CN114048790B - Road base strain analysis method based on coupling bidirectional LSTM and convolution structure network - Google Patents
Road base strain analysis method based on coupling bidirectional LSTM and convolution structure network Download PDFInfo
- Publication number
- CN114048790B CN114048790B CN202110877678.8A CN202110877678A CN114048790B CN 114048790 B CN114048790 B CN 114048790B CN 202110877678 A CN202110877678 A CN 202110877678A CN 114048790 B CN114048790 B CN 114048790B
- Authority
- CN
- China
- Prior art keywords
- data
- layer
- road
- strain
- network
- 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.)
- Active
Links
- 230000002457 bidirectional effect Effects 0.000 title claims abstract description 24
- 238000004458 analytical method Methods 0.000 title claims abstract description 18
- 230000008878 coupling Effects 0.000 title claims abstract description 10
- 238000010168 coupling process Methods 0.000 title claims abstract description 10
- 238000005859 coupling reaction Methods 0.000 title claims abstract description 10
- 238000012544 monitoring process Methods 0.000 claims abstract description 43
- 238000012549 training Methods 0.000 claims abstract description 21
- 238000007781 pre-processing Methods 0.000 claims abstract description 6
- 239000010410 layer Substances 0.000 claims description 54
- 238000000034 method Methods 0.000 claims description 20
- 239000002689 soil Substances 0.000 claims description 12
- 238000010219 correlation analysis Methods 0.000 claims description 10
- 230000008569 process Effects 0.000 claims description 9
- 238000004422 calculation algorithm Methods 0.000 claims description 7
- 230000006870 function Effects 0.000 claims description 7
- 238000005457 optimization Methods 0.000 claims description 6
- 230000004913 activation Effects 0.000 claims description 5
- 239000010426 asphalt Substances 0.000 claims description 4
- 230000003204 osmotic effect Effects 0.000 claims description 4
- 238000011176 pooling Methods 0.000 claims description 4
- 238000012935 Averaging Methods 0.000 claims description 3
- 230000003044 adaptive effect Effects 0.000 claims description 3
- 238000013135 deep learning Methods 0.000 claims description 3
- 238000012545 processing Methods 0.000 claims description 3
- 238000012360 testing method Methods 0.000 claims description 3
- 239000002344 surface layer Substances 0.000 claims description 2
- 239000013589 supplement Substances 0.000 claims 1
- 230000007774 longterm Effects 0.000 abstract description 5
- 238000012423 maintenance Methods 0.000 abstract description 4
- 238000013508 migration Methods 0.000 abstract description 2
- 238000010220 Pearson correlation analysis Methods 0.000 abstract 1
- 230000009286 beneficial effect Effects 0.000 abstract 1
- 230000003446 memory effect Effects 0.000 abstract 1
- 238000005728 strengthening Methods 0.000 abstract 1
- 238000013527 convolutional neural network Methods 0.000 description 6
- 230000000694 effects Effects 0.000 description 5
- ORILYTVJVMAKLC-UHFFFAOYSA-N Adamantane Natural products C1C(C2)CC3CC1CC2C3 ORILYTVJVMAKLC-UHFFFAOYSA-N 0.000 description 4
- 230000008859 change Effects 0.000 description 3
- 238000010586 diagram Methods 0.000 description 3
- 238000013528 artificial neural network Methods 0.000 description 2
- 230000008901 benefit Effects 0.000 description 2
- 238000007405 data analysis Methods 0.000 description 2
- 238000000605 extraction Methods 0.000 description 2
- 238000013459 approach Methods 0.000 description 1
- 238000009412 basement excavation Methods 0.000 description 1
- 230000005540 biological transmission Effects 0.000 description 1
- 230000015572 biosynthetic process Effects 0.000 description 1
- 238000004364 calculation method Methods 0.000 description 1
- 238000007796 conventional method Methods 0.000 description 1
- 238000012217 deletion Methods 0.000 description 1
- 230000037430 deletion Effects 0.000 description 1
- 201000010099 disease Diseases 0.000 description 1
- 208000037265 diseases, disorders, signs and symptoms Diseases 0.000 description 1
- 230000004927 fusion Effects 0.000 description 1
- 230000006872 improvement Effects 0.000 description 1
- 238000010801 machine learning Methods 0.000 description 1
- 238000004519 manufacturing process Methods 0.000 description 1
- 238000013507 mapping Methods 0.000 description 1
- 238000005259 measurement Methods 0.000 description 1
- 230000009467 reduction Effects 0.000 description 1
- 238000005070 sampling Methods 0.000 description 1
