CN113657544A - Sensor node data prediction method based on fusion neural network - Google Patents
Sensor node data prediction method based on fusion neural network Download PDFInfo
- Publication number
- CN113657544A CN113657544A CN202111002471.2A CN202111002471A CN113657544A CN 113657544 A CN113657544 A CN 113657544A CN 202111002471 A CN202111002471 A CN 202111002471A CN 113657544 A CN113657544 A CN 113657544A
- Authority
- CN
- China
- Prior art keywords
- neural network
- prediction model
- training
- data
- network prediction
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Pending
Links
Images
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/21—Design or setup of recognition systems or techniques; Extraction of features in feature space; Blind source separation
- G06F18/214—Generating training patterns; Bootstrap methods, e.g. bagging or boosting
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/23—Clustering techniques
-
- 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
Landscapes
- Engineering & Computer Science (AREA)
- Data Mining & Analysis (AREA)
- Theoretical Computer Science (AREA)
- Physics & Mathematics (AREA)
- Evolutionary Computation (AREA)
- Artificial Intelligence (AREA)
- General Engineering & Computer Science (AREA)
- Life Sciences & Earth Sciences (AREA)
- General Physics & Mathematics (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Bioinformatics & Computational Biology (AREA)
- Bioinformatics & Cheminformatics (AREA)
- Evolutionary Biology (AREA)
- General Health & Medical Sciences (AREA)
- Computing Systems (AREA)
- Mathematical Physics (AREA)
- Software Systems (AREA)
- Molecular Biology (AREA)
- Computational Linguistics (AREA)
- Biophysics (AREA)
- Biomedical Technology (AREA)
- Health & Medical Sciences (AREA)
- Testing Or Calibration Of Command Recording Devices (AREA)
Abstract
The invention provides a sensor node data prediction method based on a fusion neural network, which is applied to the technical field of data prediction; building a neural network prediction model, and creating a data preprocessing instruction and a model training instruction; selecting data classification sampling of sensor nodes to obtain a first training data set and a second training data set; performing data preprocessing on a first training data set and a second training data set obtained by sampling to obtain a first training sample and a second training sample; respectively training a neural network prediction model by using the obtained first training sample and the second training sample to obtain a first neural network prediction model and a second neural network prediction model; respectively inputting vectors of the first neural network prediction model and the second neural network prediction model to the sensor nodes and performing data preprocessing; and selecting the optimal prediction model from the first neural network prediction model and the second neural network prediction model to output vectors so as to enable the sensor node data prediction value to reach the most accurate prediction value.
Description
Technical Field
The invention relates to the technical field of data prediction, in particular to a sensor node data prediction method based on a fusion neural network.
Background
Sensor networks are an emerging, rapidly developing field of research. The sensor nodes collect energy from the surrounding environment, the service life of the wireless sensor network is prolonged, and the defect that the nodes cannot work due to energy consumption is overcome. However, new technology introduction also brings new challenges. The environmental energy has the characteristics of time randomness, periodicity and uneven regional distribution, so that the capability of the sensor node for acquiring data has asymmetry and instability. Thus, predicting the collected data ahead of time may help the node to take better strategies. If the data collected by the future period prediction is less, the node needs to reduce the self energy consumption; and on the contrary, more data acquisition, processing and transmission tasks can be assumed.
In the related sensor node training method, one method is to initially select a single hidden layer neural network with a larger structure, train the network to preliminarily converge through a training sample, extract an output value of a hidden layer node of the converged neural network, then perform correlation analysis-Principal Component Analysis (PCA) on the output value of the hidden layer node, remove data with larger correlation, further reduce dimensionality, combine the nodes with larger correlation corresponding to the hidden layer, and reduce the number of neuron nodes; another is based on the correlation analysis between the hidden layer output and the network output: firstly, a single hidden layer neural network with a larger structure is initially selected, the network is trained to be preliminarily converged through a training sample, output values of hidden layer nodes of the converged neural network are extracted, then output values of the network output layers are extracted, then a correlation coefficient of each hidden layer node output value and the network output layer node output value is calculated, when the correlation coefficient is smaller than a certain threshold value set in advance, the corresponding hidden layer node is deleted, and the number of the hidden layer nodes is determined appropriately.
Although the existing prediction model based on the depth model has achieved higher prediction accuracy, there still exist some problems to be solved. Most existing approaches rely on a classical sequence-to-sequence architecture, inputting a historical time sequence to the encoder and initializing the decoder with its final state to make predictions. However, there is no real sample during testing, and the decoder completely depends on the output generated by the model itself to predict the next output, which may cause the difference of the next action generated by the model during training and testing, and the errors generated in this process may accumulate continuously, although methods such as planned sampling, generation of a network countermeasures, etc. are also used to avoid errors, these methods have some drawbacks, so it is very necessary to improve the prediction accuracy to effectively avoid errors.
In view of this, the invention provides a sensor node data prediction method based on a fusion neural network, so as to improve the sensor node data prediction accuracy.
Disclosure of Invention
The invention aims to improve the problem of the data prediction accuracy of sensor nodes, and provides a sensor node data prediction method based on a fusion neural network.
The invention provides a sensor node data prediction method based on a fusion neural network, which comprises the following steps:
building a neural network prediction model, and creating a data preprocessing instruction and a model training instruction;
selecting data classification sampling of sensor nodes to obtain a first training data set and a second training data set;
performing a data preprocessing instruction on a first training data set and a second training data set obtained by sampling to obtain a first training sample and a second training sample;
respectively training a neural network prediction model by using the obtained first training sample and the second training sample to obtain a first neural network prediction model and a second neural network prediction model;
respectively inputting vectors of a first neural network prediction model and a second neural network prediction model to the sensor nodes and performing a data preprocessing instruction;
and selecting the optimal prediction model from the first neural network prediction model and the second neural network prediction model to output vectors so as to enable the sensor node data prediction value to reach the most accurate prediction value.
The sensor node data classification step by step constructs a decision tree according to classification labels, the number n of the classification labels at the bottommost layer of the decision tree is set as an initial X value of an X-means clustering algorithm, the initial X value is used as an initial value of the clustering algorithm, two node data are selected from all sensor node data to serve as mass center nodes of each classification, and data of the two mass center nodes are collected to obtain a first training data set and a second training data set.
Further, the step of performing data preprocessing on the first training data set and the second training data set obtained by sampling to obtain a first training sample and a second training sample further includes: the first training data set and the second training data set are separately subjected to data preprocessing, and negative value data in the data sets are eliminated; and restoring the negative value of the negative value data by zero padding, and restoring the missing data by using a linear interpolation method.
Further, the step of respectively training the neural network prediction model by using the obtained first training sample and the second training sample to obtain a first neural network prediction model and a second neural network prediction model is as follows: the neural network model has self-adaptive elastic characteristics, and the numerical values of the data set and the training sample are adjusted in a self-adaptive mode according to errors in the training process.
Further, the step of inputting vectors of the first neural network prediction model and the second neural network prediction model to the sensor nodes respectively and performing data preprocessing comprises: and acquiring data of the first 24 hours of prediction in the first neural network prediction model and the second neural network prediction model as vectors input to the sensor nodes, and performing normalization processing to obtain processed input vectors.
Further, the optimal prediction model output vector is selected from the first neural network prediction model and the second neural network prediction model, so that the predicted value of the sensor node data reaches the most accurate value: and the output vectors of the first neural network prediction model and the second neural network prediction model are the most accurate values obtained by performing inverse normalization calculation on data in 24 hours after prediction.
The invention also provides a sensor node data prediction device based on the fusion neural network, which comprises the following steps: the creating unit is used for constructing a neural network prediction model, creating a data preprocessing instruction and a model training instruction, and selecting data classification sampling of sensor nodes to obtain a first training data set and a second training data set;
the training unit is used for carrying out data preprocessing on a first training data set and a second training data set obtained by sampling so as to obtain a first training sample and a second training sample; respectively training a neural network prediction model by using the obtained first training sample and the second training sample to obtain a first neural network prediction model and a second neural network prediction model;
the processing unit is used for respectively inputting vectors of the first neural network prediction model and the second neural network prediction model to the sensor nodes and carrying out data preprocessing; and selecting the optimal prediction model from the first neural network prediction model and the second neural network prediction model to output vectors so as to enable the sensor node data prediction value to reach the most accurate prediction value.
Further, the creating unit further includes: the creating subunit is used for constructing a neural network prediction model, creating a data preprocessing instruction and a model training instruction;
and the selecting subunit is used for selecting data classification sampling of the sensor nodes to obtain a first training data set and a second training data set.
Further, the training unit comprises: the acquisition subunit is used for carrying out data preprocessing on the first training data set and the second training data set obtained by sampling so as to obtain a first training sample and a second training sample;
and the training subunit is used for respectively training the neural network prediction model by using the obtained first training sample and the second training sample to obtain a first neural network prediction model and a second neural network prediction model.
Further, the processing unit further comprises: the input subunit is used for respectively inputting vectors of the first neural network prediction model and the second neural network prediction model to the sensor nodes and performing data preprocessing;
and the output subunit selects the optimal prediction model output vector from the first neural network prediction model and the second neural network prediction model so as to enable the sensor node data prediction value to reach the most accurate prediction value.
The invention provides a sensor node data prediction method based on a fusion neural network, which has the following beneficial effects:
the method disclosed by the invention integrates the neural network into the prediction model so as to improve the prediction accuracy of the existing sensor node data and simultaneously avoid errors and/or deviations generated in the process of predicting the result.
Drawings
FIG. 1 is an overall flow chart of an embodiment of a method for predicting data based on a sensor node of a converged neural network according to the present invention;
FIG. 2 is a flowchart illustrating an embodiment of a data prediction apparatus based on a sensor node of a neural network according to the present invention;
the implementation, functional features and advantages of the present invention will be further described with reference to the accompanying drawings.
Detailed Description
It should be understood that the specific embodiments described herein are merely illustrative of the invention and are not intended to limit the invention.
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. All other embodiments, which can be derived by a person skilled in the art from the embodiments given herein without making any creative effort, shall fall within the protection scope of the present invention.
Referring to fig. 1, a method for predicting data based on a sensor node of a neural network in an embodiment of the present invention includes:
building a neural network prediction model, and creating a data preprocessing instruction and a model training instruction;
selecting data classification sampling of sensor nodes to obtain a first training data set and a second training data set;
performing a data preprocessing instruction on a first training data set and a second training data set obtained by sampling to obtain a first training sample and a second training sample;
respectively training a neural network prediction model by using the obtained first training sample and the second training sample to obtain a first neural network prediction model and a second neural network prediction model;
respectively inputting vectors of a first neural network prediction model and a second neural network prediction model to the sensor nodes and performing a data preprocessing instruction;
and selecting the optimal prediction model from the first neural network prediction model and the second neural network prediction model to output vectors so as to enable the sensor node data prediction value to reach the most accurate prediction value.
In a specific embodiment: constructing an initial model of a neural network prediction model, and creating an instruction for data preprocessing and an instruction for model training; then, selecting data of the sensor nodes for classified sampling to obtain a first training data set and a second training data set, and carrying out a data preprocessing instruction on the first training data set and the second training data set obtained by sampling to obtain a first training sample and a second training sample; then, respectively training a neural network prediction model by using the obtained first training sample and the second training sample to obtain a first neural network prediction model and a second neural network prediction model; and inputting vectors of the first neural network prediction model and the second neural network prediction model into the sensor nodes, performing data preprocessing, and selecting the prediction model with the best data prediction value from the two neural network prediction models, namely the sensor node data prediction value reaches the most accurate prediction value.
In one embodiment: the method comprises the following steps of selecting data classification samples of sensor nodes to obtain a first training data set and a second training data set: the sensor node data classification step by step constructs a decision tree according to classification labels, the number n of the classification labels at the bottommost layer of the decision tree is set as an initial X value of an X-means clustering algorithm, the initial X value is used as an initial value of the clustering algorithm, two node data are selected from all sensor node data to serve as mass center nodes of each classification, and data of the two mass center nodes are collected to obtain a first training data set and a second training data set.
In a specific embodiment: the number of leaf nodes at the bottom of the decision tree is the number of classes into which the whole sample is divided, i.e. the initial X value of the X-means clustering algorithm to be executed next. The initial X value has great influence on the whole clustering algorithm, and the selection of the X value directly influences the subsequent clustering conditions, including the accuracy of the position of the sink node, the clustering convergence speed and the like. The initial value of X is not chosen randomly, and too large or too small results in less than optimal clustering effect, and even results in erroneous clustering. Excessive node classification can cause uneven energy consumption of the whole sensor node, is not beneficial to improving the whole service life of the wireless sensor network, and can also increase the cost of the network node, while too little node classification can not ensure that the wireless sensor network covers the whole prediction area, and nodes which are far away from each other cannot successfully transmit data to a sink node; after a complete decision tree is established, two node data are selected as the centroid nodes of each classification, and the data of the two centroid nodes are collected to obtain a first training data set and a second training data set.
In one embodiment: the step of performing a data preprocessing instruction on the first training data set and the second training data set obtained by sampling to obtain a first training sample and a second training sample further includes:
the first training data set and the second training data set are separately subjected to data preprocessing, and negative value data in the data sets are eliminated; and restoring the negative value of the negative value data by zero padding, and restoring the missing data by using a linear interpolation method.
In a specific embodiment: the first training data set and the second training data set are subjected to data preprocessing separately, abnormal data in the data sets are eliminated, and the abnormal data include but are not limited to negative value data; repairing a negative value by zero filling, and repairing missing data by using a linear interpolation method, wherein the formula is as follows:
wherein z is the data set to be interpolated, z0 is the data value before z, and z1 is the data value after zT0 and t1 are z0 and z1 data value variables, respectively, and t is a linear interpolated data set. z0 and z1 take data at the previous time and the next time of z, respectively.
Predicting a data value variable of data at the next moment according to the data set sequence, wherein the data value variable at the moment is a dependent variable value, and the data value variable at the previous moment is an independent variable value; the dependent variable value and the independent variable value are respectively expressed.
In one embodiment: respectively training a neural network prediction model by using the obtained first training sample and the second training sample to obtain a first neural network prediction model and a second neural network prediction model, wherein the step of obtaining the first neural network prediction model and the second neural network prediction model comprises the following steps:
the neural network model has self-adaptive elastic characteristics, and the numerical values of the data set and the training sample are adjusted in a self-adaptive mode according to errors in the training process.
In a specific embodiment: self-adaptive adjustment is carried out on the elastic characteristic, the data set and the training sample value which are set in the neural network model according to the error in the training process; when the neural network prediction model tests the time sequence data of the nodes of the sensor, the data set data to be tested is input into the neural network prediction model, and the data set data to be tested are sequentially output one by one through the neural network prediction model to obtain the output value of the data set data to be tested. And comparing the output value of the data set data to be tested with the actual value of the data set data to be tested, judging that the data set data to be tested has an error when the absolute value of the difference between the output value of the data set data to be tested and the actual value of the data set data to be tested is greater than a preset threshold value, and carrying out self-adaptive adjustment on the neural network prediction model.
In one embodiment: respectively inputting vectors of a first neural network prediction model and a second neural network prediction model to sensor nodes and performing a data preprocessing instruction, wherein the method comprises the following steps: and acquiring data of the first 24 hours of prediction in the first neural network prediction model and the second neural network prediction model as vectors input to the sensor nodes, and performing normalization processing to obtain processed input vectors.
In a specific embodiment: collecting data of the first 24 hours of prediction in the first neural network prediction model and the second neural network prediction model as vectors input to the sensor nodes, and excluding abnormal data, wherein the abnormal data comprises but is not limited to negative value data and maximum value data larger than a theoretical vector; and then normalization processing is carried out, wherein the input vector is normalized according to the maximum value and the minimum value of the input vector of the training sample, and the input vector after normalization processing is obtained.
In one embodiment: selecting the optimal prediction model output vector from the first neural network prediction model and the second neural network prediction model so as to enable the sensor node data prediction value to reach the most accurate value:
and the output vectors of the first neural network prediction model and the second neural network prediction model are the most accurate values obtained by performing inverse normalization calculation on data in 24 hours after prediction.
In a specific embodiment: and performing reverse normalization on the output vector of the neural network prediction model according to the maximum value and the minimum value of the output vector of the training sample to obtain an output vector after the reverse normalization, wherein the output vector is the output vector which is input into the neural network prediction model to enable the prediction result to reach the most accurate value.
Referring to fig. 2, a device for predicting data based on a sensor node of a converged neural network according to an embodiment of the present invention includes:
the creating unit is used for constructing a neural network prediction model, creating a data preprocessing instruction and a model training instruction, and selecting data classification sampling of sensor nodes to obtain a first training data set and a second training data set;
the training unit is used for carrying out data preprocessing on a first training data set and a second training data set obtained by sampling so as to obtain a first training sample and a second training sample; respectively training a neural network prediction model by using the obtained first training sample and the second training sample to obtain a first neural network prediction model and a second neural network prediction model;
the processing unit is used for respectively inputting vectors of the first neural network prediction model and the second neural network prediction model to the sensor nodes and carrying out data preprocessing; and selecting the optimal prediction model from the first neural network prediction model and the second neural network prediction model to output vectors so as to enable the sensor node data prediction value to reach the most accurate prediction value.
In a specific embodiment: the creating unit builds an initial model of the neural network prediction model and creates an instruction for data preprocessing and an instruction for model training; then, selecting data classification sampling of the sensor nodes to obtain a first training data set and a second training data set; the training unit carries out data preprocessing instructions on a first training data set and a second training data set obtained through sampling, and therefore a first training sample and a second training sample are obtained; then, respectively training a neural network prediction model by using the obtained first training sample and the second training sample to obtain a first neural network prediction model and a second neural network prediction model; the processing unit inputs vectors of the first neural network prediction model and the second neural network prediction model into the sensor nodes for data preprocessing, and selects the prediction model with the best data prediction value from the two neural network prediction models, namely the sensor node data prediction value reaches the most accurate prediction value.
The method disclosed by the invention integrates the neural network into the prediction model so as to improve the prediction accuracy of the existing sensor node data and simultaneously avoid errors and/or deviations generated in the process of predicting the result.
Although embodiments of the present invention have been shown and described, it will be appreciated by those skilled in the art that changes, modifications, substitutions and alterations can be made in these embodiments without departing from the principles and spirit of the invention, the scope of which is defined in the appended claims and their equivalents.
Claims (10)
1. A sensor node data prediction method based on a fusion neural network is characterized by comprising the following steps:
building a neural network prediction model, and creating a data preprocessing instruction and a model training instruction;
selecting data classification sampling of sensor nodes to obtain a first training data set and a second training data set;
performing a data preprocessing instruction on a first training data set and a second training data set obtained by sampling to obtain a first training sample and a second training sample;
respectively training a neural network prediction model by using the obtained first training sample and the second training sample to obtain a first neural network prediction model and a second neural network prediction model;
respectively inputting vectors of a first neural network prediction model and a second neural network prediction model to the sensor nodes and performing a data preprocessing instruction;
and selecting the optimal prediction model from the first neural network prediction model and the second neural network prediction model to output vectors so as to enable the sensor node data prediction value to reach the most accurate prediction value.
2. The method for predicting the data of the sensor nodes based on the fusion neural network as claimed in claim 1, wherein the step of obtaining the first training data set and the second training data set by selecting the data classification samples of the sensor nodes comprises:
the sensor node data classification step by step constructs a decision tree according to classification labels, the number n of the classification labels at the bottommost layer of the decision tree is set as an initial X value of an X-means clustering algorithm, the initial X value is used as an initial value of the clustering algorithm, two node data are selected from all sensor node data to serve as mass center nodes of each classification, and data of the two mass center nodes are collected to obtain a first training data set and a second training data set.
3. The method of claim 1, wherein the step of performing a data preprocessing instruction on the first training data set and the second training data set obtained by sampling to obtain the first training sample and the second training sample further comprises:
the first training data set and the second training data set are separately subjected to data preprocessing, and negative value data in the data sets are eliminated; and restoring the negative value of the negative value data by zero padding, and restoring the missing data by using a linear interpolation method.
4. The method for predicting the data of the sensor nodes based on the converged neural network, according to claim 1, wherein the step of using the obtained first training sample and the second training sample to respectively train the neural network prediction model to obtain the first neural network prediction model and the second neural network prediction model comprises:
the neural network model has self-adaptive elastic characteristics, and the numerical values of the data set and the training sample are adjusted in a self-adaptive mode according to errors in the training process.
5. The method for predicting the data of the sensor nodes based on the converged neural network, according to claim 1, wherein the step of inputting the vectors of the first neural network prediction model and the second neural network prediction model to the sensor nodes respectively and performing data preprocessing instructions comprises:
and acquiring data of the first time predicted in the first neural network prediction model and the second neural network prediction model as vectors input to the sensor nodes, and performing normalization processing to obtain processed input vectors.
6. The method for predicting the sensor node data based on the converged neural network, according to claim 1, wherein in the step of selecting the optimal prediction model output vector from the first neural network prediction model and the second neural network prediction model to enable the predicted value of the sensor node data to reach the most accurate value:
and the output vectors of the first neural network prediction model and the second neural network prediction model are the most accurate values obtained by performing inverse normalization calculation on data in 24 hours after prediction.
7. A prediction device for sensor node data based on a converged neural network, wherein the prediction device for sensor node data based on a converged neural network is adopted to execute the prediction method for sensor node data based on a converged neural network according to any one of claims 1 to 6, and the prediction device for sensor node data based on a converged neural network comprises:
the creating unit is used for constructing a neural network prediction model, creating a data preprocessing instruction and a model training instruction, and selecting data classification sampling of sensor nodes to obtain a first training data set and a second training data set;
the training unit is used for carrying out data preprocessing on a first training data set and a second training data set obtained by sampling so as to obtain a first training sample and a second training sample; respectively training a neural network prediction model by using the obtained first training sample and the second training sample to obtain a first neural network prediction model and a second neural network prediction model;
the processing unit is used for respectively inputting vectors of the first neural network prediction model and the second neural network prediction model to the sensor nodes and carrying out data preprocessing; and selecting the optimal prediction model from the first neural network prediction model and the second neural network prediction model to output vectors so as to enable the sensor node data prediction value to reach the most accurate prediction value.
8. The device according to claim 7, wherein the creating unit further comprises:
the creating subunit is used for constructing a neural network prediction model, creating a data preprocessing instruction and a model training instruction;
and the selecting subunit is used for selecting data classification sampling of the sensor nodes to obtain a first training data set and a second training data set.
9. The device according to claim 7, wherein the training unit comprises:
the acquisition subunit is used for carrying out data preprocessing on the first training data set and the second training data set obtained by sampling so as to obtain a first training sample and a second training sample;
and the training subunit is used for respectively training the neural network prediction model by using the obtained first training sample and the second training sample to obtain a first neural network prediction model and a second neural network prediction model.
10. The device according to claim 7, wherein the processing unit further comprises:
the input subunit is used for respectively inputting vectors of the first neural network prediction model and the second neural network prediction model to the sensor nodes and performing data preprocessing;
and the output subunit selects the optimal prediction model output vector from the first neural network prediction model and the second neural network prediction model so as to enable the sensor node data prediction value to reach the most accurate prediction value.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202111002471.2A CN113657544A (en) | 2021-08-30 | 2021-08-30 | Sensor node data prediction method based on fusion neural network |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202111002471.2A CN113657544A (en) | 2021-08-30 | 2021-08-30 | Sensor node data prediction method based on fusion neural network |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN113657544A true CN113657544A (en) | 2021-11-16 |
Family
ID=78493154
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202111002471.2A Pending CN113657544A (en) | 2021-08-30 | 2021-08-30 | Sensor node data prediction method based on fusion neural network |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN113657544A (en) |
-
2021
- 2021-08-30 CN CN202111002471.2A patent/CN113657544A/en active Pending
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN109102005B (en) | Small sample deep learning method based on shallow model knowledge migration | |
| CN108900346B (en) | Wireless network flow prediction method based on LSTM network | |
| CN110335168B (en) | Method and system for optimizing power utilization information acquisition terminal fault prediction model based on GRU | |
| CN115618296A (en) | Dam monitoring time sequence data anomaly detection method based on graph attention network | |
| CN110163429A (en) | A kind of short-term load forecasting method based on similar day optimal screening | |
| EP1654692A1 (en) | Method and apparatus for automated feature selection | |
| CN115099296A (en) | Sea wave height prediction method based on deep learning algorithm | |
| CN116935649A (en) | Urban traffic flow prediction method for multi-view fusion space-time dynamic graph convolution network | |
| CN112905436B (en) | Quality evaluation prediction method for complex software | |
| CN108984851B (en) | Weighted Gaussian model soft measurement modeling method with time delay estimation | |
| CN112149896A (en) | Attention mechanism-based mechanical equipment multi-working-condition fault prediction method | |
| Bardsiri et al. | Towards improvement of analogy-based software development effort estimation: A review | |
| CN113657544A (en) | Sensor node data prediction method based on fusion neural network | |
| CN115619028A (en) | Clustering algorithm fusion-based power load accurate prediction method | |
| CN114970674A (en) | Time sequence data concept drift adaptation method based on relevance alignment | |
| CN115081609A (en) | Acceleration method in intelligent decision, terminal equipment and storage medium | |
| CN114372640A (en) | Wind power prediction method based on fluctuation sequence classification correction | |
| CN112348275A (en) | Regional ecological environment change prediction method based on online incremental learning | |
| Kumari et al. | Remaining useful life prediction using hybrid neural network and genetic algorithm approaches | |
| CN116722548B (en) | Photovoltaic power generation prediction method based on time sequence model and related equipment | |
| CN111310907A (en) | Microwave assembly fault diagnosis method, device and equipment | |
| He et al. | An Integrated Long Short-Term Memory Based Online Log Anomaly Detection | |
| Vignisson | Anomaly Detection in Streaming Data from a Sensor Network | |
| Al Mamun et al. | Diagnostic of an Aircraft by Different Dynamics of Operating Conditions | |
| CN118040675A (en) | Transient stability evaluation method, device, equipment and medium for power system |
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 |