, sizes and the like, so that the requirements of different application fineness are met; the method has the advantages that not only is the residence time predicted, but also more related cloud cluster information such as movement trend, change rule and the like can be predicted by the reserved space of the method, and the prediction and judgment capability is improved; along with the continuous iterative upgrade of the artificial intelligent algorithm and the sample data, the prediction model and the result can be improved in real time, and the prediction early warning level is kept.
While the fundamental and principal features of the invention and advantages of the invention have been shown and described, it will be apparent to those skilled in the art that the invention is not limited to the details of the foregoing exemplary embodiments, but may be embodied in other specific forms without departing from the spirit or essential characteristics thereof. The present embodiments are, therefore, to be considered in all respects as illustrative and not restrictive, the scope of the invention being indicated by the appended claims rather than by the foregoing description, and all changes which come within the meaning and range of equivalency of the claims are therefore intended to be embraced therein. Any reference sign in a claim should not be construed as limiting the claim concerned.
Furthermore, it should be understood that although the present disclosure describes embodiments, not every embodiment is provided with a separate embodiment, and that this description is provided for clarity only, and that the disclosure is not limited to the embodiments described in detail below, and that the embodiments described in the examples may be combined as appropriate to form other embodiments that will be apparent to those skilled in the art.
Claims (10)
1. The method for predicting cloud residence time in real time with high precision is characterized by comprising the following steps:
s1, acquiring position and motion information of a cloud cluster through a satellite cloud image remote sensing technology;
s2, predicting the motion trail of the cloud cluster by using an artificial intelligent algorithm;
s3, calculating the residence time of the cloud cluster according to the prediction result and the actual observation data;
and S4, publishing the calculation result in real time.
2. The method for predicting cloud residence time in real time and high precision according to claim 1, wherein the method comprises the following steps: in the step S1, the position and motion information of the cloud cluster is obtained, and the cloud cluster further comprises a multiband camera unmanned plane, a distributed laser radar network, a multi-parameter weather sensor, a millimeter wave Yun Pu instrument, a cloud image intelligent analysis algorithm and a cloud cluster simulation system;
capturing cloud cluster images by utilizing a multiband camera unmanned aerial vehicle, and acquiring cloud cluster boundary and motion vector information by an image analysis technology;
the distributed laser radar network is used for reconstructing a three-dimensional cloud cluster structure in real time by analyzing the reflected signals of airborne particles, and capturing the movement and deformation of the cloud cluster;
using a multi-parameter meteorological sensor, using a balloon or an unmanned aerial vehicle to carry out three-dimensional arrangement, obtaining parameters of temperature, humidity and pressure in the cloud cluster, and presuming the movement of the cloud cluster through change;
by utilizing a millimeter wave Yun Pu instrument, the motion state and the evolution process of the cloud cluster are reversely deduced by measuring the millimeter wave absorption characteristics of the cloud cluster at different positions;
developing an intelligent cloud image analysis algorithm, and performing image processing on cloud video shot by a ground monitoring network to obtain cloud movement information;
and establishing a cloud cluster simulation system with a turbulence model and a cloud micro-physical model coupled, inputting meteorological elements, and simulating cloud cluster evolution and movement processes.
3. The method for predicting cloud residence time in real time and high precision according to claim 1, wherein the method comprises the following steps: in step S1, the cloud cluster position and motion information obtained by satellite cloud image remote sensing specifically includes the following steps:
a satellite cloud image sensor with double wave bands is used, one wave band is in a visible light range, and the other wave band is in a middle infrared range, and a visible light image and an infrared enhanced image of a cloud cluster are obtained at the same time;
a brand new cloud cluster boundary segmentation algorithm is applied, a convolutional neural network is used for analyzing cloud cluster texture characteristics, and cloud cluster boundaries are accurately extracted;
analyzing the change of the cloud cluster in the continuous phase images by utilizing a spatial relationship graph network, and determining the motion trail of the cloud cluster key points;
establishing an optical flow deep learning model, inputting a two-phase cloud picture, and outputting a motion vector field of each pixel in the cloud cluster;
designing a cloud cluster motion predictor, and based on an LSTM network, synthesizing a cloud cluster boundary, a key point track and a motion vector field to predict a cloud cluster state at a future moment;
using cloud computing and a distributed computing framework to realize rapid processing of mass satellite cloud image data and give out real-time cloud positioning and movement results;
constructing a cloud-end cooperative cloud image processing system, compressing and preprocessing satellite data by a cloud end, and finishing key algorithm calculation by side end equipment;
the system has the functions of autonomous learning and model enhancement, and the algorithm is continuously optimized during satellite operation, so that the extraction precision is improved.