- 238000004088 simulation Methods 0.000 description 1
- 230000001502 supplementing effect Effects 0.000 description 1
- 230000009897 systematic effect Effects 0.000 description 1
- 238000010200 validation analysis Methods 0.000 description 1
- 239000013598 vector Substances 0.000 description 1
- XLYOFNOQVPJJNP-UHFFFAOYSA-N water Substances O XLYOFNOQVPJJNP-UHFFFAOYSA-N 0.000 description 1
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
- G06N3/04—Architecture, e.g. interconnection topology
- G06N3/045—Combinations of networks
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- 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
- 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
Landscapes
- Engineering & Computer Science (AREA)
- Theoretical Computer Science (AREA)
- Physics & Mathematics (AREA)
- Data Mining & Analysis (AREA)
- Evolutionary Computation (AREA)
- Life Sciences & Earth Sciences (AREA)
- Artificial Intelligence (AREA)
- General Physics & Mathematics (AREA)
- General Engineering & Computer Science (AREA)
- Computing Systems (AREA)
- Software Systems (AREA)
- Molecular Biology (AREA)
- Computational Linguistics (AREA)
- Biophysics (AREA)
- Biomedical Technology (AREA)
- Mathematical Physics (AREA)
- General Health & Medical Sciences (AREA)
- Health & Medical Sciences (AREA)
- Bioinformatics & Cheminformatics (AREA)
- Bioinformatics & Computational Biology (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Evolutionary Biology (AREA)
- Management, Administration, Business Operations System, And Electronic Commerce (AREA)
- Image Analysis (AREA)
Abstract
The invention discloses a road base strain analysis method based on a coupling bidirectional LSTM and a convolution structure network, which comprises the steps of preprocessing original data of long-term monitoring data acquired by an automatic road monitoring system, adopting Pearson correlation analysis on the basis, researching the relationship between the data to achieve the purpose of reducing the input data dimension, reducing training errors, improving the prediction precision of a model, finally selecting data in a proper time period, inputting the data into a BiLSTM-CNN network model together for training, learning physical characteristics among sensor data, aiming at the problem that the data are discontinuous in each time period, adopting a weight self-migration training mode, ensuring that the history information characteristics are relatively more obvious in the next time period, being easier to learn by the model, being beneficial to strengthening the memory effect of the model, reducing overfitting and improving the generalization capability of the model. In addition, the invention not only can save labor and time cost, but also can lay a foundation for subsequent road maintenance work.
Description
Technical Field
The invention belongs to the field of time sequence data analysis, and relates to a road base strain analysis method based on a coupling bidirectional LSTM and a convolution structure network. The method is applied to road base strain analysis by using the road structure monitoring data acquired by the multi-sensor fusion road monitoring system.
Background
In recent years, with the basic formation of highway networks in China, future highway infrastructure investment will gradually point to the road maintenance management industry. Diseases such as cracks, ruts and the like influence the safety and the comfort of driving to a certain extent, and the pavement needs to be maintained in time. Therefore, the accurate basic strain analysis result can avoid serious structural damage of the road in the future to the greatest extent, and reliable and effective technical support is provided for subsequent road maintenance management work.
At present, the conventional basic strain prediction analysis method is generally used for analyzing and calculating stress and strain data based on finite element simulation, a mechanical-experience model, a mathematical statistics method and the like, and the conventional method is usually used for carrying out data comparison to correct model parameters and simplify a three-dimensional model, so that an analysis result is affected to a certain extent, and an analysis process is complex and time-consuming. The machine learning method is gradually applied to the field of monitoring big data analysis, can analyze and predict the obtained data relatively quickly and efficiently, saves time, reduces cost and obtains more accurate results.
At present, the deep learning-based road base strain prediction analysis often faces the problems of large input data dimension, discontinuous time of monitoring data and the like, the data correlation analysis is carried out to reduce the data dimension of the input end, and the adoption of weight self-migration in the training process can bring about considerable performance improvement, improve the accuracy and the stability and reduce the overfitting.
Therefore, the invention provides a road base strain analysis method based on a coupled bidirectional LSTM and a convolution structure network (BiLSTM-CNN). According to the invention, firstly, the sensor data collected by a road monitoring system is subjected to original data preprocessing to obtain a data set which can be input into a network for training, and then the data set is subjected to characteristic correlation analysis to obtain the Pearson correlation coefficient, so that the characteristic data with smaller correlation coefficient is removed to achieve the purpose of data dimension reduction, and finally, the BiLSTM-CNN is used for training the final data set, so that the strain data of a road base layer is subjected to predictive analysis, the strain change trend of the future base layer is known in advance, and the help is provided for the subsequent road maintenance decision.
Disclosure of Invention
The invention aims to analyze road surface monitoring data which are obtained by a sensor and are from 2012 to 2020 and up to eight years long by a road base strain prediction analysis method based on a coupling bidirectional LSTM and convolution structure network, wherein the long-term monitoring data comprise asphalt strain, base three-way strain, soil layer strain, soil pressure, temperature, osmotic pressure and soil moisture, so that the base strain is predicted according to proper influence factors, and the change trend of the base strain in a future period of time is obtained.
1. Coupled bidirectional LSTM and convolutional structured network
The coupling Bidirectional LSTM and convolution structure network (Bidirection LSTM AND CNN, BILSTM-CNN) adopted by the invention consists of one Bidirectional LSTM layer and four 1D convolution layers, as shown in figure 1. The CNN is utilized to extract fixed characteristics in the data time dimension acquired by the sensor, so that the characteristic extraction of monitoring signals of various sensors on the road is realized; meanwhile, a bidirectional LSTM neural network is selected to process the characteristic sequence output by the CNN part, time sequence information in data is mined, the signal change trend and characteristics of the monitoring sensor are extracted in the time dimension, and the prediction of the strain of the roadbed base layer is realized.
The technical scheme adopted by the invention is a road base strain prediction analysis method based on a coupling bidirectional LSTM and convolution structure network, which comprises four parts of original monitoring data preprocessing, characteristic correlation analysis, data set manufacturing and BiLSTM-CNN network base strain prediction, as shown in figure 2, and comprises the following specific steps:
step one: preprocessing original monitoring data;
First, road monitoring data acquired by various sensors are integrated.
And secondly, according to the time span of the data acquired by the road monitoring system, deleting the repeated value and the jump value of the data after selecting the data in a proper time period, and supplementing the missing value to prevent poor training effect of the monitoring data.
Finally, according to the characteristics of time sequence data, the time intervals among all data points are unified to be 1 hour.
Step two: analyzing the characteristic correlation;
in the first step, correlation analysis is carried out on the first three data points acquired by each sensor after processing, and the correlation between features is measured according to the Pearson correlation coefficient. In the second step, the soil layer strain is not removed because of the high correlation.
Step three: creating a data set;
And (3) averaging all monitoring data points of asphalt strain, osmotic pressure, soil moisture and base three-way strain, and dividing the temperature into three parts of a surface layer, a base layer and a soil base according to the embedded position of the temperature sensor. Finally, the monitored data set is calculated according to the training set and the testing set proportion of about 8:2, dividing.
Step four: biLSTM-CNN network base layer strain prediction;
The BiLSTM-CNN network consists of 4 one-dimensional convolutional layers, one layer of bidirectional LSTM and two full-connection layers. The convolution kernel sizes of the one-dimensional convolution layers are 2, the number of the convolution kernels is 18, 36, 72 and 144 in sequence, each convolution layer is connected with a largest pooling layer with the size of 2 and the stride of 1, the number of bidirectional LSTM units is 256, the ReLU activation function is used for each layer, and the number of units of the last full-connection layer is 95. The preprocessed and dimensionality reduced data set is used as an input of a time sequence convolution network, and the time step is set to be 8.
The invention can analyze the road monitoring data with long-term and multi-dimension by using the road sensor monitoring data obtained by an automatic road monitoring system for a long time through a deep learning method, so as to predict and analyze the strain of a road base layer, and meanwhile, the learning weight trained in each time period is transferred to the training in the next time period, thereby improving the accuracy and stability of a prediction model and reducing the overfitting. In addition, the invention can effectively reduce labor cost and time loss, and the processed data can lay a foundation for the analysis and excavation of the follow-up monitoring data and the training of the prediction model.
Drawings
Fig. 1 is a schematic diagram of a time-series convolutional network structure.
Fig. 2 is a process diagram of a method implementation.
Fig. 3 is a correlation analysis heat map.
Fig. 4 is a diagram of an input dataset description.
Fig. 5 is a graph of the predictive effect of coupling two-way LSTM and CNN networks.
Detailed Description
The original road surface sensor monitoring data set adopted by the invention is the data acquired by an automatic road monitoring system. The specific implementation steps are as follows:
(1) Raw monitoring data preprocessing
Due to the problems of systematic errors such as data loss and temporary faults of collected equipment during data transmission, the problems of repetition, deletion, jump and the like of original data can greatly influence the network prediction accuracy. Therefore, the original data acquired by the sensor is deleted for the repeated value and the jump value, and the missing value is supplemented. In order to ensure the time continuity and consistency of the monitoring data, the data of the proper nine time periods are selected, and the time intervals are unified into one hour, so that the time sequence data is more in line with the characteristics of time sequence data, and the model training is facilitated.
(2) Feature correlation analysis
The three-way strain gauge monitoring data located on the base layer part is an actual measurement value of predicted target data, namely the base layer strain value, all data obtained by the monitoring system comprise a plurality of characteristics, and the exploration of the relation between the characteristics and the degree of association between the characteristics and the predicted target is critical. The invention primarily envisages that the base layer strain is predicted by several factors of asphalt layer strain, soil pressure, osmotic pressure, water content, temperature and soil layer strain, and correctly reducing the number of some characteristics provides certain help for model operation, such as improving accuracy, reducing risk of overfitting, etc. It is therefore necessary to feed the model with correlation analysis between the input data. Firstly, as the dimension of the monitoring data acquired by each sensor is huge, only the data of the first three monitoring positions acquired by each sensor are selected for correlation analysis; then, the correlation between the features is measured using pearson correlation coefficients, resulting in a correlation analysis heat map, as shown in fig. 3. The closer the pearson correlation coefficient value is to 1 or-1, the more positive and negative correlation is provided. Finally, according to the analysis result, the soil layer strain is displayed to have little correlation with other characteristics, and is removed from the input end.
(3) Data set generation
To meet the need of supervised learning, the pavement monitoring dataset is prepared according to a training set and a test set ratio of about 8:2, dividing. The input dataset description is shown in fig. 4. To monitor the training effect of the model during training, 20% was divided from the training set as the validation set.
(5) BiLSTM-CNN network-based strain prediction analysis
The BiLSTM-CNN network adopted by the invention utilizes a one-dimensional convolution structure to extract the invariant feature in the time dimension of the monitoring data; the bidirectional LSTM structure is utilized to capture the bidirectional semantic dependency relationship of the historical data and learn the mapping relationship among the features, the data can be reversed, and the hidden layer synthesizes the forward and reverse information, so that cells in the network can obtain the context information at the same time. The invention combines the convolution layer with the extraction fixed property and the bidirectional LSTM, thereby better retaining the long-term dependency relationship and the historical characteristic information, improving the model training effect and improving the robustness and the accuracy of the prediction model.
The BiLSTM-CNN network consists of 4 one-dimensional convolutional layers, one layer of bidirectional LSTM and two full-connection layers, which are 7 layers in total. The one-dimensional convolution kernels are 2 in size, the number of the one-dimensional convolution kernels is 18, 36, 72 and 144 in sequence, each layer uses a ReLU activation function, each one-dimensional convolution layer is connected with a largest pooling layer with the size of 2 and the stride of 1, and four one-dimensional convolution layers are connected with a two-way LSTM layer with the unit number of 256. At this time, the input data is extracted by the network through the fixed features in the input data after one-dimensional convolution of each layer, and after each layer passes through the ReLU activation function, the input data passes through an up-sampling layer (UpSampling) with a stride of 2, and then history information of long-term dependency and data context is learned in the bidirectional LSTM structure. And finally, obtaining the prediction information at the full connection layer, and taking the prediction information obtained in sequence as output by an output end. The final predicted effect is shown in fig. 5.
The ReLU function used in the present invention is expressed as follows:
In the training process of coupling the bidirectional LSTM and the convolution structure network to the pavement monitoring data, the input multidimensional data and the target data are trained together, the physical relationship between the data characteristics and the data is continuously learned, and the optimization target is to minimize the error of the predicted value and the measured value. An adaptive learning rate adjustment algorithm Adadelta (ADAPTIVE LEARNING RATE Method) is used as an optimization algorithm for gradient descent in the back propagation process, and the algorithm has the advantage of being capable of adaptively adjusting the learning rate when the gradient descends without manual setting.
For the learning method, model parameters are optimized by using an Adam method, and the Adam method is a simple random objective function gradient optimization algorithm with high calculation efficiency. This approach has two advantages in terms of handling sparse gradients and handling non-stationary targets. Adam is used in the present invention because it can be well suited to a wide range of non-convex optimization problems.
Adam maintains the trend of the average square gradient vt in the past to decay exponentially. It also has an average value of past gradients mt of exponential decay trend and a preference for a flat minimum value in the error plane. Then, the past attenuation average value and the past square gradients m t and v t are calculated as follows, respectively:
mt＝β1mt-1+(1-β1)gt (2)
Where m t and v t are estimates of the first moment (mean) and second moment (no center variance) of the gradient, respectively. The algorithm keeps the random gradient descent of the multidimensional data, keeps a single learning rate and updates all weights in the time sequence convolution network.
Since m t and v t initialize to vectors of 0, they are biased toward 0, these biases can be calculated as:
these t are then used and the update parameters are:
Beta 1 defaults to 0.9, beta 2 defaults to 0.999 and E defaults to 10 -8. Each epoch is the entire process of training the neural network through the entire data set once, including forward and backward. The learning rate in the present invention was 0.002.
Through observation, the loss value is continuously reduced along with the increase of the iteration times, and after training for 100 generations, the predicted value of the BiLSTM-CNN network can be well fitted with the measured value.
Claims (3)
1. A road base strain analysis method based on a coupling bidirectional LSTM and a convolution structure network is characterized by comprising the following steps of: the road surface history monitoring data is obtained by utilizing an automatic road monitoring system, and the key data is predicted by combining a deep learning method, and the specific steps are as follows:
step one: preprocessing original monitoring data;
firstly, integrating road monitoring data acquired by various sensors;
Secondly, according to the time span of the data acquired by the road monitoring system, deleting the data repetition value and the jump value after selecting the data of the time period, and completing the supplement of the missing value;
Finally, unifying the time intervals among all the data points of the time sequence data to be 1 hour;
Step two: analyzing the characteristic correlation;
Firstly, carrying out correlation analysis on the first three data points acquired by each sensor after processing, and measuring the correlation between features according to the Pearson correlation coefficient; secondly, removing the soil layer strain due to extremely small correlation;
Step three: creating a data set;
Averaging all monitoring data points of asphalt strain, osmotic pressure, soil moisture and base three-way strain, dividing the temperature into three parts of a surface layer, a base layer and a soil base according to the embedded position of the temperature sensor, and averaging; finally, the monitoring data set is set to 8 according to the proportion of the training set to the testing set: 2, dividing;
step four: predicting the strain of a time sequence convolution network base layer;
Extracting the invariant features in the time dimension of the monitoring data by using a BiLSTM-CNN network by using a one-dimensional convolution structure; biLSTM-CNN network is composed of 4 one-dimensional convolution layers, one layer of bidirectional LSTM and two layers of full connection layers; the convolution kernel sizes of the one-dimensional convolution layers are 2, and the numbers of the one-dimensional convolution layers are 18, 36, 72 and 144 respectively; a largest pooling layer with the size of 2 and the stride of 1 is connected, and each layer uses a ReLU activation function; the preprocessed and dimension reduced dataset is taken as input to BiLSTM-CNN network, and the time step is set to 8.
2. The method for analyzing the strain of the road base based on the coupled bidirectional LSTM and the convolution structure network according to claim 1, wherein in the training process of the BiLSTM-CNN network on the road surface monitoring data, the input multidimensional data and the target data are trained together, the physical relationship between the data characteristics and the data is continuously learned, and the optimization target is to minimize the error of the predicted value and the measured value; an adaptive learning rate adjustment algorithm Adadelta is used as an optimization algorithm for gradient descent in the back propagation process.
3. The method for analyzing the strain of the road base based on the coupled bidirectional LSTM and the convolution structure network according to claim 1, wherein the BiLSTM-CNN network is adopted for processing the monitoring data: (1) acquiring context information of data by adopting a bidirectional LSTM; (2) Adding a one-dimensional convolution layer to extract invariant features in the monitoring data; (3) a one-dimensional convolution layer is followed by a maximum pooling layer; (4) all layers use the ReLU activation function.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202110877678.8A CN114048790B (en) | 2021-08-01 | 2021-08-01 | Road base strain analysis method based on coupling bidirectional LSTM and convolution structure network |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202110877678.8A CN114048790B (en) | 2021-08-01 | 2021-08-01 | Road base strain analysis method based on coupling bidirectional LSTM and convolution structure network |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN114048790A CN114048790A (en) | 2022-02-15 |
| CN114048790B true CN114048790B (en) | 2024-05-24 |
Family
ID=80204457
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202110877678.8A Active CN114048790B (en) | 2021-08-01 | 2021-08-01 | Road base strain analysis method based on coupling bidirectional LSTM and convolution structure network |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN114048790B (en) |
Families Citing this family (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN114413832B (en) * | 2021-12-02 | 2023-12-15 | 中国通信建设第三工程局有限公司 | Road monitoring system and method based on optical fiber sensing |
| CN115077474A (en) * | 2022-06-14 | 2022-09-20 | 天津华北地质勘查总院 | Ground settlement trend prediction method and system based on machine learning |
| CN115081706B (en) * | 2022-06-16 | 2023-04-28 | 中国安能集团第三工程局有限公司 | Loess collapse prediction method and device based on bidirectional LSTM network |
| CN115329812B (en) * | 2022-08-10 | 2023-07-21 | 贵州桥梁建设集团有限责任公司 | Bridge infrastructure anomaly monitoring method based on artificial intelligence |
Citations (8)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN106355330A (en) * | 2016-08-31 | 2017-01-25 | 郑州航空工业管理学院 | Multi-response parameter optimization method based on radial basis function neural network prediction model |
| CN111209620A (en) * | 2019-12-30 | 2020-05-29 | 浙江大学 | LSTM-cGAN-based prediction method for residual bearing capacity and crack propagation path of crack-containing structure |
| CN112001177A (en) * | 2020-08-24 | 2020-11-27 | 浪潮云信息技术股份公司 | Electronic medical record named entity identification method and system integrating deep learning and rules |
| CN112069831A (en) * | 2020-08-21 | 2020-12-11 | 三峡大学 | Unreal information detection method based on BERT model and enhanced hybrid neural network |
| CN112733442A (en) * | 2020-12-31 | 2021-04-30 | 交通运输部公路科学研究所 | Road surface long-term performance prediction model based on deep learning and construction method thereof |
| CN112747779A (en) * | 2020-12-23 | 2021-05-04 | 武汉理工光科股份有限公司 | Asphalt pavement intelligent monitoring system based on fiber bragg grating sensing |
| CN112819136A (en) * | 2021-01-20 | 2021-05-18 | 南京邮电大学 | Time sequence prediction method and system based on CNN-LSTM neural network model and ARIMA model |
| CN112953862A (en) * | 2021-02-07 | 2021-06-11 | 哈尔滨工程大学 | High-speed railway scene channel estimation method based on deep learning |
-
2021
- 2021-08-01 CN CN202110877678.8A patent/CN114048790B/en active Active
Patent Citations (8)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN106355330A (en) * | 2016-08-31 | 2017-01-25 | 郑州航空工业管理学院 | Multi-response parameter optimization method based on radial basis function neural network prediction model |
| CN111209620A (en) * | 2019-12-30 | 2020-05-29 | 浙江大学 | LSTM-cGAN-based prediction method for residual bearing capacity and crack propagation path of crack-containing structure |
| CN112069831A (en) * | 2020-08-21 | 2020-12-11 | 三峡大学 | Unreal information detection method based on BERT model and enhanced hybrid neural network |
| CN112001177A (en) * | 2020-08-24 | 2020-11-27 | 浪潮云信息技术股份公司 | Electronic medical record named entity identification method and system integrating deep learning and rules |
| CN112747779A (en) * | 2020-12-23 | 2021-05-04 | 武汉理工光科股份有限公司 | Asphalt pavement intelligent monitoring system based on fiber bragg grating sensing |
| CN112733442A (en) * | 2020-12-31 | 2021-04-30 | 交通运输部公路科学研究所 | Road surface long-term performance prediction model based on deep learning and construction method thereof |
| CN112819136A (en) * | 2021-01-20 | 2021-05-18 | 南京邮电大学 | Time sequence prediction method and system based on CNN-LSTM neural network model and ARIMA model |
| CN112953862A (en) * | 2021-02-07 | 2021-06-11 | 哈尔滨工程大学 | High-speed railway scene channel estimation method based on deep learning |
Also Published As
| Publication number | Publication date |
|---|---|
| CN114048790A (en) | 2022-02-15 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN114048790B (en) | Road base strain analysis method based on coupling bidirectional LSTM and convolution structure network | |
| CN115758212B (en) | Mechanical equipment fault diagnosis method based on parallel network and transfer learning | |
| CN111784070A (en) | Intelligent landslide short-term early warning method based on XGboost algorithm | |
| CN112070322B (en) | High-voltage cable line running state prediction method based on long-short term memory network | |
| CN113128113A (en) | Poor information building load prediction method based on deep learning and transfer learning | |
| CN115271186B (en) | Reservoir water level prediction and early warning method based on delay factor and PSO RNN Attention model | |
| CN114169638A (en) | Water quality prediction method and device | |
| CN115796059A (en) | Electrical equipment service life prediction method and system based on deep learning | |
| CN113988210A (en) | Method and device for restoring distorted data of structure monitoring sensor network and storage medium | |
| CN114399073A (en) | Ocean surface temperature field prediction method based on deep learning | |
| CN110851897B (en) | Aqueduct stress-strain prediction method under multi-factor association | |
| CN117809164A (en) | Substation equipment fault detection method and system based on multi-mode fusion | |
| CN116662920B (en) | Abnormal data identification method, system, equipment and medium for drilling and blasting method construction equipment | |
| CN117521512A (en) | Bearing residual service life prediction method based on multi-scale Bayesian convolution transducer model | |
| CN116663126A (en) | Bridge temperature effect prediction method based on channel attention BiLSTM model | |
| CN115455791B (en) | Method for improving landslide displacement prediction accuracy based on numerical simulation technology | |
| CN116364203A (en) | Water quality prediction method, system and device based on deep learning | |
| CN115983333A (en) | Intelligent shield tunneling speed prediction method based on hybrid neural network | |
| CN116070126A (en) | Aviation plunger pump oil distribution disc abrasion detection method and system based on countermeasure self-supervision | |
| CN112560252B (en) | Method for predicting residual life of aeroengine | |
| CN114048835A (en) | Road base strain analysis method based on time sequence convolution network | |
| CN113591371A (en) | Bridge cluster structure damage positioning method based on space-time correlation model | |
| CN113756786A (en) | Method for predicting time sequence indicator diagram of rod-pumped well based on deep learning | |
| CN117709488B (en) | Dam seepage prediction method based on RUN-XGBoost | |
| CN117421601B (en) | Sea surface evaporation waveguide near-future rapid forecasting method |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |