CN108197012A - Remotely-sensed data distributed treatment cluster scale and performance measure method - Google Patents
Remotely-sensed data distributed treatment cluster scale and performance measure method Download PDFInfo
- Publication number
- CN108197012A CN108197012A CN201711413516.9A CN201711413516A CN108197012A CN 108197012 A CN108197012 A CN 108197012A CN 201711413516 A CN201711413516 A CN 201711413516A CN 108197012 A CN108197012 A CN 108197012A
- Authority
- CN
- China
- Prior art keywords
- cluster
- model
- processing node
- resource consumption
- scale
- 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
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F11/00—Error detection; Error correction; Monitoring
- G06F11/30—Monitoring
- G06F11/34—Recording or statistical evaluation of computer activity, e.g. of down time, of input/output operation ; Recording or statistical evaluation of user activity, e.g. usability assessment
- G06F11/3447—Performance evaluation by modeling
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F11/00—Error detection; Error correction; Monitoring
- G06F11/30—Monitoring
- G06F11/34—Recording or statistical evaluation of computer activity, e.g. of down time, of input/output operation ; Recording or statistical evaluation of user activity, e.g. usability assessment
- G06F11/3452—Performance evaluation by statistical analysis
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F11/00—Error detection; Error correction; Monitoring
- G06F11/30—Monitoring
- G06F11/34—Recording or statistical evaluation of computer activity, e.g. of down time, of input/output operation ; Recording or statistical evaluation of user activity, e.g. usability assessment
- G06F11/3466—Performance evaluation by tracing or monitoring
- G06F11/3476—Data logging
Landscapes
- Engineering & Computer Science (AREA)
- Theoretical Computer Science (AREA)
- General Engineering & Computer Science (AREA)
- Physics & Mathematics (AREA)
- Computer Hardware Design (AREA)
- General Physics & Mathematics (AREA)
- Quality & Reliability (AREA)
- Bioinformatics & Cheminformatics (AREA)
- Evolutionary Biology (AREA)
- Bioinformatics & Computational Biology (AREA)
- Life Sciences & Earth Sciences (AREA)
- Probability & Statistics with Applications (AREA)
- Management, Administration, Business Operations System, And Electronic Commerce (AREA)
Abstract
The present invention proposes a kind of remotely-sensed data distributed treatment cluster scale and performance measure method, including:Obtain the configuration information of single processing node in cluster;Establish the ability Rating Model of processing node;It establishes cluster resource consumption and increases S curve model;It is modified by successive ignition, obtains cluster resource consumption best fit Growth Curve Model, and acquire the parameter in the model;The cluster resource that the setting period is calculated according to model consumes discreet value, and the deviation with surveying consumed resource is calculated according to Bai Saier deviation evaluations model;By the resource consumption prediction period input model of cluster processing node, computing resource consumes discreet value and is maximized with the sum of deviation, in ability Rating Model as maximum consumption resource input processing node, the configuration information needed for the single processing node of cluster and cluster scale number is calculated.The present invention effectively improves remotely-sensed data distributed treatment cluster scale and performance prediction and assessment.
Description
Technical field
The invention belongs to the fields of remote sensing satellite distributed proccessing, are related to a kind of remotely-sensed data distributed treatment cluster
Scale and performance measure method.
Background technology
The Remote Sensing Data Processing system in China is started late, and existing many Ground Processing Systems use the meter of centralized cluster
Calculation mode stores and processs image data, and the setting of cluster scale and performance often regards ground system construction scale, root
It is set according to artificial previous experience, subjectivity is very strong, and the unified standard of neither one and model, causes existing all kinds of satellites
The phenomenon that data floor treatment cluster resource utilization rate remains high or resource is largely left unused simultaneously is deposited, and further visits company-data
It asks that efficiency and processing speed decline, therefore establishes the model and standard of a unified ground data processing cluster building, carry
The cost performance that high disposal cluster utilizes is imperative.
Satellite remote sensing date distributed treatment cluster scale is set, the data volume that main with good grounds daily satellite passes down at present
The experience of estimated value or the cluster building of existing ground satellite data processing system, carry out satellite data processing cluster scale and
Performance setting.The data of the satellites such as series are painted according to environment series of satellites, the day of the daily down-transmitting data amount of satellite, such as China
Handle cluster building mode;Cluster, such as wind and cloud series of satellites, ocean series of satellites are built according to existing ground system.In order to
Realize that accurately distributed treatment cluster is built, particularly ensure satellite data processing cluster is stable, under reliable requirement,
Rational cluster scale is provided and performance setting is more and more necessary, whether improves the economic cost performance of data processing cluster, also
It is to have strong demand in reduction ground surface works scale.
Traditional Remote Sensing Data Processing cluster building is estimated according to prior model or according to daily processing satellite data amount
It calculates and predicts, the artificial experience estimation of prior model is not accurate enough, with other influence prior models such as the iteration updates of IT performances
Factor change, differ and be surely updated in time, causing to build distributed treatment cluster scale and performance has centainly
Loss;There is gross data to support but without fixed model according to daily processing satellite data amount estimation cluster scale and performance
With reference to.First, satellite data amount is too relied on, does not fully consider the uncontrollable factor generated in processing procedure, such as certain
The intermediate result data that a processing module generates, does not bring entire process flow into time, thus there is additional data production
It is raw, calculating distributed treatment cluster models is caused relatively large deviation occur;In addition, such algorithm lacks strict theoretical foundation conduct
Guidance, can not also eliminate theoretic deviation and convincingness;Rise and fall violent situation finally, for satellite data, can not ensure
The stability and reliability of entire distributed treatment cluster scale.
Therefore, how to establish accurately cluster scale to set with performance, using individual node configuration and performance, with reference to existing
Floor treatment cluster resource consumption calculates Gong's platinum thatch S curve model of growth of cluster resource consumption, according to computation model reality
Existing project operation process of refinement cluster and performance setting are the problems to be solved of the present invention.
Invention content
Problems to be solved by the invention are how to establish accurately cluster scale to set with performance, are matched using individual node
It puts and performance, with reference to existing floor treatment cluster resource consumption, to ensure the stability of entire distributed treatment cluster scale
And reliability.
Technological means for solving subject is:
The present invention proposes a kind of remotely-sensed data distributed treatment cluster scale and performance measure method, includes the following steps:
It builds remotely-sensed data floor treatment cluster and obtains the configuration information of single processing node in cluster;
According to the configuration parameter of single processing node, shared weight is normalized in the cluster, and by hundred-mark system pair
It handles node and carries out dynamic grading, establish the ability Rating Model of each processing node;
Node resource consumption parameter substitution within setting period and period will be handled in the cluster of acquisition and establishes processing section
The ability Rating Model of point carries out scale, increases S curve model, and calculate in the model to establish cluster resource consumption
Curve coefficients;
According to cluster resource consumption of the processing node in several periods, disappear by successive ignition to the cluster resource of foundation
Consumption increases S curve model and is modified, and obtains cluster resource consumption best fit Growth Curve Model, and acquire in the model
Parameter;
The cluster resource of setting period is calculated according to gained cluster resource consumption best fit Growth Curve Model to disappear
Discreet value is consumed, calculate cluster resource consumption discreet value according to Bai Saier deviation evaluation models surveys with cluster in the setting period
The deviation of consumed resource;
The resource consumption prediction period of cluster processing node is inputted into cluster resource consumption best fit Growth Curve Model,
Cluster resource consumption discreet value of the processing node in prediction period is calculated and is maximized with the sum of deviation, as most
In the ability Rating Model of big consumption resource input processing node, the configuration information needed for the single processing node of cluster is calculated
With cluster scale number.
A preferred technical solution of the present invention is further used as, the single processing node configuration obtained in the method
Information includes:The cpu busy percentage of node and dominant frequency, exchange storage region, hard at memory size and IO performances, network rate information
Disk IO information.
Further as a preferred technical solution of the present invention, according to Bai Saier deviation evaluation models in the method
It calculates deviation and uses equation below:
Wherein, m is deviation;V is.
