CN111428419A - Suspended sediment concentration prediction method and device, computer equipment and storage medium - Google Patents
Suspended sediment concentration prediction method and device, computer equipment and storage medium Download PDFInfo
- Publication number
- CN111428419A CN111428419A CN202010131794.0A CN202010131794A CN111428419A CN 111428419 A CN111428419 A CN 111428419A CN 202010131794 A CN202010131794 A CN 202010131794A CN 111428419 A CN111428419 A CN 111428419A
- Authority
- CN
- China
- Prior art keywords
- concentration
- suspended sediment
- prediction model
- prediction
- concentration data
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Pending
Links
- 239000013049 sediment Substances 0.000 title claims abstract description 255
- 238000000034 method Methods 0.000 title claims abstract description 57
- 238000012549 training Methods 0.000 claims abstract description 82
- 238000004422 calculation algorithm Methods 0.000 claims abstract description 29
- 239000013598 vector Substances 0.000 claims description 33
- 238000012360 testing method Methods 0.000 claims description 31
- 230000015654 memory Effects 0.000 claims description 26
- 238000004590 computer program Methods 0.000 claims description 25
- 230000000007 visual effect Effects 0.000 claims description 11
- 238000007637 random forest analysis Methods 0.000 claims description 10
- 230000008569 process Effects 0.000 description 11
- 238000010586 diagram Methods 0.000 description 5
- 238000011160 research Methods 0.000 description 5
- 238000005516 engineering process Methods 0.000 description 4
- 238000004364 calculation method Methods 0.000 description 3
- 125000004122 cyclic group Chemical group 0.000 description 3
- 230000000694 effects Effects 0.000 description 3
- 239000003344 environmental pollutant Substances 0.000 description 3
- 231100000719 pollutant Toxicity 0.000 description 3
- 238000012545 processing Methods 0.000 description 3
- 238000004088 simulation Methods 0.000 description 3
- XLYOFNOQVPJJNP-UHFFFAOYSA-N water Substances O XLYOFNOQVPJJNP-UHFFFAOYSA-N 0.000 description 3
- 230000008859 change Effects 0.000 description 2
- 238000004140 cleaning Methods 0.000 description 2
- 238000010276 construction Methods 0.000 description 2
- 230000006870 function Effects 0.000 description 2
- 238000012544 monitoring process Methods 0.000 description 2
- 230000003287 optical effect Effects 0.000 description 2
- 239000002245 particle Substances 0.000 description 2
- 150000003839 salts Chemical class 0.000 description 2
- 238000001179 sorption measurement Methods 0.000 description 2
- 230000004913 activation Effects 0.000 description 1
- 238000013528 artificial neural network Methods 0.000 description 1
- 230000009286 beneficial effect Effects 0.000 description 1
- 239000000969 carrier Substances 0.000 description 1
- 230000010354 integration Effects 0.000 description 1
- 230000003993 interaction Effects 0.000 description 1
- 238000010801 machine learning Methods 0.000 description 1
- 239000000463 material Substances 0.000 description 1
- 230000007246 mechanism Effects 0.000 description 1
- 230000005012 migration Effects 0.000 description 1
- 238000013508 migration Methods 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 235000015097 nutrients Nutrition 0.000 description 1
- 230000000306 recurrent effect Effects 0.000 description 1
- 238000010187 selection method Methods 0.000 description 1
- 230000035945 sensitivity Effects 0.000 description 1
- 230000006403 short-term memory Effects 0.000 description 1
- 230000003068 static effect Effects 0.000 description 1
Images
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
- G06N3/044—Recurrent networks, e.g. Hopfield 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
- G06N3/04—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)
- Data Mining & Analysis (AREA)
- General Health & Medical Sciences (AREA)
- Biomedical Technology (AREA)
- Biophysics (AREA)
- Computational Linguistics (AREA)
- Life Sciences & Earth Sciences (AREA)
- Evolutionary Computation (AREA)
- Artificial Intelligence (AREA)
- Molecular Biology (AREA)
- Computing Systems (AREA)
- General Engineering & Computer Science (AREA)
- General Physics & Mathematics (AREA)
- Mathematical Physics (AREA)
- Software Systems (AREA)
- Health & Medical Sciences (AREA)
- Management, Administration, Business Operations System, And Electronic Commerce (AREA)
Abstract
The method comprises the steps of obtaining suspended sediment concentration data to form a concentration data sample set, randomly extracting a plurality of concentration data samples from the concentration data sample set to form a concentration data training set, generating a prediction model based on an L STM algorithm, training the prediction model by using the concentration data training set, storing model parameters of the prediction model when a training result meets a set condition to obtain a suspended sediment concentration prediction model, and predicting the suspended sediment concentration according to the suspended sediment concentration prediction model.
Description
Technical Field
The application relates to the technical field of suspended sediment, in particular to a method and a device for predicting the concentration of suspended sediment, computer equipment and a storage medium.
Background
The suspended sediment at the river mouth has important influence on the river mouth and the surrounding environment, and the content of the suspended sediment directly influences the optical properties of water transparency, turbidity, water color and the like. The suspended silt particles have an adsorption effect, wherein fine silt particles are important carriers of various nutrient salts and pollutants, and the pollutants and the silt have an adsorption effect, so that the transportation of the silt has an important effect on the migration and circulation of the pollutants, and the offshore marine organism geochemical process is influenced. Therefore, accurate and timely prediction of the concentration of the suspended sediment has important scientific significance for protecting and treating estuary and coast.
The existing suspended sediment prediction method is used for predicting the suspended sediment concentration based on numerical simulation or a comprehensive numerical model combined with a data assimilation technology, and the method has the following steps: according to the characteristics of the research area, selecting a corresponding numerical model, setting corresponding initial time points and boundary conditions, discretizing basic equations (mass, momentum, heat and salt conservation equations, energy balance equations and the like) into differential equations according to certain step lengths, and solving an equation set by using a computer to simulate the dynamic field of the research area. Finally, by combining the characteristics of the research area and the past experience, debugging corresponding parameters (such as vertical layering and bottom friction coefficient in the three-dimensional model) of the numerical model to obtain a prediction model, and predicting the suspended sediment in the research area by using the prediction model.
However, in the existing suspended sediment prediction method, the numerical simulation is the time-space discrete processing of the marine power process, and the discrete processing starts from the initial time point, so that the modeling process is long, and the error of the prediction model is easily large due to the manual setting of the initial time point and the boundary condition.
