CN110222823A - A kind of hydrology flowed fluctuation situation recognition methods and its system - Google Patents
A kind of hydrology flowed fluctuation situation recognition methods and its system Download PDFInfo
- Publication number
- CN110222823A CN110222823A CN201910466916.9A CN201910466916A CN110222823A CN 110222823 A CN110222823 A CN 110222823A CN 201910466916 A CN201910466916 A CN 201910466916A CN 110222823 A CN110222823 A CN 110222823A
- Authority
- CN
- China
- Prior art keywords
- neural network
- parameter
- convolutional neural
- experimental data
- water ripple
- 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 50
- XLYOFNOQVPJJNP-UHFFFAOYSA-N water Substances O XLYOFNOQVPJJNP-UHFFFAOYSA-N 0.000 claims abstract description 150
- 238000013527 convolutional neural network Methods 0.000 claims abstract description 126
- 238000012360 testing method Methods 0.000 claims abstract description 74
- 238000012549 training Methods 0.000 claims abstract description 21
- 238000013528 artificial neural network Methods 0.000 claims description 28
- 238000004590 computer program Methods 0.000 claims description 11
- 238000004364 calculation method Methods 0.000 claims description 7
- 238000010276 construction Methods 0.000 claims description 5
- 238000012545 processing Methods 0.000 abstract description 7
- 238000010586 diagram Methods 0.000 description 7
- 230000006870 function Effects 0.000 description 7
- 230000008569 process Effects 0.000 description 6
- 238000004891 communication Methods 0.000 description 3
- 230000005540 biological transmission Effects 0.000 description 2
- 238000012986 modification Methods 0.000 description 2
- 230000004048 modification Effects 0.000 description 2
- 238000012935 Averaging Methods 0.000 description 1
- 241000283690 Bos taurus Species 0.000 description 1
- 238000013500 data storage Methods 0.000 description 1
- 238000002474 experimental method Methods 0.000 description 1
- 230000006872 improvement Effects 0.000 description 1
- 230000003993 interaction Effects 0.000 description 1
- 230000003287 optical effect Effects 0.000 description 1
- 239000002245 particle Substances 0.000 description 1
- 238000011160 research Methods 0.000 description 1
- 238000012827 research and development Methods 0.000 description 1
- 238000004088 simulation Methods 0.000 description 1
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
- G06N3/04—Architecture, e.g. interconnection topology
- G06N3/045—Combinations of networks
-
- 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
Landscapes
- Engineering & Computer Science (AREA)
- Theoretical Computer Science (AREA)
- Physics & Mathematics (AREA)
- General Health & Medical Sciences (AREA)
- Computing Systems (AREA)
- Biomedical Technology (AREA)
- Biophysics (AREA)
- Computational Linguistics (AREA)
- Data Mining & Analysis (AREA)
- Evolutionary Computation (AREA)
- Life Sciences & Earth Sciences (AREA)
- Molecular Biology (AREA)
- Artificial Intelligence (AREA)
- General Engineering & Computer Science (AREA)
- General Physics & Mathematics (AREA)
- Mathematical Physics (AREA)
- Software Systems (AREA)
- Health & Medical Sciences (AREA)
- Image Analysis (AREA)
- Management, Administration, Business Operations System, And Electronic Commerce (AREA)
Abstract
The present embodiments relate to technical field of data processing, in particular to a kind of hydrology flowed fluctuation situation recognition methods and its system.This method is respectively using every group of water wave wave experimental data as test set, the convolutional neural networks for completing training are used to be identified to the test set to obtain the second recognition accuracy, and the first parameter of the convolutional neural networks for completing training is adjusted to obtain the second parameter and cache according to whether the second recognition accuracy reaches the second setting value, the parameter of caching is averaged to obtain third parameter, third time configuration is carried out to constructed convolutional neural networks according to third parameter, and current water wave wave number is completed into the convolutional neural networks of third time configuration according to input to realize the identification of water wave wave fluctuation situation, so, the stronger convolutional neural networks of generalization ability can be built based on limited water wave wave experimental data and realize that the fluctuation situation of water wave wave is accurately predicted and identified based on the convolutional neural networks.
Description
Technical Field
The embodiment of the invention relates to the technical field of data processing, in particular to a hydrologic flow fluctuation situation identification method and a hydrologic flow fluctuation situation identification system.
Background
The water ripple wave is a relatively special wave, the water ripple wave does not belong to longitudinal wave nor transverse wave, particle points in the water ripple wave move along an elliptical orbit, the projection of the movement in the vertical direction is just like transverse wave, the projection in the horizontal direction is just like longitudinal wave, if two rows of waves are emitted from two wave sources at the same time, and the energy of the wave sources cannot appear to accord with the two laws of cattle at the same time. Therefore, it becomes a difficult problem how to accurately predict and identify the fluctuation situation of the water ripple.
Disclosure of Invention
In view of this, the invention provides a method and a system for recognizing a fluctuation situation of hydrologic flow.
The embodiment of the invention provides a hydrologic flow fluctuation situation identification method, which comprises the following steps:
acquiring a plurality of groups of water ripple experimental data;
constructing a convolutional neural network;
aiming at each group of water ripple experimental data in the multiple groups of water ripple experimental data, inputting the water ripple experimental data as a test set into the convolutional neural network for training, so that the first identification accuracy of the trained convolutional neural network for identifying each group of water ripple experimental data in the other groups of water ripple experimental data reaches a first set value, and acquiring a first parameter of the trained convolutional neural network;
performing first configuration on the trained convolutional neural network based on the first parameter, and inputting the test set into the convolutional neural network after the first configuration is completed so as to obtain a second identification accuracy rate obtained by identifying the test set by the convolutional neural network after the first configuration is completed;
judging whether the second identification accuracy reaches a second set value or not; if so, caching the first parameter; if not, adjusting the first parameter to obtain a second parameter, performing second configuration on the convolutional neural network which completes the first configuration based on the second parameter so that a second identification accuracy rate of the convolutional neural network which completes the second configuration for identifying the test set reaches a second set value, and caching the first parameter and the second parameter according to a set rule;
calculating the average value of the parameters obtained by caching as a third parameter, and carrying out third configuration on the constructed convolutional neural network according to the third parameter;
and acquiring current water ripple data, and inputting the current water ripple data into a convolution neural network which is configured for the third time so as to realize the identification of the water ripple fluctuation situation.
Optionally, each set of ripple experimental data in the multiple sets of ripple experimental data includes n consecutive three-dimensional coordinates of at least one first pixel within a preset time period;
the first recognition accuracy is calculated by the following method:
wherein,
pre1is the first recognition accuracy;
n1predicting the three-dimensional coordinates of at least one first pixel in each group of water ripple experimental data at the next moment of each three-dimensional coordinate in a preset time period by adopting a trained convolutional neural network and predicting the correct time sum;
in order to sum the times of predicting the three-dimensional coordinates of at least one first pixel in each group of water ripple experimental data at the next moment of each three-dimensional coordinate within a preset time period by adopting the trained convolutional neural network,
the second recognition accuracy is calculated by the following method:
pre2a second recognition accuracy;
n2predicting three-dimensional coordinates of at least one first pixel in the water wave experimental data corresponding to the test set at the next moment of each three-dimensional coordinate in a preset time period by adopting a convolution neural network which is configured for the first time, and predicting the correct time sum;
in order to adopt the convolution neural network which completes the first configuration to predict the three-dimensional coordinates of at least one first pixel in the water wave experimental data corresponding to the test set at the next moment of each three-dimensional coordinate in the preset time period,
optionally, the caching the first parameter and the second parameter according to a set rule includes:
carrying out weighted summation on the first parameter and the second parameter according to a set formula, and caching a weighted summation value; the set formula is as follows:
wherein,
sumitaking the ith group of water ripple experimental data as a weighted summation value when the test set is formed, wherein i belongs to [1, m ∈]∩i∈Z；
para1-iTaking the ith group of water ripple experimental data as a first parameter of the convolutional neural network which is correspondingly trained when the test set is used;
para2-iwhen the ith group of water wave experimental data is used as a test set and the corresponding convolutional neural network which finishes training identifies the ith group of water wave experimental data to obtain a second identification accuracy rate which does not reach a second set value, the first parameter para is subjected to the first parameter para1-iAnd adjusting the obtained second parameter.
Optionally, the step of inputting the current water ripple data into a convolution neural network configured for the third time to realize identification of the water ripple fluctuation situation includes:
inputting the current-time three-dimensional coordinate into the convolution neural network which completes the third configuration so that the convolution neural network which completes the third configuration identifies the next-time three-dimensional coordinate of the at least one second pixel.
The embodiment of the invention also provides a hydrologic flow fluctuation situation recognition system, which comprises:
the experimental data acquisition module is used for acquiring a plurality of groups of water wave experimental data;
the convolutional neural network construction module is used for constructing a convolutional neural network;
a first parameter obtaining module, configured to, for each group of water ripple experimental data in the multiple groups of water ripple experimental data, use the water ripple experimental data as a test set, input other groups of water ripple experimental data in the multiple groups of water ripple experimental data except the test set into the convolutional neural network for training, so that a first recognition accuracy of the trained convolutional neural network when recognizing each group of water ripple experimental data in the other groups of water ripple experimental data reaches a first set value, and obtain a first parameter of the trained convolutional neural network;
the identification accuracy rate calculation module is used for carrying out first configuration on the trained convolutional neural network based on the first parameter, and inputting the test set into the convolutional neural network which completes the first configuration to obtain second identification accuracy rate obtained by identifying the test set by the convolutional neural network which completes the first configuration;
the parameter caching module is used for judging whether the second identification accuracy reaches a second set value or not; if so, caching the first parameter; if not, adjusting the first parameter to obtain a second parameter, performing second configuration on the convolutional neural network which completes the first configuration based on the second parameter so that a second identification accuracy rate of the convolutional neural network which completes the second configuration for identifying the test set reaches a second set value, and caching the first parameter and the second parameter according to a set rule;
the convolutional neural network configuration module is used for calculating the average value of the parameters obtained by caching as a third parameter and carrying out third configuration on the constructed convolutional neural network according to the third parameter;
and the identification module is used for acquiring current water ripple data and inputting the current water ripple data into the convolution neural network which completes the third configuration so as to realize the identification of the water ripple fluctuation situation.
Optionally, each set of ripple experimental data in the multiple sets of ripple experimental data includes n consecutive three-dimensional coordinates of at least one first pixel within a preset time period;
the first recognition accuracy is calculated by the following method:
wherein,
pre1is the first recognition accuracy;
n1predicting the three-dimensional coordinates of at least one first pixel in each group of water ripple experimental data at the next moment of each three-dimensional coordinate in a preset time period by adopting a trained convolutional neural network and predicting the correct time sum;
in order to sum the times of predicting the three-dimensional coordinates of at least one first pixel in each group of water ripple experimental data at the next moment of each three-dimensional coordinate within a preset time period by adopting the trained convolutional neural network,
the second recognition accuracy is calculated by the following method:
pre2a second recognition accuracy;
n2predicting three-dimensional coordinates of at least one first pixel in the water wave experimental data corresponding to the test set at the next moment of each three-dimensional coordinate in a preset time period by adopting a convolution neural network which is configured for the first time, and predicting the correct time sum;
in order to adopt the convolution neural network which completes the first configuration to predict the three-dimensional coordinates of at least one first pixel in the water wave experimental data corresponding to the test set at the next moment of each three-dimensional coordinate in the preset time period,
optionally, the multiple groups of water ripple experimental data are m groups, and the parameter caching module caches the first parameter and the second parameter according to a set rule in the following manner:
carrying out weighted summation on the first parameter and the second parameter according to a set formula, and caching a weighted summation value; the set formula is as follows:
wherein,
sumitaking the ith group of water ripple experimental data as a weighted summation value when the test set is formed, wherein i belongs to [1, m ∈]∩i∈Z；
para1-iFor the ith group of water wavesWhen the test data is used as a test set, the first parameter of the convolutional neural network which completes training correspondingly;
para2-iwhen the ith group of water wave experimental data is used as a test set and the corresponding convolutional neural network which finishes training identifies the ith group of water wave experimental data to obtain a second identification accuracy rate which does not reach a second set value, the first parameter para is subjected to the first parameter para1-iAnd adjusting the obtained second parameter.
Optionally, the current water wave data includes a current-time three-dimensional coordinate of at least one second pixel, and the identification module inputs the current water wave data into a convolutional neural network configured for the third time to realize identification of the water ripple fluctuation situation by:
inputting the current-time three-dimensional coordinate into the convolution neural network which completes the third configuration so that the convolution neural network which completes the third configuration identifies the next-time three-dimensional coordinate of the at least one second pixel.
The embodiment of the invention also provides electronic equipment which comprises a memory, a processor and a computer program which is stored on the memory and can be run on the processor, wherein the processor realizes the hydrologic flow fluctuation situation recognition method when executing the computer program.
The embodiment of the invention also provides a computer readable storage medium, which comprises a computer program, and the computer program controls the electronic equipment where the readable storage medium is located to execute the hydrologic flow fluctuation situation recognition method when running
The hydrographic flow fluctuation situation recognition method and the hydrographic flow fluctuation situation recognition system provided by the embodiment of the invention respectively take each group of hydrographic wave experimental data in a plurality of groups of hydrographic wave experimental data as a test set, recognize the test set by adopting a trained convolutional neural network to obtain a second recognition accuracy, adjust a first parameter of the trained convolutional neural network according to whether the second recognition accuracy reaches a second set value to obtain a second parameter, cache the first parameter and the second parameter, average the cached parameter to obtain a third parameter, configure the constructed convolutional neural network for the third time according to the third parameter, and input the obtained current hydrographic wave data into the convolutional neural network which completes the third configuration to realize the recognition of the hydrographic flow fluctuation situation, so that the convolutional neural network with stronger generalization capability can be built on the basis of the limited hydrographic wave experimental data and the water ripple situation can be realized on the basis of the convolutional neural network The fluctuation situation is accurately predicted and identified.
Drawings
In order to more clearly illustrate the technical solutions of the embodiments of the present invention, the drawings needed to be used in the embodiments will be briefly described below, it should be understood that the following drawings only illustrate some embodiments of the present invention and therefore should not be considered as limiting the scope, and for those skilled in the art, other related drawings can be obtained according to the drawings without inventive efforts.
Fig. 1 is a block diagram of an electronic device according to an embodiment of the present invention.
Fig. 2 is a flowchart of a method for recognizing a fluctuation situation of a hydrological flow according to an embodiment of the present invention.
Fig. 3 is a block diagram of a hydrologic flow fluctuation situation recognition system according to an embodiment of the present invention.
Icon:
10-an electronic device; 11-a memory; 12-a processor; 13-a network module;
20-a hydrologic flow fluctuation situation recognition system; 21-an experimental data acquisition module; 22-a convolutional neural network construction module; 23-a first parameter acquisition module; 24-an identification accuracy calculation module; 25-parameter cache module; 26-a convolutional neural network configuration module; 27-identification module.
Detailed Description
In order to make the objects, technical solutions and advantages of the embodiments of the present invention clearer, the technical solutions in the embodiments of the present invention will be clearly and completely described below with reference to the drawings in the embodiments of the present invention, and it is obvious that the described embodiments are only a part of the embodiments of the present invention, and not all of the embodiments. The components of embodiments of the present invention generally described and illustrated in the figures herein may be arranged and designed in a wide variety of different configurations.
The inventor finds that the prior art is difficult to accurately predict and identify the fluctuation situation of the water ripple. On one hand, the accuracy of the common prediction and identification method is low; on the other hand, a large amount of experimental data is needed for simulation in the early stage of some methods with high prediction and identification accuracy, the labor cost and the research and development cost are high, and in addition, the generalization capability of the prediction and identification methods with high accuracy is low.
The above prior art solutions have shortcomings which are the results of practical and careful study of the inventor, and therefore, the discovery process of the above problems and the solutions proposed by the following embodiments of the present invention to the above problems should be the contribution of the inventor to the present invention in the course of the present invention.
Based on the research, the embodiment of the invention provides a method and a system for recognizing the fluctuation situation of the hydrological flow, which can build a convolutional neural network with strong generalization capability based on limited experimental data of the water ripple and realize accurate prediction and recognition of the fluctuation situation of the water ripple based on the convolutional neural network.
Fig. 1 shows a block diagram of an electronic device 10 according to an embodiment of the present invention. The electronic device 10 in the embodiment of the present invention has functions of data storage, transmission, and processing, and as shown in fig. 1, the electronic device 10 includes: the device comprises a memory 11, a processor 12, a network module 13 and a hydrologic flow fluctuation situation recognition system 20.
The memory 11, the processor 12 and the network module 13 are electrically connected directly or indirectly to realize data transmission or interaction. For example, the components may be electrically connected to each other via one or more communication buses or signal lines. The memory 11 stores a hydrological flow fluctuation situation recognition system 20, the hydrological flow fluctuation situation recognition system 20 includes at least one software function module which can be stored in the memory 11 in a form of software or firmware (firmware), and the processor 12 executes various function applications and data processing by running a software program and a module stored in the memory 11, such as the hydrological flow fluctuation situation recognition system 20 in the embodiment of the present invention, so as to implement the hydrological flow fluctuation situation recognition method in the embodiment of the present invention.
The memory 11 is used for storing a program, and the processor 12 executes the program after receiving the execution instruction. The processor 12 may be an integrated circuit chip having data processing capabilities. The Processor 12 may be a general-purpose Processor including a Central Processing Unit (CPU), a Network Processor (NP), and the like. The various methods, steps and logic blocks disclosed in embodiments of the present invention may be implemented or performed. A general purpose processor may be a microprocessor or the processor may be any conventional processor or the like. The network module 13 is used for establishing communication connection between the electronic device 10 and other communication terminal devices through a network, and implementing transceiving operation of network signals and data. The network signal may include a wireless signal or a wired signal.
The embodiment of the present invention further provides a computer-readable storage medium, where the computer-readable storage medium includes a computer program, and the computer program controls, when running, the electronic device 10 where the computer-readable storage medium is located to execute the following method for recognizing the fluctuation situation of the hydrologic flow rate.
Fig. 2 shows a flowchart of a method for recognizing a fluctuation situation of a hydrological flow according to an embodiment of the present invention. The method steps defined by the method-related flow, as applied to the electronic device 10, may be implemented by the processor 12. The specific process shown in FIG. 2 will be described in detail below:
and step S21, acquiring multiple groups of water ripple test data.
In this example, the water wave test data are m groups.
And step S22, constructing a convolutional neural network.
And constructing the obtained convolutional neural network for subsequent training and recognition.
And step S23, training the convolutional neural network, and acquiring a first parameter of the trained convolutional neural network.
In this embodiment, each of the m groups of water wave test data is used as a test set, and the m-1 groups of water wave test data are used as a training set to train the convolutional neural network constructed in step S22.
In the embodiment, each set of water ripple test data includes n consecutive three-dimensional coordinates of at least one first pixel within a preset time period.
Wherein, the training process is as follows:
taking the ith group of water wave experimental data as an example of a test set, predicting the three-dimensional coordinate of at least one first pixel in each group of water wave experimental data in the m-1 groups of water wave experimental data at the next moment corresponding to each three-dimensional coordinate in n continuous three-dimensional coordinates in a preset time period, and further obtaining a first identification accuracy pre1. As can be appreciated, the first recognition accuracy pre is obtained1A total of m-1, if m-1, first recognition accuracy rates pre1All reach the first set value (in this embodiment, the first set value may be set to 80%), and the first parameter para of the convolutional neural network which is trained at this time is obtained1-iIt is understood that para1-iAnd taking the ith group of water ripple experimental data as a first parameter of the convolutional neural network which is correspondingly trained when the test set is finished.
In this embodiment, the first recognition accuracy is calculated by:
wherein,
pre1is the first recognition accuracy;
n1predicting the three-dimensional coordinates of at least one first pixel in each group of water ripple experimental data at the next moment of each three-dimensional coordinate in a preset time period by adopting a trained convolutional neural network and predicting the correct time sum;
in order to sum the times of predicting the three-dimensional coordinates of at least one first pixel in each group of water ripple experimental data at the next moment of each three-dimensional coordinate within a preset time period by adopting the trained convolutional neural network,
for another example, taking m as 5, taking the 1 st group of water wave experimental data as an example of the test set, the 2 nd to 5 th groups of water wave experimental data are input into the convolutional neural network constructed in step S22, and parameters of the convolutional neural network are adjusted so that the convolutional neural network recognizes the 2 nd to 5 th groups of water wave experimental data to obtain the first recognition accuracy pre1Reaches 80%, the first parameter para at this time is obtained1-1。
And step S24, performing first configuration on the trained convolutional neural network based on the first parameter, and inputting the test set into the convolutional neural network after the first configuration to obtain a second identification accuracy rate obtained by identifying the test set by the convolutional neural network after the first configuration.
Take m-5 as an example, based on the first parameter para1-1Performing first configuration on the trained convolutional neural network, inputting the 1 st group of water ripple experiment data into the convolutional neural network which completes the first configuration to obtain a second identification accuracy pre obtained by identifying the test set by the convolutional neural network which completes the first configuration2Wherein:
pre2a second recognition accuracy;
n2predicting three-dimensional coordinates of at least one first pixel in the water wave experimental data corresponding to the test set at the next moment of each three-dimensional coordinate in a preset time period by adopting a convolution neural network which is configured for the first time, and predicting the correct time sum;
in order to adopt the convolution neural network which completes the first configuration to predict the three-dimensional coordinates of at least one first pixel in the water wave experimental data corresponding to the test set at the next moment of each three-dimensional coordinate in the preset time period,
it can be understood that the second recognition accuracy rate obtained by recognizing the 1 st group of water wave experimental data through the convolutional neural network trained by taking the 2 nd to 5 th groups of water wave experimental data as the training set is pre2。
And step S25, determining whether the second recognition accuracy reaches a second set value, and performing a corresponding operation according to the determination result.
In the embodiment, the second set value is higher than the first set value, for example, the second set value may be 90%.
If pre2Up to 90% of para1-1Buffer memory if pre2Not up to 90% for para1-1Adjusting to obtain the second parameter para using the 1 st group of water wave experimental data as the test set2-1. Wherein the second parameter para is based on2-1The second configuration of the convolutional neural network which completes the first configuration can enable the convolutional neural network which completes the second configuration to identify the second identification accuracy pre of the test set2Up to 90%.
Further, for para1-1And para2-1And carrying out weighted summation according to a set formula, and caching a weighted summation value.
In this embodiment, the formula is set as:
wherein,
sumitaking the ith group of water ripple experimental data as a weighted summation value when the test set is formed, wherein i belongs to [1, m ∈]∩i∈Z；
para1-iTaking the ith group of water ripple experimental data as a first parameter of the convolutional neural network which is correspondingly trained when the test set is used;
para2-iwhen the ith group of water wave experimental data is used as a test set and the corresponding convolutional neural network which finishes training identifies the ith group of water wave experimental data to obtain a second identification accuracy rate which does not reach a second set value, the first parameter para is subjected to the first parameter para1-iAdjusted secondAnd (4) parameters.
For example, for para1-1And para2-1And carrying out weighted summation according to a set formula to obtain:
will su m1And carrying out caching.
It is understood that the above steps are iterated 5 times, in other words, the total number of buffered parameters is 5, the buffered parameters include at least one of the first parameter and the weighted sum value, and for another example, the buffered parameters are: sum1、para1-2、para1-3、sum4And para1-5。
And step S26, calculating the average value of the parameters obtained by caching as a third parameter, and carrying out third configuration on the constructed convolutional neural network according to the third parameter.
For example, for sum1、para1-2、para1-3、sum4And para1-5And averaging to obtain a third parameter, and performing third configuration on the constructed convolutional neural network according to the third parameter.
And step S27, acquiring current water wave data, and inputting the current water wave data into the convolution neural network which completes the third configuration so as to realize the identification of the water wave fluctuation situation.
In this embodiment, the current water wave data includes a current time three-dimensional coordinate of at least one second pixel.
The method for identifying and predicting the current water wave data comprises the following steps: and inputting the current-time three-dimensional coordinate of the at least one second pixel into the convolution neural network which completes the third configuration so that the convolution neural network which completes the third configuration identifies the next-time three-dimensional coordinate of the at least one second pixel, thereby realizing accurate prediction and identification of the water ripple fluctuation situation.
In this embodiment, the convolutional neural network is trained by using each group of water ripple test data in m groups of water ripple test data as a test set and using m-1 groups of water ripple test data as a training set, so that the generalization capability of the convolutional neural network can be increased on the basis of a limited group of water ripple test data.
In addition, the first parameter is adjusted to obtain the second parameter on the premise that the second identification accuracy rate obtained by identifying the test set by adopting the convolutional neural network which is configured for the first time does not reach the second set value, and the weighted summation is carried out on the basis of the first parameter and the second parameter, so that the weights of the training set and the test set can be taken into consideration, and the accuracy of the water wave fluctuation situation identification is further improved.
Furthermore, the convolutional neural network is configured by adopting the third parameter obtained by calculation, so that the smooth processing of the convolutional neural network can be realized, the volatility of the recognition result is reduced, and the accuracy of the water wave fluctuation situation recognition is further improved.
On the basis, as shown in fig. 3, an embodiment of the present invention provides a hydrologic flow fluctuation situation recognition system 20, where the hydrologic flow fluctuation situation recognition system 20 includes: the device comprises an experimental data acquisition module 21, a convolutional neural network construction module 22, a first parameter acquisition module 23, an identification accuracy calculation module 24, a parameter cache module 25, a convolutional neural network configuration module 26 and an identification module 27.
And the experimental data acquisition module 21 is configured to acquire multiple groups of water ripple experimental data.
Since the experimental data acquisition module 21 is similar to the implementation principle of step S21 in fig. 2, it will not be further described here.
And the convolutional neural network construction module 22 is used for constructing a convolutional neural network.
Since the convolutional neural network building block 22 is similar to the implementation principle of step S22 in fig. 2, it will not be further described here.
A first parameter obtaining module 23, configured to, for each group of water ripple experimental data in the multiple groups of water ripple experimental data, use the water ripple experimental data as a test set, input other groups of water ripple experimental data in the multiple groups of water ripple experimental data except the test set into the convolutional neural network for training, so that a first recognition accuracy of the trained convolutional neural network when recognizing each group of water ripple experimental data in the other groups of water ripple experimental data reaches a first set value, and obtain a first parameter of the trained convolutional neural network; .
Since the first parameter obtaining module 23 is similar to the implementation principle of step S23 in fig. 2, it will not be further described here.
And the identification accuracy calculation module 24 is configured to perform first configuration on the trained convolutional neural network based on the first parameter, and input the test set into the first configured convolutional neural network to obtain a second identification accuracy obtained by identifying the test set by the first configured convolutional neural network.
Since the recognition accuracy calculation module 24 is similar to the implementation principle of step S24 in fig. 2, it will not be further described here.
The parameter caching module 25 is configured to determine whether the second recognition accuracy reaches a second set value; if so, caching the first parameter; if not, adjusting the first parameter to obtain a second parameter, performing second configuration on the convolutional neural network which completes the first configuration based on the second parameter, so that a second identification accuracy rate of the convolutional neural network which completes the second configuration for identifying the test set reaches a second set value, and caching the first parameter and the second parameter according to a set rule.
Since the implementation principle of the parameter caching module 25 is similar to that of step S25 in fig. 2, no further description is provided here.
And the convolutional neural network configuration module 26 is configured to calculate an average value of the parameters obtained by caching as a third parameter, and perform third configuration on the constructed convolutional neural network according to the third parameter.
Since the convolutional neural network configuration module 26 is similar to the implementation principle of step S26 in fig. 2, it will not be further described here.
And the identification module 27 is configured to acquire current water ripple data, and input the current water ripple data into the convolution neural network configured for the third time to identify the water ripple fluctuation situation.
Since the identification module 27 is similar to the implementation principle of step S27 in fig. 2, it will not be further described here.
In summary, the method and the system for recognizing the fluctuation situation of the hydrological flow provided by the embodiment of the invention can build a convolutional neural network with strong generalization capability based on limited experimental data of the water ripple and realize accurate prediction and recognition of the fluctuation situation of the water ripple based on the convolutional neural network.
In the embodiments provided in the present invention, it should be understood that the disclosed apparatus and method can be implemented in other ways. The apparatus and method embodiments described above are illustrative only, as the flowcharts and block diagrams in the figures illustrate the architecture, functionality, and operation of possible implementations of apparatus, methods and computer program products according to various embodiments of the present invention. In this regard, each block in the flowchart or block diagrams may represent a module, segment, or portion of code, which comprises one or more executable instructions for implementing the specified logical function(s). It should also be noted that, in some alternative implementations, the functions noted in the block may occur out of the order noted in the figures. For example, two blocks shown in succession may, in fact, be executed substantially concurrently, or the blocks may sometimes be executed in the reverse order, depending upon the functionality involved. It will also be noted that each block of the block diagrams and/or flowchart illustration, and combinations of blocks in the block diagrams and/or flowchart illustration, can be implemented by special purpose hardware-based systems which perform the specified functions or acts, or combinations of special purpose hardware and computer instructions.
In addition, the functional modules in the embodiments of the present invention may be integrated together to form an independent part, or each module may exist separately, or two or more modules may be integrated to form an independent part.
The functions, if implemented in the form of software functional modules and sold or used as a stand-alone product, may be stored in a computer readable storage medium. Based on such understanding, the technical solution of the present invention may be embodied in the form of a software product, which is stored in a storage medium and includes instructions for causing a computer device (which may be a personal computer, an electronic device, or a network device) to perform all or part of the steps of the method according to the embodiments of the present invention. And the aforementioned storage medium includes: various media capable of storing program codes, such as a usb disk, a removable hard disk, a Read-Only Memory (ROM), a Random Access Memory (RAM), a magnetic disk, or an optical disk. It should be noted that, in this document, the terms "comprises," "comprising," or any other variation thereof, are intended to cover a non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements does not include only those elements but may include other elements not expressly listed or inherent to such process, method, article, or apparatus. Without further limitation, an element defined by the phrase "comprising an … …" does not exclude the presence of other identical elements in a process, method, article, or apparatus that comprises the element.
The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention, and various modifications and changes may be made by those skilled in the art. Any modification, equivalent replacement, or improvement made within the spirit and principle of the present invention should be included in the protection scope of the present invention.
Claims (10)
1. A hydrologic flow fluctuation situation recognition method is characterized by comprising the following steps:
acquiring a plurality of groups of water ripple experimental data;
constructing a convolutional neural network;
aiming at each group of water ripple experimental data in the multiple groups of water ripple experimental data, inputting the water ripple experimental data as a test set into the convolutional neural network for training, so that the first identification accuracy of the trained convolutional neural network for identifying each group of water ripple experimental data in the other groups of water ripple experimental data reaches a first set value, and acquiring a first parameter of the trained convolutional neural network;
performing first configuration on the trained convolutional neural network based on the first parameter, and inputting the test set into the convolutional neural network after the first configuration is completed so as to obtain a second identification accuracy rate obtained by identifying the test set by the convolutional neural network after the first configuration is completed;
judging whether the second identification accuracy reaches a second set value or not; if so, caching the first parameter; if not, adjusting the first parameter to obtain a second parameter, performing second configuration on the convolutional neural network which completes the first configuration based on the second parameter so that a second identification accuracy rate of the convolutional neural network which completes the second configuration for identifying the test set reaches a second set value, and caching the first parameter and the second parameter according to a set rule;
calculating the average value of the parameters obtained by caching as a third parameter, and carrying out third configuration on the constructed convolutional neural network according to the third parameter;
and acquiring current water ripple data, and inputting the current water ripple data into a convolution neural network which is configured for the third time so as to realize the identification of the water ripple fluctuation situation.
2. The hydrologic flow fluctuation situation identification method according to claim 1, wherein each of the plurality of sets of water ripple experimental data comprises n consecutive three-dimensional coordinates of at least one first pixel within a preset time period;
the first recognition accuracy is calculated by the following method:
wherein,
pre1is the first recognition accuracy;
n1predicting the three-dimensional coordinates of at least one first pixel in each group of water ripple experimental data at the next moment of each three-dimensional coordinate in a preset time period by adopting a trained convolutional neural network and predicting the correct time sum;
in order to sum the times of predicting the three-dimensional coordinates of at least one first pixel in each group of water ripple experimental data at the next moment of each three-dimensional coordinate within a preset time period by adopting the trained convolutional neural network,
the second recognition accuracy is calculated by the following method:
pre2a second recognition accuracy;
n2predicting three-dimensional coordinates of at least one first pixel in the water wave experimental data corresponding to the test set at the next moment of each three-dimensional coordinate in a preset time period by adopting a convolution neural network which is configured for the first time, and predicting the correct time sum;
in order to adopt the convolution neural network which completes the first configuration to predict the three-dimensional coordinates of at least one first pixel in the water wave experimental data corresponding to the test set at the next moment of each three-dimensional coordinate in the preset time period,
3. the method for recognizing the fluctuation situation of the hydrological flow according to claim 1, wherein the multiple groups of water ripple experimental data are m groups, and the step of caching the first parameter and the second parameter according to a set rule comprises the steps of:
carrying out weighted summation on the first parameter and the second parameter according to a set formula, and caching a weighted summation value; the set formula is as follows:
wherein,
sumitaking the ith group of water ripple experimental data as a weighted summation value when the test set is formed, wherein i belongs to [1, m ∈]∩i∈Z；
para1-iTaking the ith group of water ripple experimental data as a first parameter of the convolutional neural network which is correspondingly trained when the test set is used;
para2-iwhen the ith group of water wave experimental data is used as a test set and the corresponding convolutional neural network which finishes training identifies the ith group of water wave experimental data to obtain a second identification accuracy rate which does not reach a second set value, the first parameter para is subjected to the first parameter para1-iAnd adjusting the obtained second parameter.