Further as a preferred technical solution of the present invention, the method further includes single to cluster is calculated
Configuration information and cluster scale needed for processing node judge whether the processing capacity for meeting cluster.
Invention effect:
The remotely-sensed data distributed treatment cluster scale of the present invention and performance measure method, for remote sensing satellite data processing
The scale and performance setting of cluster, a kind of cluster resource consumption for being used for production daily according to existing floor treatment cluster of design
Gong's platinum thatch S curve resource consumption model of growth is built, for this method by establishing every 3 days as a stage, 30 days are a week
Phase resource consumption determines that its cluster resource consumes model of growth, and its model bias is corrected according to bessel formula.
Distributed type assemblies stock number consumption growth S curve model method of establishing in inventive algorithm is then to utilize single section
The scoring restriction relation of point Performance Score and clustering performance, based on Gompertz curve models and Bai Saier deviation evaluation models,
Individual node and cluster scale and the conversion relation of performance are established, and then realizes the estimation to distributed treatment cluster scale;Together
The cluster resource consumption of 3 natural gift busy and idle time different periods of Shi Liyong is a period, is changed for a cycle within 30 days
Generation generation cluster resource consumption best fit Growth Curve Model, using Bai Saier deviation evaluation models, to predicted value and observation
Value carries out deviation evaluation, effectively improves remotely-sensed data distributed treatment cluster scale and performance prediction and assessment.The present invention calculates
Method breaches the limitation that traditional prior model estimates cluster with the processing data amount that follows up, to provide highly reliable, high performance-price ratio
It handles cluster and more tight and ideal technical solution is provided.
Description of the drawings
Fig. 1 is remotely-sensed data distributed treatment cluster scale and the flow diagram of performance measure method in the present invention.
Specific embodiment
Hereinafter, it is described in detail based on attached drawing for the present invention.
As shown in Figure 1, the present invention proposes a kind of remotely-sensed data distributed treatment cluster scale and performance measure method, we
Method specifically comprises the following steps:
1. building remotely-sensed data floor treatment cluster, and cluster is initialized, start management node and processing node.
2. obtain the configuration information that node is handled in the cluster, it preferably includes cpu busy percentage and dominant frequency, memory size with
IO performances, exchange the parameter informations such as storage region, hard disk IO information at network rate information.
3. the weights according to the dominant frequency processing capacity m of single CUP are 5, wherein 1.2GHz<m<4.2GHz;Memory size n's
Weights are 2, wherein 1GB<n<1TB;The weights of network transmission speed s are 2, wherein 100Mb<s<10Gb;The power of hard disk I/O rate y
It is 3 to be worth, wherein reading 3000Mb<y1<5000Mb writes 1000Mb<y2<3000Mb;The weights for exchanging slow c are 3, wherein 256MB<c<
2GB;Based on the auxiliary datas such as above-mentioned, after weighting normalization processing, 100 are multiplied by, establishes the single processing node capacity scoring of cluster
Model.
4. obtaining in the processing cluster in the resource consumption parameter of setting period, the resource in three day busy period is such as obtained
The parameters such as consumption, period distribution situation.
5. step 4 is obtained in cluster at the single place of cluster that the resource consumption parameter substitution step 3 of setting period is established
Node capacity Rating Model is managed, can be consumed with the cluster resource of establishment stage and increase S curve model, and calculate the song in the model
Linear system number.The present embodiment was a setting period with three days, and 30 days are a cycle, are disappeared according to the maximum resource of daily busy
Consumption calculates cluster resource consumption and increases S curve model inner curve correlation unknowm coefficient, and determines that the stage curve increases
Model.
And S curve model is increased according to determining cluster resource consumption, next stage forecast consumed resource is recorded,
By the busy period in several stages to cluster resource consumption, successive ignition brings cluster resource consumption into and increases the progress of S curve model
It corrects, regenerates to obtain cluster resource consumption best fit Growth Curve Model, it is higher to be configured to a new degree of fitting
Cluster resource consumes best fit Growth Curve Model, and acquires the parameter in the model.In the present embodiment, according to multiple periods
Busy period to cluster resource consumption, cluster resource is consumed by a cycle successive ignition and increases S curve model and carries out
It corrects, the cluster resource consumption best fit Growth Curve Model wirelessly approached, and the curve model acquired at this time is related
Parameter.
The detailed process is as follows:
(1) characteristic of Gong's platinum thatch curve of analysis cluster resource consumption, including:Curve initial stage increasess slowly, later gradually
Accelerate, after reaching a certain level, growth rate is gradually reduced again, finally close to a horizontal line.
(2) determine that the general type that the cluster resource consumption increases S curve model is:
Wherein, K, a, b are unknown constant, and t is the time constant to be predicted;And K>0,0<A ≠ 1,0<B ≠ 1,.
There is asymptote at curve model both ends, and upper asymptote is Y=K, and lower asymptote is Y=0.
(3) formula one is rewritten as logarithmic form using Prediction by Modified Index Curve and its feature:
(4) constant that curve is solved using fair curve determines method, if the local summation of three of observed value is respectively S1,
S2, S3:
(5) it is acquired according to three and method and the lg a, lg K, b is obtained:
(6) its antilogarithm is taken to acquire a and K values according to gained lg a, lg K;
(7) it iterates and establishes cluster resource consumption best fit growth S curve model M odel;
(8) 30 days cluster resources are calculated using gained cluster resource consumption best fit growth S curve model M odel to disappear
The discreet value of consumption.
6. judge whether the resource consumption model for meeting the cluster.Otherwise the observation of a cycle is carried out again, and is returned
4th step if meeting optimum fit curve model, continues to handle as follows.
7. the setting period is set as 30 days, increase S curve model M odel using gained cluster resource consumption best fit
The discreet value for setting the period as the cluster resource consumption of 30 days is calculated, and obtains and sets the period as the cluster actual measurement resource of 30 days
Consumption calculates cluster resource consumption discreet value according to Bai Saier deviation evaluation models and is provided with cluster actual measurement in the setting period
The deviation of source consumption;It is calculated according to Bai Saier deviation evaluations model formation:
8. wherein V in formula seven --- most or value and observation difference, take the difference of arithmetic mean of instantaneous value and observation, that is, haveL is observation, LiObservation for ith.
9. being computed according to step 7 and 8, m deviations can be obtained as ± 27 stock numbers of m=, arbitrarily observation obtains most twice
Big error is mMost=± 50 stock numbers.
10. handling node hardware configuration information according to cluster, a resource consumption prediction period is inputted, cluster is calculated and exists
Institute's maximum consumption stock number in specified predicted time, while provide specific number of nodes and cluster performance indicator;It specifically includes:
(1) resource consumption prediction period to the step 5 gained cluster resource of input cluster processing node consumes best fit
Resource consumption discreet value of the cluster processing node in prediction period is calculated in Growth Curve Model;
(2) the resource consumption discreet value that cluster processing node is calculated in prediction period is obtained with step 9
The sum of deviation be maximized.
(3) it is handled the maximum value as 3 gained of maximum consumption resource input step in the ability Rating Model of node,
The configuration information needed for the single processing node of cluster is calculated, while the specific scale of cluster can be obtained according to node configuration information
Number and performance indicator.
11. configuration information, the specific scale number of cluster and performance that judgment step 10 exports to obtain needed for processing node refer to
Mark whether meet demand, if meet if continue to execute, otherwise skip to the 4th step and continue to execute;
The configuration information of the final output cluster and specific performance information, end processing.
To sum up, the present invention is based on Gompertz (Gong's platinum thatch) Growth Curve Model, according to a stage, 30 days every three days
The cluster resource consumption of a cycle establishes cluster resource consumption Growth Curve Model, and utilize Bai Saier by iteration
Drift correction Evaluation model acquires deviation range to resource consumption estimation amount and observation, according to single resource estimation model, pushes away
Rational cluster scale and configuration are recommended, avoids traditional Remote Sensing Data Processing cluster neither one configuration standard, relies on experience
Model buildings, cluster utilization rate is not high, the wasting of resources, helps to standardize, is engineered structure remote sensing ground data processing cluster
Scale and configuration.
It should be noted that described above is only the preferred embodiment of the present invention, it should be understood that for art technology
For personnel, several changes and improvements can also be made under the premise of the technology of the present invention design is not departed from, these are included in
In protection scope of the present invention.
Claims (4)
1. remotely-sensed data distributed treatment cluster scale and performance measure method, which is characterized in that include the following steps:
It builds remotely-sensed data floor treatment cluster and obtains the configuration information of single processing node in cluster;
According to the configuration parameter of single processing node, shared weight is normalized in the cluster, and by hundred-mark system to processing
Node carries out dynamic grading, establishes the ability Rating Model of each processing node;
Node resource consumption parameter substitution within setting period and period will be handled in the cluster of acquisition and establishes processing node
Ability Rating Model carries out scale, increases S curve model, and calculate the curve in the model to establish cluster resource consumption
Coefficient;
According to cluster resource consumption of the processing node in several periods, the cluster resource of foundation is consumed by successive ignition and is increased
Long S curve model is modified, and obtains cluster resource consumption best fit Growth Curve Model, and acquire the ginseng in the model
Number;
Best fit Growth Curve Model is consumed according to gained cluster resource, the cluster resource consumption of setting period is calculated in advance
Valuation calculates cluster resource consumption discreet value according to Bai Saier deviation evaluation models and surveys resource with cluster in the setting period
The deviation of consumption;
The resource consumption prediction period of cluster processing node is inputted into cluster resource consumption best fit Growth Curve Model, is calculated
It obtains cluster resource consumption discreet value of the processing node in prediction period and is maximized with the sum of deviation, disappeared as maximum
In the ability Rating Model of cost source input processing node, the configuration information sum aggregate needed for the single processing node of cluster is calculated
Group's scale number.
2. remotely-sensed data distributed treatment cluster scale and performance measure method according to claim 1, which is characterized in that institute
The single processing node configuration information obtained in method is stated to include:The cpu busy percentage of node and dominant frequency, memory size and IO
Energy, exchanges storage region, hard disk IO information at network rate information.
3. remotely-sensed data distributed treatment cluster scale and performance measure method according to claim 1, which is characterized in that institute
It states in method and deviation is calculated using equation below according to Bai Saier deviation evaluations model:
Wherein, m is deviation;V is most or the difference of value and observation, n are observation frequency.
4. remotely-sensed data distributed treatment cluster scale and performance measure method according to claim 1, which is characterized in that institute
The method of stating is further included to be judged whether to meet collection to the configuration information needed for the single processing node of cluster and cluster scale is calculated
The calculation processing ability of group.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201711413516.9A CN108197012B (en) | 2017-12-24 | 2017-12-24 | Remote sensing data distributed processing cluster scale and performance measurement method |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201711413516.9A CN108197012B (en) | 2017-12-24 | 2017-12-24 | Remote sensing data distributed processing cluster scale and performance measurement method |
Publications (2)
Publication Number | Publication Date |
---|---|
CN108197012A true CN108197012A (en) | 2018-06-22 |
CN108197012B CN108197012B (en) | 2021-04-23 |
Family
ID=62583778
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN201711413516.9A Active CN108197012B (en) | 2017-12-24 | 2017-12-24 | Remote sensing data distributed processing cluster scale and performance measurement method |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN108197012B (en) |
Cited By (1)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN109088747A (en) * | 2018-07-10 | 2018-12-25 | 郑州云海信息技术有限公司 | The management method and device of resource in cloud computing system |
Citations (5)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN101374076A (en) * | 2008-10-06 | 2009-02-25 | 中国移动通信集团设计院有限公司 | Method for forecasting and planning server handling ability of telecom business support system |
CN102169494A (en) * | 2011-04-06 | 2011-08-31 | 北京师范大学 | Distributed remotely-sensed data managing system and method |
CN105743783A (en) * | 2016-04-12 | 2016-07-06 | 同济大学 | Car-Networking Node Selecting Method based on BS-TS and Autoencoder Network, and Accessibility Routing Mechanism Thereof |
US20170026312A1 (en) * | 2015-07-22 | 2017-01-26 | Netapp, Inc. | Methods and systems for managing a resource in a networked storage environment |
CN106790099A (en) * | 2016-12-26 | 2017-05-31 | 大连大学 | Charge network survivability Measurement Method based on network survivability entropy |
-
2017
- 2017-12-24 CN CN201711413516.9A patent/CN108197012B/en active Active
Patent Citations (5)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN101374076A (en) * | 2008-10-06 | 2009-02-25 | 中国移动通信集团设计院有限公司 | Method for forecasting and planning server handling ability of telecom business support system |
CN102169494A (en) * | 2011-04-06 | 2011-08-31 | 北京师范大学 | Distributed remotely-sensed data managing system and method |
US20170026312A1 (en) * | 2015-07-22 | 2017-01-26 | Netapp, Inc. | Methods and systems for managing a resource in a networked storage environment |
CN105743783A (en) * | 2016-04-12 | 2016-07-06 | 同济大学 | Car-Networking Node Selecting Method based on BS-TS and Autoencoder Network, and Accessibility Routing Mechanism Thereof |
CN106790099A (en) * | 2016-12-26 | 2017-05-31 | 大连大学 | Charge network survivability Measurement Method based on network survivability entropy |
Non-Patent Citations (1)
Title |
---|
胡沅等: "遥感卫星在生态环境监测领域的应用满足度分析", 《技术进展》 * |
Cited By (1)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN109088747A (en) * | 2018-07-10 | 2018-12-25 | 郑州云海信息技术有限公司 | The management method and device of resource in cloud computing system |
Also Published As
Publication number | Publication date |
---|---|
CN108197012B (en) | 2021-04-23 |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN108593990B (en) | Electricity stealing detection method based on electricity consumption behavior mode of electric energy user and application | |
CN110471024B (en) | Intelligent electric meter online remote calibration method based on measurement data analysis | |
CN105260803B (en) | A kind of system power consumption prediction technique | |
WO2023010759A1 (en) | Power distribution and sale competitive situation-based regional power distribution network gridding load interval prediction method | |
US10564315B2 (en) | Methods for location identification of renewable energy systems | |
CN102999791A (en) | Power load forecasting method based on customer segmentation in power industry | |
CN109086951B (en) | Multi-stage space load prediction method considering city development degree | |
CN110501761B (en) | Forecasting method for predicting and forecasting crops ETc in different forecast periods | |
CN110119590A (en) | A kind of water quality model particle filter assimilation method based on multi-source observation data | |
CN101916335A (en) | Prediction method of city water-requirement time series-exponent smoothing model | |
CN104112163A (en) | Construction method of electric power forecasting business index | |
CN108921425A (en) | A kind of method, system and the server of asset item classifcation of investment | |
CN108197012A (en) | Remotely-sensed data distributed treatment cluster scale and performance measure method | |
CN105262148B (en) | The planning year power balance method of meter and wind power output characteristic | |
CN104091227B (en) | Method for analyzing load density of power consumers in city | |
CN104063810B (en) | A kind of electricity consumption factor based on big data influences the construction method of model | |
CN109559046A (en) | A kind of natural gas development influence factor appraisal procedure and assessment system | |
CN107846039A (en) | Consider the cluster wind-electricity integration modeling and analysis methods and system of wind speed correlation | |
CN109657854A (en) | A kind of method, system and the equipment of water demand of crop prediction | |
CN114548609B (en) | Water resource collaborative allocation method for ecological power generation and water supply | |
CN106355264A (en) | Combined prediction method of reference crop evapotranspiration | |
CN115936545A (en) | Two-parameter monthly scale hydrological model building method based on hydrothermal complementary equation | |
CN113723719B (en) | Distribution transformation-based near-intermediate load prediction method, device, equipment and medium | |
CN115358441A (en) | New energy cluster consumption intelligent control method and device based on federal learning | |
CN112507635B (en) | Method for quantitatively evaluating flood regulation and storage functions of watershed wetland |
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 |