Therefore, the current suspended sediment prediction method has the problems of long modeling time of a prediction model and large prediction error.
Disclosure of Invention
Therefore, it is necessary to provide a suspended sediment concentration prediction method, a suspended sediment concentration prediction device, a computer device and a storage medium for solving the technical problems of long modeling time and large prediction error of the prediction model.
A method of predicting suspended sediment concentration, the method comprising:
acquiring suspended sediment concentration data to form a concentration data sample set;
randomly extracting a plurality of concentration data samples from the concentration data sample set to form a concentration data training set;
generating a prediction model based on L STM algorithm, and training the prediction model by adopting the concentration data training set;
when the training result meets the set condition, storing the model parameters of the prediction model to obtain a suspended sediment concentration prediction model;
and predicting the concentration of the suspended sediment according to the suspended sediment concentration prediction model.
In one embodiment, before forming the concentration data sample set, the method further includes:
determining characteristic variables influencing the concentration of the suspended sediment;
and acquiring characteristic variable data, and fusing the characteristic variable data and the suspended sediment concentration data to form the concentration data sample set.
In one embodiment, the determining the characteristic variables affecting the concentration of suspended sediment comprises:
determining a plurality of power factors affecting the concentration of suspended sediment;
determining the influence importance of each power factor on the concentration of the suspended sediment based on a random forest algorithm;
and selecting a plurality of dominant power factors from the plurality of power factors as the characteristic variables according to the influence importance.
In one embodiment, after the training the prediction model with the training set of concentration data, the method further includes:
adjusting model parameters of the prediction model to obtain a plurality of candidate suspended sediment concentration prediction models;
determining the prediction precision of each candidate suspended sediment concentration prediction model;
and taking the candidate suspended sediment concentration prediction model corresponding to the prediction precision meeting the preset conditions as the suspended sediment concentration prediction model.
In one embodiment, the determining the prediction accuracy of each candidate suspended sediment concentration prediction model includes:
determining an input vector and a target output concentration of the candidate suspended sediment concentration prediction model;
inputting the input vector into the candidate suspended sediment concentration prediction model to obtain the predicted output concentration of the candidate suspended sediment concentration prediction model;
calculating an error value of the predicted output concentration relative to the target output concentration, and determining the prediction accuracy according to the error value.
In one embodiment, before determining the input vector and the target output concentration of the candidate suspended sediment concentration prediction model, the method further includes:
randomly selecting a plurality of concentration data samples from the concentration data sample set to form a concentration data test set;
selecting the input vector and the target output concentration from the concentration data test set.
In one embodiment, the method further comprises:
and constructing a visual prediction platform to visualize the prediction result of the suspended sediment concentration prediction model.
An apparatus for predicting suspended sediment concentration, the apparatus comprising:
the sample set forming module is used for obtaining suspended sediment concentration data to form a concentration data sample set;
the training set forming module is used for randomly extracting a plurality of concentration data samples from the concentration data sample set to form a concentration data training set;
the model training module is used for generating a prediction model based on L STM algorithm and training the prediction model by adopting the concentration data training set;
the model determining module is used for storing the model parameters of the prediction model when the training result meets the set condition to obtain a suspended sediment concentration prediction model;
and the prediction module is used for predicting the concentration of the suspended sediment according to the suspended sediment concentration prediction model.
A computer device comprising a memory and a processor, the memory storing a computer program, the processor implementing the following steps when executing the computer program:
acquiring suspended sediment concentration data to form a concentration data sample set;
randomly extracting a plurality of concentration data samples from the concentration data sample set to form a concentration data training set;
generating a prediction model based on L STM algorithm, and training the prediction model by adopting the concentration data training set;
when the training result meets the set condition, storing the model parameters of the prediction model to obtain a suspended sediment concentration prediction model;
and predicting the concentration of the suspended sediment according to the suspended sediment concentration prediction model.
A computer-readable storage medium, on which a computer program is stored which, when executed by a processor, carries out the steps of:
acquiring suspended sediment concentration data to form a concentration data sample set;
randomly extracting a plurality of concentration data samples from the concentration data sample set to form a concentration data training set;
generating a prediction model based on L STM algorithm, and training the prediction model by adopting the concentration data training set;
when the training result meets the set condition, storing the model parameters of the prediction model to obtain a suspended sediment concentration prediction model;
and predicting the concentration of the suspended sediment according to the suspended sediment concentration prediction model.
According to the suspended sediment concentration prediction method, the suspended sediment concentration prediction device, the computer equipment and the storage medium, the prediction model based on the L STM algorithm is generated, the concentration data training set extracted from the concentration data sample set is adopted to train the prediction model, when the training result meets the preset condition, the model parameters are stored, the suspended sediment concentration prediction model is obtained, the suspended sediment concentration prediction model can be used for rapidly predicting the suspended sediment concentration, the initial time point and the process of dispersing from the initial time point are not needed to be set, the prediction efficiency of the suspended sediment concentration is improved, in addition, the requirements of the building and debugging process of the prediction model based on the L STM algorithm on the professional background of a user are low, and the objective accuracy of the prediction result is improved.
Drawings
FIG. 1 is a schematic flow chart of a method for predicting the concentration of suspended sediment in one embodiment;
FIG. 2 is a schematic flow chart of the steps of determining characteristic variables that affect the concentration of suspended silt in one embodiment;
FIG. 3 is a schematic flow chart of a method for predicting suspended sediment concentration in another embodiment;
FIG. 4 is a schematic diagram of an L STM model memory module in one embodiment;
FIG. 5 is a block diagram showing the structure of a suspended sediment concentration predicting apparatus according to an embodiment;
FIG. 6 is a diagram illustrating an internal structure of a computer device according to an embodiment.
Detailed Description
In order to make the objects, technical solutions and advantages of the present application more apparent, the present application is described in further detail below with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are merely illustrative of the present application and are not intended to limit the present application.
In an embodiment, as shown in fig. 1, a suspended sediment concentration prediction method is provided, and this embodiment is exemplified by applying the method to a server, and it is to be understood that the method may also be applied to a terminal, and may also be applied to a system including a terminal and a server, and is implemented by interaction between the terminal and the server. The terminal can be, but is not limited to, various personal computers, notebook computers, smart phones, tablet computers and portable wearable devices, and the server can be implemented by an independent server or a server cluster formed by a plurality of servers. In this embodiment, the method includes the steps of:
and S102, acquiring suspended sediment concentration data to form a concentration data sample set.
In the concrete implementation, the server can obtain the concentration data of the suspended sediment through various observation devices such as a buoy, a satellite and the like, and the related data of the influence factors are required to be obtained in consideration of the influence of factors such as wind stress, tide, runoff and meteorological elements on the concentration of the suspended sediment, and are fused with the concentration data of the suspended sediment to form a suspended sediment concentration sample in the form of (time, suspended sediment concentration, wind stress, tide value and runoff value), and further form a suspended sediment concentration data sample set.
In practical application, when suspended sediment concentration data from different observation devices are obtained in the same time period, the target suspended sediment concentration data to be extracted can be determined according to preset selection conditions. The preset selection condition may be: preferentially selecting suspended sediment concentration data of observation equipment capable of directly measuring the suspended sediment concentration; if a plurality of different observation devices can directly measure the concentration of the suspended sediment in the same time period, the suspended sediment concentration data of the observation device with longer observation time can be preferentially selected as the target suspended sediment concentration data.
And step S104, randomly extracting a plurality of concentration data samples from the concentration data sample set to form a concentration data training set.
In the concrete realization, the server can classify the concentration data sample set into after obtaining the concentration data sample set of suspended sediment: a training set of concentration data and a testing set of concentration data. The concentration data training set is used for training the prediction model, and the concentration data testing set is used for testing the prediction accuracy of the prediction model. More specifically, the acquisition of the concentration data sample set may be: and randomly selecting a plurality of concentration data samples from the concentration data sample set to form a concentration data sample set. The concentration data sample set can also be randomly divided into a concentration data training set and a concentration data testing set according to a certain proportion.
And S106, generating a prediction model based on an L STM algorithm, and training the prediction model by adopting a concentration data training set.
In the concrete implementation, L STM (L ong Short-Term Memory) represents a long-Short Term Memory network, is a time recurrent neural network and is used for processing and predicting an important event algorithm with relatively long interval and delay in a time sequence, because the acquired suspended sediment concentration data has the characteristics of randomness, periodicity and nonlinearity, a prediction model is constructed by adopting a L STM algorithm, so that the accuracy of a prediction result can be greatly improved, the L STM prediction model is of a three-layer model structure consisting of an input layer, a cyclic hidden layer and an output layer, after the prediction model is generated based on a L algorithm, a concentration data training set can be adopted to train the prediction model, and a random selection method is adopted, so that the concentration data test set and the concentration data training set have certain randomness, the problem of overfitting training the prediction model and reducing the accuracy of prediction can be avoided.
And S108, when the training result meets the set condition, storing the model parameters of the prediction model to obtain the suspended sediment concentration prediction model.
In specific implementation, the set condition may be a set training frequency or training precision, that is, when training of the prediction model reaches the set frequency, the model parameters are saved; or when the training result of the prediction model reaches the set training precision, namely the error of the training result is within the allowable range, the model parameters of the prediction model are stored, and the obtained prediction model is used as the suspended sediment concentration prediction model.
And S110, predicting the concentration of the suspended sediment according to the suspended sediment concentration prediction model.
In the concrete implementation, after the suspended sediment concentration prediction model is obtained, the suspended sediment concentration prediction model can be adopted to predict the concentration of suspended sediment. More specifically, the obtained suspended sediment concentration data is used as an input vector, input into a circulating hidden layer of a suspended sediment concentration prediction model through an input layer of the suspended sediment concentration prediction model, and output to an output layer of the suspended sediment concentration prediction model after being processed by the circulating hidden layer, so that the predicted concentration of the suspended sediment is obtained.
In the method for predicting the concentration of the suspended sediment, a prediction model based on the L STM algorithm is generated, the prediction model is trained by a concentration data training set extracted from a concentration data sample set, and when a training result meets a preset condition, model parameters are stored to obtain the suspended sediment concentration prediction model.
In one embodiment, after the step of obtaining the suspended sediment concentration data, the method further includes: determining characteristic variables influencing the concentration of the suspended sediment; and acquiring characteristic variable data, and fusing the characteristic variable data and the suspended sediment concentration data to form a concentration data sample set.
In the concrete implementation, after the suspended sediment concentration data is obtained through the monitoring equipment, the suspended sediment is considered to be influenced by factors such as wind stress, ocean current, tide, runoff and meteorological elements, therefore, the data corresponding to the influencing factors which have obvious influence on the suspended sediment concentration can be obtained through the observation equipment and used as characteristic variable data, after the characteristic variable data is obtained, the characteristic variable data and the suspended sediment concentration data are fused, and the fused data are used as a concentration data sample set of the suspended sediment.
In the embodiment, the influence of other influencing factors on the concentration of the suspended sediment is fully considered, and the relevant data of the characteristic variable influencing the concentration of the suspended sediment is fused with the concentration data of the suspended sediment to form a concentration data sample set. In addition, factors influencing the concentration of the suspended sediment are fused with the concentration data and used as input vectors to construct a prediction model, so that the prediction model is not limited by a research area, and the universality of the prediction model is improved.
In one embodiment, as shown in fig. 2, the step of determining the characteristic variables affecting the concentration of suspended sediment further comprises:
step S202, a plurality of power factors influencing the concentration of the suspended sediment are determined.
In this step, the power factor that influences suspended sediment includes: wind stress, ocean current, tide, runoff, air pressure, rainfall and other meteorological elements. The power factor can be determined based on characteristics of the material transportation process in estuary dynamics and physical oceanography, and factors influencing the concentration and distribution of suspended sediment are comprehensively considered.
And S204, determining the influence importance of each power factor on the concentration of the suspended sediment based on a random forest algorithm.
In the step, a random forest (random forest) algorithm belongs to a machine learning algorithm, and the influence importance of each power factor on the concentration of suspended sediment is determined through the random forest algorithm, so that the importance sequence of each power factor is obtained.
And S206, selecting a plurality of dominant power factors from the plurality of power factors as characteristic variables according to the influence importance.
In the step, after the importance degrees of the power factors are obtained and ranked, a plurality of power factors with larger influence importance degrees are selected from the plurality of determined power factors according to the influence importance degrees of the power factors. Specifically, the first power factors can be selected in the order from large to small according to the numerical value of the influence importance degree, and are used as the dominant power factors to form the characteristic variable. The number of the dominant power factors is not less than two, and the sum of the influence importance degrees of the selected dominant power factors is not less than 80% of the sum of the influence importance degrees of each power factor, so that the effectiveness of the selected dominant power factors is ensured.
In the embodiment, the influence of the power factors on the distribution and the change of the suspended sediment concentration is considered, the influence importance of each power factor influencing the suspended sediment concentration is calculated and evaluated based on a random forest algorithm, the contribution mechanism and the difference of the power factors influencing the suspended sediment concentration change are determined, and the dominant power factor is selected to facilitate further construction of the prediction model, so that the building process of the prediction model is simplified, the calculation cost of the model is reduced, the universality of the model is improved, and the flexibility and the accuracy of the suspended sediment concentration prediction model are improved.
In an embodiment, after the step of training the prediction model by using the training set of concentration data, the method further includes: adjusting model parameters of the prediction model to obtain a plurality of candidate suspended sediment concentration prediction models; determining the prediction precision of each candidate suspended sediment concentration prediction model; and taking the candidate suspended sediment concentration prediction model corresponding to the prediction precision meeting the preset conditions as a suspended sediment concentration prediction model.
In specific implementation, the model parameters may be the number of nodes in the hidden layer, the number of cases for predicting model learning, the time step length, the training times and the like. When the prediction model is trained by adopting the concentration data training set, model parameters of the prediction model can be adjusted through a sensitivity test to obtain prediction results of different parameters, a group of parameters with the optimal prediction results are selected by comparing and analyzing the relation between parameter changes and the corresponding prediction results, and the prediction model formed by the group of parameters is used as a suspended sediment concentration prediction model. And each model parameter is changed to correspond to a new prediction model, so that when the model parameters of the prediction models are adjusted, a plurality of prediction models are generated and serve as candidate suspended sediment concentration prediction models. And determining the prediction precision of each candidate suspended sediment concentration prediction model one by one, and taking the candidate suspended sediment concentration prediction model corresponding to the prediction precision meeting the preset conditions as a final suspended sediment concentration prediction model. The preset condition may be that the prediction accuracy value is the highest, for example, the prediction accuracies of the candidate suspended sediment concentration prediction models are sorted, and the candidate suspended sediment concentration prediction model corresponding to the prediction accuracy value with the highest value is used as the suspended sediment concentration prediction model.
In this embodiment, a plurality of candidate suspended sediment concentration prediction models are obtained by adjusting parameters of the prediction model, and an optimal suspended sediment concentration prediction model is further determined according to the prediction accuracy of each candidate suspended sediment concentration prediction model, so as to improve the accuracy of suspended sediment concentration prediction.
In one embodiment, the step of determining the prediction accuracy of each candidate suspended sediment concentration prediction model includes: determining an input vector and a target output concentration of a candidate suspended sediment concentration prediction model; inputting the input vector into a candidate suspended sediment concentration prediction model to obtain the predicted output concentration of the candidate suspended sediment concentration prediction model; an error value of the predicted output concentration relative to the target output concentration is calculated, and the prediction accuracy is determined based on the error value.
In specific implementation, the input vector and the target output concentration are both obtained actual observation data of the suspended sediment, and can be selected from a concentration data sample set. After the input vector and the target output concentration are determined, the input vector is input into the circulating hiding layer through an input layer of the candidate suspended sediment concentration prediction model, and after the input vector is processed in the circulating hiding layer, a predicted value of the suspended sediment concentration can be output through an output layer and recorded as the predicted output concentration. And finally, calculating an error value of the predicted output concentration relative to the target output concentration, and determining the prediction precision of the candidate suspended sediment concentration prediction model according to the error value. The error value of the predicted output concentration relative to the target output concentration may be a mean square error value or an absolute error value.
In practical application, in order to improve the accuracy of judging the precision of the prediction model, when the prediction precision of the candidate suspended sediment concentration prediction model is determined, multiple groups of input vectors and target output concentrations can be selected, and the prediction precision of the candidate suspended sediment concentration prediction model is determined by calculating an average error value.
In this embodiment, the prediction accuracy is determined by calculating the error value of the candidate suspended sediment concentration prediction model, and then the optimal suspended sediment concentration prediction model, that is, the suspended sediment concentration prediction model with the highest prediction accuracy, can be selected from the candidate suspended sediment concentration prediction models according to the prediction accuracy to predict the concentration of suspended sediment, so that the prediction accuracy is improved.
In an embodiment, before determining the input vector and the target output concentration of the candidate suspended sediment concentration prediction model, the method further includes: randomly selecting a plurality of concentration data samples from the concentration data sample set to form a concentration data test set; an input vector and a target output concentration are selected from the concentration data test set.
In the specific implementation, after the concentration data sample set is obtained, a plurality of concentration data samples are randomly selected from the concentration data sample set to form a concentration data test set. In addition, the concentration data test set can also be composed of concentration data samples of the concentration data sample set except the concentration data training set. And after the concentration data test set is obtained, testing the prediction result of the candidate suspended sediment concentration prediction model by adopting the concentration data test set. More specifically, a plurality of input vectors for testing and target output concentrations can be selected from the concentration data test set, the target output concentrations are compared with actual predicted concentrations, prediction errors are obtained through calculation, and then prediction accuracy of the candidate suspended sediment concentration prediction model is determined.
In this embodiment, a concentration data test set is formed, an input vector and a target output concentration are selected from the concentration data test set, and a candidate suspended sediment concentration prediction model is tested, so that a prediction error of the candidate suspended sediment concentration prediction model is obtained, the prediction accuracy of the candidate suspended sediment concentration prediction model is determined, an optimal suspended sediment concentration prediction model is further selected from the candidate suspended sediment concentration prediction model, the suspended sediment concentration is predicted, and the accuracy of a prediction result is improved.
In one embodiment, further comprising: and constructing a visual prediction platform to visualize the prediction result of the suspended sediment concentration prediction model.
In the concrete implementation, after the prediction result of the concentration of the suspended sediment is obtained by adopting the suspended sediment concentration prediction model, a visual prediction platform can be constructed to visualize the predicted concentration of the suspended sediment so as to observe the prediction result of the concentration of the suspended sediment visually. And moreover, according to the demand, the prediction result of each time period or the variation trend of the suspended sediment concentration can be checked in the visual prediction platform, so that treatment measures for river mouths or coasts and the like can be further formulated according to the variation trend. In addition, the 3S integration technology, the knowledge map and component technology, the GIS technology and the like can be applied to a visual prediction platform to display three-dimensional information of a river mouth or a coastal environment.
In the embodiment, the visual prediction platform is constructed, so that the prediction result of the suspended sediment concentration can be visually presented, a user can conveniently analyze the prediction result, decision support is provided for the management of estuaries or coasts, and the effective utilization of water resources is realized.
In an embodiment, in order to more clearly illustrate the technical solution provided by the embodiment of the present application, the solution will be described below with reference to fig. 3, where fig. 3 is a schematic flow chart of a suspended sediment concentration prediction method in an application example, and the specific flow chart of the method is as follows:
(1) and acquiring suspended sediment concentration data, and extracting, inverting and cleaning the suspended sediment concentration data to form a suspended sediment concentration data set. The inversion represents the process of obtaining the suspended sediment concentration data through inversion of the obtained other data and data when the monitoring equipment cannot directly obtain the suspended sediment concentration data; cleaning means to remove invalid and duplicate values from the observation.
(2) Determining power factors (such as wind stress, tide, runoff, meteorological elements and the like) influencing the concentration of the suspended sediment, selecting a plurality of dominant power factors from the power factors, acquiring related data of the dominant power factors, and fusing the related data and the suspended sediment concentration data to form a suspended sediment concentration data sample set.
(3) The method comprises the steps of establishing a prediction model based on an L STM algorithm, randomly dividing a suspended sediment concentration data sample set into a concentration data training set and a concentration data testing set according to a certain proportion, training the prediction model by using the concentration data training set, testing the prediction model by using the concentration data testing set, and debugging model parameters to obtain the optimal suspended sediment concentration prediction model.
(4) And constructing a visual prediction platform to visualize the prediction result of the suspended sediment concentration.
The L STM model consists of an input layer, a cyclic hiding layer and an output layer, the basic unit of the L STM model hiding layer is a memory module, the memory module comprises a cell state unit (CEC) and 3 special operation units called gates, the memory module comprises 3 gate structures of an input gate, a forgetting gate and an output gate, and the 3 gate structures can control information flow in the memory module, the principle of the L STM model memory module is shown in FIG. 41,X2,…XT) Then the hidden layer state can be obtained by the following cycle training:
f1＝σ(XtUf+St-1Wf+bf)
it＝σ(XtUi+St-1Wi+bi)
ot＝σ(XtUo+St-1Wo+bo)
wherein, Wf、Wi、Wo、WcInput weight matrixes of a forgetting gate, an input gate, an output gate and a CEC respectively; u shapef、Ui、Uo、UcThe cyclic weight matrixes are respectively a forgetting gate, an input gate, an output gate and a CEC; bf、bi、bo、bcDeviation vectors of a forgetting gate, an input gate, an output gate and a CEC are respectively; xtRepresenting the inputs of a first layer L STM model memory modulet、it、ot、Ct、StThe t time forgetting gate, the input gate, the output gate, the CEC and the output of the unit are respectively; st-1Represents the output of the cell at time t-1; ct-1Represents the CEC output at time t-1;
and
respectively representing vector summation and vector dot product operation; σ (-) is a standard sigmoid function; tanh (-) is a hyperbolic tangent activation function.
In the embodiment, the prediction model based on the L STM algorithm is generated, so that the suspended sediment concentration can be rapidly predicted, the prediction efficiency and the prediction precision of the suspended sediment concentration are improved, and the problem that the simulation precision cannot meet the actual requirement easily because a calculation formula of the suspended sediment is in a semi-empirical form when a numerical model is adopted in the traditional method is solved.
It should be understood that although the various steps in the flow charts of fig. 1-3 are shown in order as indicated by the arrows, the steps are not necessarily performed in order as indicated by the arrows. The steps are not performed in the exact order shown and described, and may be performed in other orders, unless explicitly stated otherwise. Moreover, at least some of the steps in fig. 1-3 may include multiple steps or multiple stages, which are not necessarily performed at the same time, but may be performed at different times, which are not necessarily performed in sequence, but may be performed in turn or alternately with other steps or at least some of the other steps.
In one embodiment, as shown in fig. 5, there is provided a suspended sediment concentration prediction apparatus comprising: a sample set forming module 502, a training set forming module 504, a model training module 506, a model determining module 508, and a prediction module 510, wherein:
a sample set forming module 502, configured to obtain suspended sediment concentration data to form a concentration data sample set;
a training set forming module 504, configured to randomly extract a plurality of concentration data samples from the concentration data sample set to form a concentration data training set;
the model training module 506 is used for generating a prediction model based on L STM algorithm and training the prediction model by adopting a concentration data training set;
the model determining module 508 is configured to, when the training result meets the set condition, save the model parameters of the prediction model to obtain a suspended sediment concentration prediction model;
and the prediction module 510 is configured to predict the concentration of the suspended sediment according to the suspended sediment concentration prediction model.
In one embodiment, the above apparatus further comprises:
a variable determination module for determining characteristic variables affecting the concentration of suspended sediment
And the data acquisition module is used for acquiring the characteristic variable data and fusing the characteristic variable data and the suspended sediment concentration data to form a concentration data sample set.
In an embodiment, the variable determining module is specifically configured to: determining a plurality of power factors affecting the concentration of suspended sediment; determining the influence importance of each power factor on the concentration of the suspended sediment based on a random forest algorithm; and selecting a plurality of dominant power factors from the plurality of power factors as characteristic variables according to the influence importance.
In one embodiment, the above apparatus further comprises:
the parameter adjusting module is used for adjusting model parameters of the prediction model to obtain a plurality of candidate suspended sediment concentration prediction models;
the accuracy determination module is used for determining the prediction accuracy of each candidate suspended sediment concentration prediction model;
and the precision judgment module is used for taking the candidate suspended sediment concentration prediction model corresponding to the prediction precision meeting the preset conditions as the suspended sediment concentration prediction model.
In an embodiment, the accuracy determining module is specifically configured to: determining an input vector and a target output concentration of a candidate suspended sediment concentration prediction model; inputting the input vector into a candidate suspended sediment concentration prediction model to obtain the predicted output concentration of the candidate suspended sediment concentration prediction model; an error value of the predicted output concentration relative to the target output concentration is calculated, and the prediction accuracy is determined based on the error value.
In one embodiment, the above apparatus further comprises:
the test set forming module is used for randomly selecting a plurality of concentration data samples from the concentration data sample set to form a concentration data test set;
and the selection module is used for selecting the input vector and the target output concentration from the concentration data test set.
In one embodiment, the above apparatus further comprises:
and the platform construction module is used for constructing a visual prediction platform so as to visualize the prediction result of the suspended sediment concentration prediction model.
It should be noted that, the suspended sediment concentration prediction apparatus of the present application corresponds to the suspended sediment concentration prediction method of the present application one to one, and the technical features and the beneficial effects thereof described in the embodiments of the suspended sediment concentration prediction method are all applicable to the embodiments of the suspended sediment concentration prediction apparatus, and specific contents may refer to the description in the embodiments of the method of the present application, which is not described herein again, and thus the present application claims.
In addition, all or part of the modules in the suspended sediment concentration prediction device can be realized by software, hardware and a combination thereof. The modules can be embedded in a hardware form or independent from a processor in the computer device, and can also be stored in a memory in the computer device in a software form, so that the processor can call and execute operations corresponding to the modules.
In one embodiment, a computer device is provided, which may be a server, and its internal structure diagram may be as shown in fig. 6. The computer device includes a processor, a memory, and a network interface connected by a system bus. Wherein the processor of the computer device is configured to provide computing and control capabilities. The memory of the computer device comprises a nonvolatile storage medium and an internal memory. The non-volatile storage medium stores an operating system, a computer program, and a database. The internal memory provides an environment for the operation of an operating system and computer programs in the non-volatile storage medium. The database of the computer device is used for storing data generated in the process of predicting the concentration of the suspended sediment. The network interface of the computer device is used for communicating with an external terminal through a network connection. The computer program is executed by a processor to implement a suspended sediment concentration prediction method.
Those skilled in the art will appreciate that the architecture shown in fig. 6 is merely a block diagram of some of the structures associated with the disclosed aspects and is not intended to limit the computing devices to which the disclosed aspects apply, as particular computing devices may include more or less components than those shown, or may combine certain components, or have a different arrangement of components.
In one embodiment, a computer device is provided, comprising a memory and a processor, the memory having a computer program stored therein, the processor implementing the following steps when executing the computer program:
acquiring suspended sediment concentration data to form a concentration data sample set;
randomly extracting a plurality of concentration data samples from the concentration data sample set to form a concentration data training set;
generating a prediction model based on L STM algorithm, and training the prediction model by adopting a concentration data training set;
when the training result meets the set condition, storing the model parameters of the prediction model to obtain a suspended sediment concentration prediction model;
and predicting the concentration of the suspended sediment according to the suspended sediment concentration prediction model.
In one embodiment, the processor, when executing the computer program, further performs the steps of: determining characteristic variables influencing the concentration of the suspended sediment; and acquiring characteristic variable data, and fusing the characteristic variable data and the suspended sediment concentration data to form a concentration data sample set.
In one embodiment, the processor, when executing the computer program, further performs the steps of: determining a plurality of power factors affecting the concentration of suspended sediment; determining the influence importance of each power factor on the concentration of the suspended sediment based on a random forest algorithm; and selecting a plurality of dominant power factors from the plurality of power factors as characteristic variables according to the influence importance.
In one embodiment, the processor, when executing the computer program, further performs the steps of: adjusting model parameters of the prediction model to obtain a plurality of candidate suspended sediment concentration prediction models; determining the prediction precision of each candidate suspended sediment concentration prediction model; and taking the candidate suspended sediment concentration prediction model corresponding to the prediction precision meeting the preset conditions as a suspended sediment concentration prediction model.
In one embodiment, the processor, when executing the computer program, further performs the steps of: determining an input vector and a target output concentration of a candidate suspended sediment concentration prediction model; inputting the input vector into a candidate suspended sediment concentration prediction model to obtain the predicted output concentration of the candidate suspended sediment concentration prediction model; an error value of the predicted output concentration relative to the target output concentration is calculated, and the prediction accuracy is determined based on the error value.
In one embodiment, the processor, when executing the computer program, further performs the steps of: randomly selecting a plurality of concentration data samples from the concentration data sample set to form a concentration data test set; an input vector and a target output concentration are selected from the concentration data test set.
In one embodiment, the processor, when executing the computer program, further performs the steps of: and constructing a visual prediction platform to visualize the prediction result of the suspended sediment concentration prediction model.
In one embodiment, a computer-readable storage medium is provided, having a computer program stored thereon, which when executed by a processor, performs the steps of:
acquiring suspended sediment concentration data to form a concentration data sample set;
randomly extracting a plurality of concentration data samples from the concentration data sample set to form a concentration data training set;
generating a prediction model based on L STM algorithm, and training the prediction model by adopting a concentration data training set;
when the training result meets the set condition, storing the model parameters of the prediction model to obtain a suspended sediment concentration prediction model;
and predicting the concentration of the suspended sediment according to the suspended sediment concentration prediction model.
In one embodiment, the computer program when executed by the processor further performs the steps of: determining characteristic variables influencing the concentration of the suspended sediment; and acquiring characteristic variable data, and fusing the characteristic variable data and the suspended sediment concentration data to form a concentration data sample set.
In one embodiment, the computer program when executed by the processor further performs the steps of: determining a plurality of power factors affecting the concentration of suspended sediment; determining the influence importance of each power factor on the concentration of the suspended sediment based on a random forest algorithm; and selecting a plurality of dominant power factors from the plurality of power factors as characteristic variables according to the influence importance.
In one embodiment, the computer program when executed by the processor further performs the steps of: adjusting model parameters of the prediction model to obtain a plurality of candidate suspended sediment concentration prediction models; determining the prediction precision of each candidate suspended sediment concentration prediction model; and taking the candidate suspended sediment concentration prediction model corresponding to the prediction precision meeting the preset conditions as a suspended sediment concentration prediction model.
In one embodiment, the computer program when executed by the processor further performs the steps of: determining an input vector and a target output concentration of a candidate suspended sediment concentration prediction model; inputting the input vector into a candidate suspended sediment concentration prediction model to obtain the predicted output concentration of the candidate suspended sediment concentration prediction model; an error value of the predicted output concentration relative to the target output concentration is calculated, and the prediction accuracy is determined based on the error value.
In one embodiment, the computer program when executed by the processor further performs the steps of: randomly selecting a plurality of concentration data samples from the concentration data sample set to form a concentration data test set; an input vector and a target output concentration are selected from the concentration data test set.
In one embodiment, the computer program when executed by the processor further performs the steps of: and constructing a visual prediction platform to visualize the prediction result of the suspended sediment concentration prediction model.
It will be understood by those skilled in the art that all or part of the processes of the methods of the embodiments described above can be implemented by hardware instructions of a computer program, which can be stored in a non-volatile computer-readable storage medium, and when executed, can include the processes of the embodiments of the methods described above. Any reference to memory, storage, database or other medium used in the embodiments provided herein can include at least one of non-volatile and volatile memory. Non-volatile Memory may include Read-Only Memory (ROM), magnetic tape, floppy disk, flash Memory, optical storage, or the like. Volatile Memory can include Random Access Memory (RAM) or external cache Memory. By way of illustration and not limitation, RAM can take many forms, such as Static Random Access Memory (SRAM) or Dynamic Random Access Memory (DRAM), among others.
The technical features of the above embodiments can be arbitrarily combined, and for the sake of brevity, all possible combinations of the technical features in the above embodiments are not described, but should be considered as the scope of the present specification as long as there is no contradiction between the combinations of the technical features.
The above-mentioned embodiments only express several embodiments of the present application, and the description thereof is more specific and detailed, but not construed as limiting the scope of the invention. It should be noted that, for a person skilled in the art, several variations and modifications can be made without departing from the concept of the present application, which falls within the scope of protection of the present application. Therefore, the protection scope of the present patent shall be subject to the appended claims.
Claims (10)
1. A method for predicting suspended sediment concentration, the method comprising:
acquiring suspended sediment concentration data to form a concentration data sample set;
randomly extracting a plurality of concentration data samples from the concentration data sample set to form a concentration data training set;
generating a prediction model based on L STM algorithm, and training the prediction model by adopting the concentration data training set;
when the training result meets the set condition, storing the model parameters of the prediction model to obtain a suspended sediment concentration prediction model;
and predicting the concentration of the suspended sediment according to the suspended sediment concentration prediction model.
2. The method of claim 1, wherein prior to forming the concentration data sample set, further comprising:
determining characteristic variables influencing the concentration of the suspended sediment;
and acquiring characteristic variable data, and fusing the characteristic variable data and the suspended sediment concentration data to form the concentration data sample set.
3. The method of claim 2, wherein determining the characteristic variables that affect the concentration of suspended silt comprises:
determining a plurality of power factors affecting the concentration of suspended sediment;
determining the influence importance of each power factor on the concentration of the suspended sediment based on a random forest algorithm;
and selecting a plurality of dominant power factors from the plurality of power factors as the characteristic variables according to the influence importance.
4. The method of claim 1, wherein after training the predictive model using the training set of concentration data, further comprising:
adjusting model parameters of the prediction model to obtain a plurality of candidate suspended sediment concentration prediction models;
determining the prediction precision of each candidate suspended sediment concentration prediction model;
and taking the candidate suspended sediment concentration prediction model corresponding to the prediction precision meeting the preset conditions as the suspended sediment concentration prediction model.
5. The method of claim 4, wherein determining the prediction accuracy of each candidate suspended sediment concentration prediction model comprises:
determining an input vector and a target output concentration of the candidate suspended sediment concentration prediction model;
inputting the input vector into the candidate suspended sediment concentration prediction model to obtain the predicted output concentration of the candidate suspended sediment concentration prediction model;
calculating an error value of the predicted output concentration relative to the target output concentration, and determining the prediction accuracy according to the error value.
6. The method of claim 5, wherein prior to determining the input vector and the target output concentration of the candidate suspended sediment concentration prediction model, further comprising:
randomly selecting a plurality of concentration data samples from the concentration data sample set to form a concentration data test set;
selecting the input vector and the target output concentration from the concentration data test set.
7. The method of claim 1, further comprising:
and constructing a visual prediction platform to visualize the prediction result of the suspended sediment concentration prediction model.
8. A suspended sediment concentration prediction apparatus, the apparatus comprising:
the sample set forming module is used for obtaining suspended sediment concentration data to form a concentration data sample set;
the training set forming module is used for randomly extracting a plurality of concentration data samples from the concentration data sample set to form a concentration data training set;
the model training module is used for generating a prediction model based on L STM algorithm and training the prediction model by adopting the concentration data training set;
the model determining module is used for storing the model parameters of the prediction model when the training result meets the set condition to obtain a suspended sediment concentration prediction model;
and the prediction module is used for predicting the concentration of the suspended sediment according to the suspended sediment concentration prediction model.
9. A computer device comprising a memory and a processor, the memory storing a computer program, wherein the processor implements the steps of the method of any one of claims 1 to 7 when executing the computer program.
10. A computer-readable storage medium, on which a computer program is stored, which, when being executed by a processor, carries out the steps of the method of any one of claims 1 to 7.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202010131794.0A CN111428419A (en) | 2020-02-29 | 2020-02-29 | Suspended sediment concentration prediction method and device, computer equipment and storage medium |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202010131794.0A CN111428419A (en) | 2020-02-29 | 2020-02-29 | Suspended sediment concentration prediction method and device, computer equipment and storage medium |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN111428419A true CN111428419A (en) | 2020-07-17 |
Family
ID=71547354
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202010131794.0A Pending CN111428419A (en) | 2020-02-29 | 2020-02-29 | Suspended sediment concentration prediction method and device, computer equipment and storage medium |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN111428419A (en) |
Cited By (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN112215366A (en) * | 2020-12-07 | 2021-01-12 | 成都数联铭品科技有限公司 | Model interpretation method and system based on result portrait, calculation and storage equipment |
| CN113359726A (en) * | 2021-06-04 | 2021-09-07 | 中山大学 | Method and system for predicting maximum turbid zone |
| CN116415508A (en) * | 2023-06-12 | 2023-07-11 | 珠江水利委员会珠江水利科学研究院 | Method and system for generating two-dimensional sediment model of estuary |
| CN116486305A (en) * | 2023-04-27 | 2023-07-25 | 中国海洋大学 | Image recognition-based deep sea suspended particulate matter concentration prediction method |
Citations (7)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN106599520A (en) * | 2016-12-31 | 2017-04-26 | 中国科学技术大学 | LSTM-RNN model-based air pollutant concentration forecast method |
| CN108009674A (en) * | 2017-11-27 | 2018-05-08 | 上海师范大学 | Air PM2.5 concentration prediction methods based on CNN and LSTM fused neural networks |
| US20190041320A1 (en) * | 2015-09-14 | 2019-02-07 | Luxembourg Institute Of Science And Technology (List) | Method For Determining In-Situ Suspended Sediment Properties |
| CN109615011A (en) * | 2018-12-14 | 2019-04-12 | 河海大学 | A kind of middle and small river short time flood forecast method based on LSTM |
| CN110147877A (en) * | 2019-05-24 | 2019-08-20 | 福州大学 | PM2.5 prediction technique based on the long network in short-term of convolution |
| CN110288199A (en) * | 2019-05-29 | 2019-09-27 | 北京航空航天大学 | The method of product quality forecast |
| CN110823190A (en) * | 2019-09-30 | 2020-02-21 | 广州地理研究所 | Island reef shallow sea water depth prediction method based on random forest |
-
2020
- 2020-02-29 CN CN202010131794.0A patent/CN111428419A/en active Pending
Patent Citations (7)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20190041320A1 (en) * | 2015-09-14 | 2019-02-07 | Luxembourg Institute Of Science And Technology (List) | Method For Determining In-Situ Suspended Sediment Properties |
| CN106599520A (en) * | 2016-12-31 | 2017-04-26 | 中国科学技术大学 | LSTM-RNN model-based air pollutant concentration forecast method |
| CN108009674A (en) * | 2017-11-27 | 2018-05-08 | 上海师范大学 | Air PM2.5 concentration prediction methods based on CNN and LSTM fused neural networks |
| CN109615011A (en) * | 2018-12-14 | 2019-04-12 | 河海大学 | A kind of middle and small river short time flood forecast method based on LSTM |
| CN110147877A (en) * | 2019-05-24 | 2019-08-20 | 福州大学 | PM2.5 prediction technique based on the long network in short-term of convolution |
| CN110288199A (en) * | 2019-05-29 | 2019-09-27 | 北京航空航天大学 | The method of product quality forecast |
| CN110823190A (en) * | 2019-09-30 | 2020-02-21 | 广州地理研究所 | Island reef shallow sea water depth prediction method based on random forest |
Non-Patent Citations (4)
| Title |
|---|
| REN, L 等: "Short-Term Forecasting of Coastal Surface Currents Using High Frequency Radar Data and Artificial Neural Networks", 《 REMOTE SENSING》 * |
| 卢淑怡 等: "一种空气污染物浓度预测深度学习平台", 《上海师范大学学报（自然科学版）》 * |
| 林开春 等: "基于随机森林和神经网络的空气质量预测研究", 《青岛大学学报（工程技术版）》 * |
| 王科等: "基于LSTM网络的变压器油中溶解气体浓度预测", 《电子测量技术》 * |
Cited By (7)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN112215366A (en) * | 2020-12-07 | 2021-01-12 | 成都数联铭品科技有限公司 | Model interpretation method and system based on result portrait, calculation and storage equipment |
| CN113359726A (en) * | 2021-06-04 | 2021-09-07 | 中山大学 | Method and system for predicting maximum turbid zone |
| CN113359726B (en) * | 2021-06-04 | 2022-12-13 | 中山大学 | Method and system for predicting maximum turbid zone |
| CN116486305A (en) * | 2023-04-27 | 2023-07-25 | 中国海洋大学 | Image recognition-based deep sea suspended particulate matter concentration prediction method |
| CN116486305B (en) * | 2023-04-27 | 2023-11-21 | 中国海洋大学 | Image recognition-based deep sea suspended particulate matter concentration prediction method |
| CN116415508A (en) * | 2023-06-12 | 2023-07-11 | 珠江水利委员会珠江水利科学研究院 | Method and system for generating two-dimensional sediment model of estuary |
| CN116415508B (en) * | 2023-06-12 | 2023-10-13 | 珠江水利委员会珠江水利科学研究院 | Method and system for generating two-dimensional sediment model of estuary |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| JP7175415B1 (en) | Marine environment element prediction method based on STEOF-LSTM | |
| CN111428419A (en) | Suspended sediment concentration prediction method and device, computer equipment and storage medium | |
| CN111199270B (en) | Regional wave height forecasting method and terminal based on deep learning | |
| CN111814956B (en) | Multi-task learning air quality prediction method based on multi-dimensional secondary feature extraction | |
| Lee et al. | Cultural heritage and the intelligent internet of things | |
| CN110837921A (en) | Real estate price prediction research method based on gradient lifting decision tree mixed model | |
| CN112001110A (en) | Structural damage identification monitoring method based on vibration signal space real-time recursive graph convolutional neural network | |
| CN111753461A (en) | Tidal water level correction method, target residual water level acquisition method, device and equipment | |
| Bai et al. | A forecasting method of forest pests based on the rough set and PSO-BP neural network | |
| CN113011455B (en) | Air quality prediction SVM model construction method | |
| CN112347537B (en) | Calibration method and device for engineering structure numerical model, electronic equipment and medium | |
| CN114065653A (en) | Construction method of power load prediction model and power load prediction method | |
| CN111507505A (en) | Method for constructing reservoir daily input prediction model | |
| CN115545334A (en) | Land use type prediction method, land use type prediction device, electronic device, and storage medium | |
| CN117408167A (en) | Debris flow disaster vulnerability prediction method based on deep neural network | |
| Xu et al. | Dynamic graph neural network with adaptive edge attributes for air quality prediction: A case study in China | |
| Wen et al. | An APPSO–SVM approach building the monitoring model of dam safety | |
| CN116960962A (en) | Mid-long term area load prediction method for cross-area data fusion | |
| CN111428420B (en) | Method and device for predicting sea surface flow velocity, computer equipment and storage medium | |
| Ray et al. | Solar irradiance forecasting using an artificial intelligence model | |
| CN114692529B (en) | CFD high-dimensional response uncertainty quantification method and device, and computer equipment | |
| CN115510763A (en) | Air pollutant concentration prediction method and system based on data-driven exploration | |
| CN115330082A (en) | PM2.5 concentration prediction method of LSTM network based on attention mechanism | |
| CN114021847A (en) | Short-term power load prediction method and system based on ISGU (inverse transformation and aggregation) hybrid model | |
| CN113011657A (en) | Method and device for predicting typhoon water level, electronic equipment and storage medium |
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 | ||
| RJ01 | Rejection of invention patent application after publication | ||
| RJ01 | Rejection of invention patent application after publication |
Application publication date: 20200717 |