4. The method for recognizing the fluctuation situation of the hydrological flow according to claim 1, wherein the current water ripple data includes a current-time three-dimensional coordinate of at least one second pixel, and the step of inputting the current water ripple data into a convolutional neural network which is configured for the third time to realize recognition of the fluctuation situation of the water ripple comprises:
inputting the current-time three-dimensional coordinate into the convolution neural network which completes the third configuration so that the convolution neural network which completes the third configuration identifies the next-time three-dimensional coordinate of the at least one second pixel.
5. A hydrologic flow fluctuation situation recognition system is characterized by comprising:
the experimental data acquisition module is used for acquiring a plurality of groups of water wave experimental data;
the convolutional neural network construction module is used for constructing a convolutional neural network;
a first parameter obtaining module, configured to, for each group of water ripple experimental data in the multiple groups of water ripple experimental data, use the water ripple experimental data as a test set, input other groups of water ripple experimental data in the multiple groups of water ripple experimental data except the test set into the convolutional neural network for training, so that a first recognition accuracy of the trained convolutional neural network when recognizing each group of water ripple experimental data in the other groups of water ripple experimental data reaches a first set value, and obtain a first parameter of the trained convolutional neural network;
the identification accuracy rate calculation module is used for carrying out first configuration on the trained convolutional neural network based on the first parameter, and inputting the test set into the convolutional neural network which completes the first configuration to obtain second identification accuracy rate obtained by identifying the test set by the convolutional neural network which completes the first configuration;
the parameter caching module is used for judging whether the second identification accuracy reaches a second set value or not; if so, caching the first parameter; if not, adjusting the first parameter to obtain a second parameter, performing second configuration on the convolutional neural network which completes the first configuration based on the second parameter so that a second identification accuracy rate of the convolutional neural network which completes the second configuration for identifying the test set reaches a second set value, and caching the first parameter and the second parameter according to a set rule;
the convolutional neural network configuration module is used for calculating the average value of the parameters obtained by caching as a third parameter and carrying out third configuration on the constructed convolutional neural network according to the third parameter;
and the identification module is used for acquiring current water ripple data and inputting the current water ripple data into the convolution neural network which completes the third configuration so as to realize the identification of the water ripple fluctuation situation.
6. The hydrologic flow fluctuation situation recognition system of claim 5, wherein each set of the plurality of sets of water ripple experimental data comprises n consecutive three-dimensional coordinates of at least one first pixel within a preset time period;
the first recognition accuracy is calculated by the following method:
wherein,
pre1is the first recognition accuracy;
n1predicting the three-dimensional coordinates of at least one first pixel in each group of water ripple experimental data at the next moment of each three-dimensional coordinate in a preset time period by adopting a trained convolutional neural network and predicting the correct time sum;
in order to sum the times of predicting the three-dimensional coordinates of at least one first pixel in each group of water ripple experimental data at the next moment of each three-dimensional coordinate within a preset time period by adopting the trained convolutional neural network,
the second recognition accuracy is calculated by the following method:
pre2a second recognition accuracy;
n2predicting three-dimensional coordinates of at least one first pixel in the water wave experimental data corresponding to the test set at the next moment of each three-dimensional coordinate in a preset time period by adopting a convolution neural network which is configured for the first time, and predicting the correct time sum;
in order to adopt the convolution neural network which completes the first configuration to predict the three-dimensional coordinates of at least one first pixel in the water wave experimental data corresponding to the test set at the next moment of each three-dimensional coordinate in the preset time period,
7. the hydrologic flow fluctuation situation recognition system of claim 5, wherein the multiple groups of water ripple experimental data are m groups, and the parameter caching module caches the first parameter and the second parameter according to a set rule by:
carrying out weighted summation on the first parameter and the second parameter according to a set formula, and caching a weighted summation value; the set formula is as follows:
wherein,
sumitaking the ith group of water ripple experimental data as a weighted summation value when the test set is formed, wherein i belongs to [1, m ∈]∩i∈Z；
para1-iAs a test for the ith set of water ripple experimental dataA first parameter of the convolutional neural network which completes training correspondingly in set time;
para2-iwhen the ith group of water wave experimental data is used as a test set and the corresponding convolutional neural network which finishes training identifies the ith group of water wave experimental data to obtain a second identification accuracy rate which does not reach a second set value, the first parameter para is subjected to the first parameter para1-iAnd adjusting the obtained second parameter.
8. The hydrologic flow fluctuation situation recognition system of claim 5, wherein the current water wave data comprises current-time three-dimensional coordinates of at least one second pixel, and the recognition module inputs the current water wave data into a convolution neural network configured for the third time to realize recognition of the water wave fluctuation situation by:
inputting the current-time three-dimensional coordinate into the convolution neural network which completes the third configuration so that the convolution neural network which completes the third configuration identifies the next-time three-dimensional coordinate of the at least one second pixel.
9. An electronic device comprising a memory, a processor and a computer program stored on the memory and operable on the processor, wherein the processor implements the hydrologic flow fluctuation situation recognition method according to any one of claims 1 to 4 when executing the computer program.
10. A computer-readable storage medium, wherein the computer-readable storage medium comprises a computer program, and the computer program controls an electronic device where the computer-readable storage medium is located to execute the hydrologic flow fluctuation situation recognition method according to any one of claims 1 to 4 when running.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201910466916.9A CN110222823B (en) | 2019-05-31 | 2019-05-31 | Hydrological flow fluctuation situation identification method and system |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201910466916.9A CN110222823B (en) | 2019-05-31 | 2019-05-31 | Hydrological flow fluctuation situation identification method and system |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN110222823A true CN110222823A (en) | 2019-09-10 |
| CN110222823B CN110222823B (en) | 2022-12-30 |
Family
ID=67818950
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201910466916.9A Active CN110222823B (en) | 2019-05-31 | 2019-05-31 | Hydrological flow fluctuation situation identification method and system |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN110222823B (en) |
Citations (9)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20140247973A1 (en) * | 2013-02-28 | 2014-09-04 | Auxogyn, Inc. | Apparatus, method, and system for automated, non-invasive cell activity tracking |
| CN105095658A (en) * | 2015-07-21 | 2015-11-25 | 武汉大学 | Recognition method and system for fluctuation situation of hydrologic flow |
| CN106156734A (en) * | 2016-06-28 | 2016-11-23 | 浙江工业大学 | Water flow velocity measurement method based on convolutional neural network image recognition |
| US20180082212A1 (en) * | 2016-09-20 | 2018-03-22 | Intel Corporation | Optimizing machine learning running time |
| CN108304668A (en) * | 2018-02-11 | 2018-07-20 | 河海大学 | A kind of Forecasting Flood method of combination hydrologic process data and history priori data |
| US20180276527A1 (en) * | 2017-03-23 | 2018-09-27 | Hitachi, Ltd. | Processing Method Using Convolutional Neural Network, Convolutional Neural Network Learning Method, and Processing Device Including Convolutional Neural Network |
| CN109063656A (en) * | 2018-08-08 | 2018-12-21 | 厦门市美亚柏科信息股份有限公司 | A kind of method and device carrying out face inquiry using multiple face engines |
| CN109508810A (en) * | 2018-09-29 | 2019-03-22 | 天津大学 | A kind of system based on realization monthly average hydrology volume forecasting |
| CN109688611A (en) * | 2018-12-20 | 2019-04-26 | 惠州Tcl移动通信有限公司 | A kind of configuration method, device, terminal and the storage medium of frequency range parameter |
-
2019
- 2019-05-31 CN CN201910466916.9A patent/CN110222823B/en active Active
Patent Citations (9)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20140247973A1 (en) * | 2013-02-28 | 2014-09-04 | Auxogyn, Inc. | Apparatus, method, and system for automated, non-invasive cell activity tracking |
| CN105095658A (en) * | 2015-07-21 | 2015-11-25 | 武汉大学 | Recognition method and system for fluctuation situation of hydrologic flow |
| CN106156734A (en) * | 2016-06-28 | 2016-11-23 | 浙江工业大学 | Water flow velocity measurement method based on convolutional neural network image recognition |
| US20180082212A1 (en) * | 2016-09-20 | 2018-03-22 | Intel Corporation | Optimizing machine learning running time |
| US20180276527A1 (en) * | 2017-03-23 | 2018-09-27 | Hitachi, Ltd. | Processing Method Using Convolutional Neural Network, Convolutional Neural Network Learning Method, and Processing Device Including Convolutional Neural Network |
| CN108304668A (en) * | 2018-02-11 | 2018-07-20 | 河海大学 | A kind of Forecasting Flood method of combination hydrologic process data and history priori data |
| CN109063656A (en) * | 2018-08-08 | 2018-12-21 | 厦门市美亚柏科信息股份有限公司 | A kind of method and device carrying out face inquiry using multiple face engines |
| CN109508810A (en) * | 2018-09-29 | 2019-03-22 | 天津大学 | A kind of system based on realization monthly average hydrology volume forecasting |
| CN109688611A (en) * | 2018-12-20 | 2019-04-26 | 惠州Tcl移动通信有限公司 | A kind of configuration method, device, terminal and the storage medium of frequency range parameter |
Non-Patent Citations (2)
| Title |
|---|
| SEUNG SOOK SHIN ET AL.: "Measurement of Flow Velocity Using Video Image of Spherical", 《12TH INTERNATIONAL CONFERENCE ON HYDROINFORMATICS, HIC 2016》 * |
| 张振 等: "河流水面成像测速中的水流示踪物综述", 《水利水电科技进展》 * |
Also Published As
| Publication number | Publication date |
|---|---|
| CN110222823B (en) | 2022-12-30 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN111401516B (en) | Searching method for neural network channel parameters and related equipment | |
| CN110807109A (en) | Data enhancement strategy generation method, data enhancement method and device | |
| US11586903B2 (en) | Method and system of controlling computing operations based on early-stop in deep neural network | |
| CN105447498A (en) | A client device configured with a neural network, a system and a server system | |
| CN106796533B (en) | System and method for adaptively selecting execution mode | |
| JP7461344B2 (en) | Convolution processing method, device, and storage medium for convolutional neural network | |
| CN110211119B (en) | Image quality evaluation method and device, electronic equipment and readable storage medium | |
| CN110824587B (en) | Image prediction method, image prediction device, computer equipment and storage medium | |
| CN110263628B (en) | Obstacle detection method, obstacle detection device, electronic apparatus, and storage medium | |
| CN105718848B (en) | Quality evaluation method and device for fingerprint image | |
| CN110363098B (en) | Violent behavior early warning method and device, readable storage medium and terminal equipment | |
| CN111898578A (en) | Crowd density acquisition method and device, electronic equipment and computer program | |
| CN109663359B (en) | Game intelligent agent training optimization method and device, terminal device and storage medium | |
| JP2017068608A (en) | Arithmetic unit, method and program | |
| CN111028182A (en) | Image sharpening method and device, electronic equipment and computer-readable storage medium | |
| CN111159481B (en) | Edge prediction method and device for graph data and terminal equipment | |
| CN114565092A (en) | Neural network structure determining method and device | |
| CN110490058B (en) | Training method, device and system of pedestrian detection model and computer readable medium | |
| CN115296984A (en) | Method, device, equipment and storage medium for detecting abnormal network nodes | |
| CN107273692B (en) | Distributed fusion method of random set theory with limited sensor sensing capability | |
| CN110222823B (en) | Hydrological flow fluctuation situation identification method and system | |
| CN112800813B (en) | Target identification method and device | |
| CN109918567B (en) | Travel mode recommendation method and device | |
| CN111861931A (en) | Model training method, image enhancement method, model training device, image enhancement device, electronic equipment and storage medium | |
| CN114998707B (en) | Attack method and device for evaluating robustness of target detection 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 |