4. The method for predicting cloud residence time in real time and high precision according to claim 1, wherein the method comprises the following steps: in step S2, the method for predicting the cloud cluster motion trail by using the artificial intelligence algorithm specifically includes the following steps:
using a convolution LSTM network as a prediction model to simulate a space-time relationship in the cloud cluster movement process;
the network input comprises a historical multi-time phase cloud chart and meteorological field data, wherein the meteorological field data comprises a wind field and a pressure field, and a cloud cluster motion law is learned;
the network structure adopts a coding-decoding architecture, a coder extracts cloud group motion related characteristics, and a decoder completes future motion prediction;
an attention module is led into the decoder to pay attention to different parts of input data, so that self-adaptive learning is realized;
the network output comprises cloud cluster boundary box sequences of N time periods in the future, and the cloud cluster boundary box sequences represent predicted cloud cluster positions;
li Yongmeng Teslaloy random sampling method enhances data and improves model generalization capability;
designing a genetic algorithm-based super-parameter optimization strategy, and searching out an optimal network structure;
an incremental learning algorithm is applied, and a model is updated by using newly acquired data, so that the prediction accuracy is continuously improved;
and model parallel training is realized on the cloud platform by utilizing multiple GPUs, and the operation process is accelerated.
5. The method for predicting cloud residence time in real time and high precision according to claim 1, wherein the method comprises the following steps: in step S3, the method steps for calculating the residence time of the cloud cluster according to the prediction result and the actual observation data include the following steps:
based on radar data and weather models: combining radar echo data and a meteorological model, predicting a movement track of the cloud cluster by utilizing the intensity, speed and direction information of the radar echo, and calculating the residence time of the cloud cluster by comparing the movement track with actual observation data;
based on a weather database and a machine learning algorithm: utilizing a large number of weather databases, combining a machine learning algorithm, establishing a correlation model between cloud cluster motion and residence time, and outputting the residence time of the cloud cluster by inputting a prediction result and actual observation data;
based on satellite cloud image texture characteristics and a deep learning algorithm: and predicting the movement track of the cloud cluster by extracting the texture features of the cloud cluster in the satellite cloud picture and calculating the residence time of the cloud cluster by utilizing a deep learning algorithm and combining actual observation data.
6. The method for predicting cloud residence time in real time and high precision according to claim 1, wherein the method comprises the following steps: in step S4, the specific step method for publishing the calculation result in real time includes the following steps:
the unmanned aerial vehicle is used for carrying and deploying a small satellite network, a plurality of microsatellites are deployed at important observation points and prediction points, and the microsatellites are carried with machine learning chips with strong calculation processing capacity, receive cloud image data in real time and perform prediction calculation;
after the calculation is completed, each micro satellite directly transmits the result back to the ground control center through a satellite communication network, and meanwhile, the micro satellite can directly communicate with a nearby unmanned aerial vehicle through a beacon;
the control center pushes the calculation result to emergency centers and user terminals in various places in real time, and the unmanned aerial vehicle nearby directly receives the calculation result and propagates the result to a farther place by utilizing the maneuverability of the unmanned aerial vehicle nearby.
7. The method for predicting cloud residence time in real time and high precision according to claim 6, wherein the method comprises the following steps: the machine learning chip comprises, but is not limited to, an AI acceleration processor, an ARM high-performance CPU, an FPGA programmable logic device, a high-capacity Flash or SSD memory, a general PCIe/M.2 interface, a low power consumption, an API, a bit operation and communication interface;
the AI acceleration processor comprises an NVIDIA Jetson series singlechip and is used for preprocessing cloud picture data, deep learning feature extraction and model prediction calculation;
the ARM high-performance central processing unit is matched with the SDRAM with larger capacity to perform online training of models and data;
the FPGA programmable logic device realizes a customized AI acceleration processing function and module, and improves the calculation efficiency;
the high-capacity Flash or SSD memory is provided with a pre-trained machine learning model and an instance database;
the universal PCIe/M.2 interface supports updating and replacing an AI processing module and an algorithm at any time;
the low-power design is suitable for a satellite miniaturized operation environment;
the API is used for integrating a third-party AI algorithm framework and a model;
the bit manipulation and communication interface enables data transmission and result output.
8. The method for predicting cloud residence time in real time and high precision according to claim 1, wherein the method comprises the following steps: the weather forecast and disaster early warning provides accurate data support including aerosol sensing, and specifically includes the following steps:
a miniature aerosol sensor that releases a specific component in an atmosphere within a cloud target area;
the micro aerosol sensor capsule is spread along with wind, and the cloud form change information of the passing region is perceived and recorded;
when the number of the sensor capsules reaches a threshold value, a temporary climate sensing network is formed through spontaneous polymerization;
the sensing network completes real-time summarizing and transmitting of cloud information in the area through gas signal interaction;
the weather prediction mechanism collects signals in the atmosphere through specific equipment and decodes the signals into a prediction result;
thus, the instant distributed application in the physical space is realized, and a new supporting channel is provided for decision making.
9. The method for predicting cloud residence time in real time and high precision according to claim 8, wherein the method comprises the following steps: the weather forecast and disaster early warning provides accurate data support to calculate and save cloud cluster prediction results by using a distributed computing network based on a block chain, and specifically comprises the following steps:
each satellite/unmanned aerial vehicle node participating in calculation and storage is used as a block in a calculation network, and an encryption technology is used for generating a unique ID;
the calculation results are not collected to the central server, but written into each participating node at the same time, and a tamper-proof result chain is formed through a consensus mechanism;
the meteorological and disaster early warning mechanism is directly connected to the distributed network, and the latest snapshot state of the prediction result is obtained in real time;
cloud images and motion data are shared and uploaded to a network in a decentralization mode, so that the safety and auditability of the data are ensured;
and the calculation result directly triggers the release of the early warning event through an intelligent contract mechanism, so that the decision response time is shortened.
10. The method for predicting cloud residence time in real time and high precision according to claim 9, wherein the method comprises the following steps: the weather forecast and disaster early warning provide accurate data support to utilize a wired network of the unmanned aerial vehicle and a ground base station to transmit calculation results in real time, and specifically comprise the following steps:
arranging a plurality of ground base stations and a low-altitude spiral unmanned aerial vehicle network in a cloud picture monitoring area;
each base station and a plurality of unmanned aerial vehicles nearby form a small-sized wired communication network;
each unmanned plane is provided with a calculation and sensing module which participates in predictive calculation and information exchange;
after the calculation is finished, the result is immediately uploaded to a base station through a wired channel and then is sent to an emergency command center;
meanwhile, the wired network between the bases transmits the original data in real time for calculation reference;
by increasing the coverage area of the base station, the low-altitude affinity network can rapidly reach large-scale coverage, and the early warning aging requirement is met.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202311656804.2A CN117369026B (en) | 2023-12-06 | 2023-12-06 | Real-time high-precision cloud cluster residence time prediction method |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202311656804.2A CN117369026B (en) | 2023-12-06 | 2023-12-06 | Real-time high-precision cloud cluster residence time prediction method |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN117369026A true CN117369026A (en) | 2024-01-09 |
| CN117369026B CN117369026B (en) | 2024-03-08 |
Family
ID=89404502
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202311656804.2A Active CN117369026B (en) | 2023-12-06 | 2023-12-06 | Real-time high-precision cloud cluster residence time prediction method |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN117369026B (en) |
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN117849907A (en) * | 2024-03-07 | 2024-04-09 | 江苏省气象台 | Meteorological disaster targeted early warning method and system based on multi-source data |
Citations (8)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| WO2011129473A1 (en) * | 2010-04-16 | 2011-10-20 | (주)뉴멀티테크 | Automatic sky state observation system and method |
| US20120072579A1 (en) * | 2010-09-17 | 2012-03-22 | Microsoft Corporation | Monitoring cloud-runtime operations |
| CN103513295A (en) * | 2013-09-25 | 2014-01-15 | 青海中控太阳能发电有限公司 | Weather monitoring system and method based on multi-camera real-time shoot and image processing |
| CN111754539A (en) * | 2020-06-23 | 2020-10-09 | 上海电力大学 | Full-automatic cloud picture acquisition device and cloud cluster motion trajectory prediction method thereof |
| WO2021064524A1 (en) * | 2019-10-04 | 2021-04-08 | International Business Machines Corporation | Predicting weather radar images |
| CN113936041A (en) * | 2021-10-27 | 2022-01-14 | 浙江科技学院 | Cloud cluster tracking and motion trend prediction method for photovoltaic power generation |
| CN115170619A (en) * | 2022-06-10 | 2022-10-11 | 山东电力建设第三工程有限公司 | Cloud occlusion prediction method based on dense optical flow method |
| CN115546657A (en) * | 2022-09-20 | 2022-12-30 | 湖南防灾科技有限公司 | Regional earth surface radiation intensity prediction method based on cloud cluster motion trail |
-
2023
- 2023-12-06 CN CN202311656804.2A patent/CN117369026B/en active Active
Patent Citations (8)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| WO2011129473A1 (en) * | 2010-04-16 | 2011-10-20 | (주)뉴멀티테크 | Automatic sky state observation system and method |
| US20120072579A1 (en) * | 2010-09-17 | 2012-03-22 | Microsoft Corporation | Monitoring cloud-runtime operations |
| CN103513295A (en) * | 2013-09-25 | 2014-01-15 | 青海中控太阳能发电有限公司 | Weather monitoring system and method based on multi-camera real-time shoot and image processing |
| WO2021064524A1 (en) * | 2019-10-04 | 2021-04-08 | International Business Machines Corporation | Predicting weather radar images |
| CN111754539A (en) * | 2020-06-23 | 2020-10-09 | 上海电力大学 | Full-automatic cloud picture acquisition device and cloud cluster motion trajectory prediction method thereof |
| CN113936041A (en) * | 2021-10-27 | 2022-01-14 | 浙江科技学院 | Cloud cluster tracking and motion trend prediction method for photovoltaic power generation |
| CN115170619A (en) * | 2022-06-10 | 2022-10-11 | 山东电力建设第三工程有限公司 | Cloud occlusion prediction method based on dense optical flow method |
| CN115546657A (en) * | 2022-09-20 | 2022-12-30 | 湖南防灾科技有限公司 | Regional earth surface radiation intensity prediction method based on cloud cluster motion trail |
Cited By (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN117849907A (en) * | 2024-03-07 | 2024-04-09 | 江苏省气象台 | Meteorological disaster targeted early warning method and system based on multi-source data |
| CN117849907B (en) * | 2024-03-07 | 2024-05-24 | 江苏省气象台 | Meteorological disaster targeted early warning method and system based on multi-source data |
Also Published As
| Publication number | Publication date |
|---|---|
| CN117369026B (en) | 2024-03-08 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| Lin et al. | Exploiting spatiotemporal patterns for accurate air quality forecasting using deep learning | |
| US20200272899A1 (en) | Systems and Methods for Deploying and Updating Neural Networks at the Edge of a Network | |
| CN117369026B (en) | Real-time high-precision cloud cluster residence time prediction method | |
| WO2022261675A1 (en) | Systems and methods for generating 3d models from drone imaging | |
| CN111033318A (en) | Local weather prediction | |
| CN110414732B (en) | Travel future trajectory prediction method and device, storage medium and electronic equipment | |
| CN112418245B (en) | Electromagnetic emission point positioning method based on urban environment physical model | |
| CN112256589B (en) | Simulation model training method and point cloud data generation method and device | |
| CN112307884A (en) | Forest fire spreading prediction method based on continuous time sequence remote sensing situation data and electronic equipment | |
| CN110189304A (en) | Remote sensing image target on-line quick detection method based on artificial intelligence | |
| CN111339826B (en) | Landslide unmanned aerial vehicle linear sensor network frame detecting system | |
| JP7054255B2 (en) | Virtual intelligence and optimization with multi-source, real-time, context-aware real-world data | |
| CN110221360A (en) | A kind of power circuit thunderstorm method for early warning and system | |
| Kim et al. | DeepTC: ConvLSTM network for trajectory prediction of tropical cyclone using spatiotemporal atmospheric simulation data | |
| CN116192251A (en) | Method for determining satellite-ground laser link communication availability based on sky infrared remote sensing image | |
| CN116310898A (en) | Forest fire spread prediction method and system based on neural network and Huygens principle | |
| US11115468B2 (en) | Live management of real world via a persistent virtual world system | |
| CN116699731B (en) | Tropical cyclone path short-term forecasting method, system and storage medium | |
| Wang et al. | Hybrid model for prediction of carbon monoxide and fine particulate matter concentrations near a road intersection | |
| CN116581756B (en) | Wind power prediction method, model training method, device, equipment and medium | |
| CN116070676B (en) | Expressway road surface temperature forecasting method based on attention mechanism and self-encoder | |
| CN116824372B (en) | Urban rainfall prediction method based on Transformer | |
| CN116324523A (en) | Turbulence prediction system and turbulence prediction method | |
| CN102364488B (en) | Synthetic environment simulation platform | |
| Yuan et al. | Spacecraft components detection based on a lightweight YOLOv3 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 | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |