CN118015839B - Expressway road domain risk prediction method and device - Google Patents
Expressway road domain risk prediction method and device Download PDFInfo
- Publication number
- CN118015839B CN118015839B CN202410413853.1A CN202410413853A CN118015839B CN 118015839 B CN118015839 B CN 118015839B CN 202410413853 A CN202410413853 A CN 202410413853A CN 118015839 B CN118015839 B CN 118015839B
- Authority
- CN
- China
- Prior art keywords
- data
- expressway
- risk prediction
- domain risk
- road domain
- 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
- 238000000034 method Methods 0.000 title claims abstract description 67
- 238000013058 risk prediction model Methods 0.000 claims abstract description 58
- 238000005457 optimization Methods 0.000 claims abstract description 31
- 238000013179 statistical model Methods 0.000 claims abstract description 22
- 238000012549 training Methods 0.000 claims description 69
- 238000013528 artificial neural network Methods 0.000 claims description 36
- 238000012360 testing method Methods 0.000 claims description 28
- 230000004927 fusion Effects 0.000 claims description 26
- 238000000605 extraction Methods 0.000 claims description 20
- 238000012706 support-vector machine Methods 0.000 claims description 19
- 238000012795 verification Methods 0.000 claims description 19
- 238000007781 pre-processing Methods 0.000 claims description 17
- 230000015654 memory Effects 0.000 claims description 16
- 238000012544 monitoring process Methods 0.000 claims description 16
- 238000012545 processing Methods 0.000 claims description 16
- 238000004590 computer program Methods 0.000 claims description 15
- 238000013527 convolutional neural network Methods 0.000 claims description 10
- 238000004458 analytical method Methods 0.000 claims description 9
- 238000004422 calculation algorithm Methods 0.000 claims description 9
- 238000003860 storage Methods 0.000 claims description 8
- 238000004140 cleaning Methods 0.000 claims description 6
- 125000004122 cyclic group Chemical group 0.000 claims description 6
- 238000004364 calculation method Methods 0.000 claims description 5
- 238000010200 validation analysis Methods 0.000 claims description 3
- 206010039203 Road traffic accident Diseases 0.000 description 16
- 230000000875 corresponding effect Effects 0.000 description 10
- 238000010586 diagram Methods 0.000 description 10
- 238000011002 quantification Methods 0.000 description 6
- 230000000903 blocking effect Effects 0.000 description 5
- 230000006870 function Effects 0.000 description 5
- 230000007246 mechanism Effects 0.000 description 5
- 230000008569 process Effects 0.000 description 4
- 238000011160 research Methods 0.000 description 4
- 238000013136 deep learning model Methods 0.000 description 3
- 238000013461 design Methods 0.000 description 3
- 238000010801 machine learning Methods 0.000 description 3
- 238000012502 risk assessment Methods 0.000 description 3
- 230000003068 static effect Effects 0.000 description 3
- 238000012800 visualization Methods 0.000 description 3
- 230000008859 change Effects 0.000 description 2
- 238000004891 communication Methods 0.000 description 2
- 238000010219 correlation analysis Methods 0.000 description 2
- 238000002790 cross-validation Methods 0.000 description 2
- 238000013135 deep learning Methods 0.000 description 2
- 238000005516 engineering process Methods 0.000 description 2
- 239000000284 extract Substances 0.000 description 2
- 239000000463 material Substances 0.000 description 2
- 238000010606 normalization Methods 0.000 description 2
- 238000007500 overflow downdraw method Methods 0.000 description 2
- 238000000513 principal component analysis Methods 0.000 description 2
- 238000007637 random forest analysis Methods 0.000 description 2
- 230000000306 recurrent effect Effects 0.000 description 2
- 230000006403 short-term memory Effects 0.000 description 2
- 230000009466 transformation Effects 0.000 description 2
- 230000000007 visual effect Effects 0.000 description 2
- 230000002159 abnormal effect Effects 0.000 description 1
- 230000003044 adaptive effect Effects 0.000 description 1
- 230000009286 beneficial effect Effects 0.000 description 1
- 238000006243 chemical reaction Methods 0.000 description 1
- 238000010276 construction Methods 0.000 description 1
- 230000002596 correlated effect Effects 0.000 description 1
- 238000001514 detection method Methods 0.000 description 1
- 238000009826 distribution Methods 0.000 description 1
- 230000000694 effects Effects 0.000 description 1
- 238000011156 evaluation Methods 0.000 description 1
- 230000010354 integration Effects 0.000 description 1
- 230000003993 interaction Effects 0.000 description 1
- 238000004519 manufacturing process Methods 0.000 description 1
- 238000013507 mapping Methods 0.000 description 1
- 238000012821 model calculation Methods 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 238000012806 monitoring device Methods 0.000 description 1
- 230000003287 optical effect Effects 0.000 description 1
- 230000008092 positive effect Effects 0.000 description 1
- 239000002243 precursor Substances 0.000 description 1
- 238000012216 screening Methods 0.000 description 1
- 230000011218 segmentation Effects 0.000 description 1
- 238000010187 selection method Methods 0.000 description 1
- 238000007619 statistical method Methods 0.000 description 1
- 230000036962 time dependent Effects 0.000 description 1
- 238000011426 transformation method Methods 0.000 description 1
- 230000001960 triggered effect Effects 0.000 description 1
- 239000013598 vector Substances 0.000 description 1
- XLYOFNOQVPJJNP-UHFFFAOYSA-N water Substances O XLYOFNOQVPJJNP-UHFFFAOYSA-N 0.000 description 1
Landscapes
- Traffic Control Systems (AREA)
Abstract
The invention provides a highway road domain risk prediction method and device, and relates to the technical field of intelligent traffic. The method comprises the following steps: acquiring multi-source data related to road domain risk prediction of the expressway; performing road domain risk prediction on the multi-source data based on a preset expressway road domain risk prediction model to obtain an expressway road domain risk prediction result; and obtaining a road domain risk optimization prediction result of the expressway according to the output result of the statistical model and the expressway road domain risk prediction result which are obtained in advance. The apparatus performs the above method. The expressway domain risk prediction method and the expressway domain risk prediction device can improve accuracy of expressway domain risk prediction.
Description
Technical Field
The invention relates to the technical field of intelligent traffic, in particular to a highway road domain risk prediction method and device.
Background
Road domain risk is caused by the superposition of multiple risk factors acting together. Risk factor superposition refers to the simultaneous existence of multiple risk factors in terms of 'people, vehicles, roads and environment' in a time period, the risk factors commonly act on a main body, and different risk factors can have linear or nonlinear superposition in the acting process. The superposition of these different risk factors creates a road risk, which in turn leads to traffic accidents.
Currently, in the method of predicting the risk of highway accidents, there are mainly two types: one is to construct a statistical model based on historical traffic accident data to macroscopically predict the occurrence frequency of future traffic accidents; another is to predict the possibility of a short term (around 5 minutes) traffic accident based on dangerous traffic flow data and historical traffic accident data. The method for predicting the road domain risk based on the historical accident occurrence frequency is mostly a statistical method, is a static prediction method, reflects the macroscopic traffic accident risk situation, and has low instantaneity. The method for predicting the road domain risk based on traffic flow data and historical accident data is a machine learning method, is a dynamic prediction method, reflects the current microscopic traffic risk situation, has high real-time performance, and is used for predicting the situation in a short time (generally about 5 minutes) and can not predict the situation for a long time in most researches. Through analysis of the current expressway road domain risk prediction method, the following technical problems can be found:
① The nonlinear negative two-term regression, the multiple nonlinear regression, the time sequence model, the Tobit model and other mathematical statistics methods are mostly used for identifying dangerous factors causing accidents through historical accident data of the expressways, so as to determine accident black spots, evaluate whether expressway safety management measures are proper or not, and are not suitable for predicting road domain risks and road domain risk factors possibly causing traffic accidents.
② The machine learning algorithms such as random forest, support vector machine, logistics regression, bayesian learner, neural network and the like basically assume that factors such as real-time traffic flow, road geometric linearity, weather conditions and the like are main risk factors of expressway road domain risks, can identify precursors of accident occurrence in a dynamic traffic system, and further predict whether traffic accidents occur in a given road section within a short time (generally 5 minutes); the problem of most of researches is that accident risks are equal to accident occurrence, quantitative researches on highway road domain risks are ignored, so that the magnitude of the road domain risks cannot be estimated according to a prediction result, and main risk factors are interfered, so that the aim of reducing the highway road domain risks to an acceptable degree is fulfilled.
Disclosure of Invention
Aiming at the problems in the prior art, the embodiment of the invention provides a method and a device for predicting the risk of a highway road domain, which can at least partially solve the problems in the prior art.
In one aspect, the invention provides a highway road domain risk prediction method, which comprises the following steps:
acquiring multi-source data related to road domain risk prediction of the expressway;
performing road domain risk prediction on the multi-source data based on a preset expressway road domain risk prediction model to obtain an expressway road domain risk prediction result;
The expressway domain risk prediction model is obtained by training a combined model according to expressway domain risk prediction sample data, wherein the combined model takes a fully-connected neural network as a meta classifier, a Bayesian network, a support vector machine and K neighbors as base classifiers, and each base classifier is respectively connected with the meta classifier;
And obtaining a road domain risk optimization prediction result of the expressway according to the output result of the statistical model and the expressway road domain risk prediction result which are obtained in advance.
The method for predicting the road domain risk of the multi-source data based on the preset expressway road domain risk prediction model comprises the following steps:
And carrying out data preprocessing and feature extraction on the multi-source data, and carrying out road domain risk prediction on the extracted features based on the preset expressway road domain risk prediction model to obtain an expressway road domain risk prediction result.
Training a combined model according to expressway road domain risk prediction sample data to obtain an expressway road domain risk prediction model, wherein the expressway road domain risk prediction model comprises the following steps:
preprocessing the acquired data;
extracting features according to the preprocessed data;
performing feature optimization on the extracted features, and performing feature fusion on the features after feature optimization;
and training a combined model according to the characteristics after the characteristic fusion to obtain the expressway road domain risk prediction model.
Wherein, the preprocessing the collected data comprises:
Collecting data to obtain a data set; the dataset includes a first type of data having an unstructured data type and a second type of data having a log data type;
Carrying out structural processing on the first type of data, and carrying out analysis and secondary calculation on the second type of data to obtain traffic flow data;
and cleaning and formatting the data in the data set to obtain standardized data, and dividing the standardized data to obtain a training set, a verification set and a test set.
The feature extraction according to the preprocessed data comprises the following steps:
oversampling is performed on the data in the training set, and no oversampling is performed on the data in the test set;
And selecting the features to be selected which are relevant to road domain risk prediction from the training set, the verification set and the test set.
Wherein the dataset comprises video surveillance image data, the traffic flow data and structured data; accordingly, the selecting the candidate features related to road domain risk prediction from the training set, the verification set and the test set includes:
Performing feature extraction on a first feature to be selected corresponding to the video monitoring image data by using a convolutional neural network;
Extracting features of a second to-be-selected feature corresponding to the traffic flow data by using a cyclic neural network or a long-short-term memory network;
and extracting the characteristics of the third candidate characteristics corresponding to the structured data by using a fully connected neural network or an embedded layer.
The method for obtaining the expressway road domain risk prediction model according to the feature training combination model after feature fusion comprises the following steps:
training each base classifier with the training set, and training the meta classifier with the verification set;
adding a plurality of output layers into the combined model, and simultaneously predicting prediction results of different combinations;
The combined model is trained using a back propagation algorithm and a gradient descent optimizer.
In one aspect, the present invention provides a highway road domain risk prediction apparatus, including:
the acquisition unit is used for acquiring multi-source data related to road domain risk prediction of the expressway;
the prediction unit is used for predicting the road domain risk of the multi-source data based on a preset expressway road domain risk prediction model to obtain an expressway road domain risk prediction result;
The expressway domain risk prediction model is obtained by training a combined model according to expressway domain risk prediction sample data, wherein the combined model takes a fully-connected neural network as a meta classifier, a Bayesian network, a support vector machine and K neighbors as base classifiers, and each base classifier is respectively connected with the meta classifier;
And the optimization unit is used for obtaining the road domain risk optimization prediction result of the expressway according to the output result of the statistical model and the expressway road domain risk prediction result which are obtained in advance.
In still another aspect, an embodiment of the present invention provides an electronic device, including: a processor, a memory, and a bus, wherein,
The processor and the memory complete communication with each other through the bus;
The memory stores program instructions executable by the processor, the processor invoking the program instructions capable of performing the method of:
acquiring multi-source data related to road domain risk prediction of the expressway;
performing road domain risk prediction on the multi-source data based on a preset expressway road domain risk prediction model to obtain an expressway road domain risk prediction result;
The expressway domain risk prediction model is obtained by training a combined model according to expressway domain risk prediction sample data, wherein the combined model takes a fully-connected neural network as a meta classifier, a Bayesian network, a support vector machine and K neighbors as base classifiers, and each base classifier is respectively connected with the meta classifier;
And obtaining a road domain risk optimization prediction result of the expressway according to the output result of the statistical model and the expressway road domain risk prediction result which are obtained in advance.
Embodiments of the present invention provide a non-transitory computer readable storage medium comprising:
the non-transitory computer readable storage medium stores computer instructions that cause the computer to perform the method of:
acquiring multi-source data related to road domain risk prediction of the expressway;
performing road domain risk prediction on the multi-source data based on a preset expressway road domain risk prediction model to obtain an expressway road domain risk prediction result;
The expressway domain risk prediction model is obtained by training a combined model according to expressway domain risk prediction sample data, wherein the combined model takes a fully-connected neural network as a meta classifier, a Bayesian network, a support vector machine and K neighbors as base classifiers, and each base classifier is respectively connected with the meta classifier;
And obtaining a road domain risk optimization prediction result of the expressway according to the output result of the statistical model and the expressway road domain risk prediction result which are obtained in advance.
The expressway domain risk prediction method and the expressway domain risk prediction device provided by the embodiment of the invention acquire multi-source data related to expressway domain risk prediction; performing road domain risk prediction on the multi-source data based on a preset expressway road domain risk prediction model to obtain an expressway road domain risk prediction result; the expressway domain risk prediction model is obtained by training a combined model according to expressway domain risk prediction sample data, wherein the combined model takes a fully-connected neural network as a meta classifier, a Bayesian network, a support vector machine and K neighbors as base classifiers, and each base classifier is respectively connected with the meta classifier; according to the output result of the statistical model and the highway road domain risk prediction result which are obtained in advance, the road domain risk optimization prediction result of the highway is obtained, and the accuracy of road domain risk prediction of the highway can be improved.
Drawings
In order to more clearly illustrate the embodiments of the invention or the technical solutions in the prior art, the drawings that are required in the embodiments or the description of the prior art will be briefly described, it being obvious that the drawings in the following description are only some embodiments of the invention, and that other drawings may be obtained according to these drawings without inventive effort for a person skilled in the art. In the drawings:
fig. 1 is a flowchart of a highway road domain risk prediction method according to an embodiment of the present invention.
Fig. 2 is a flowchart of a highway road domain risk prediction method according to another embodiment of the present invention.
Fig. 3 is a flowchart illustrating a method for predicting risk of an expressway domain according to another embodiment of the present invention.
Fig. 4 is a schematic structural diagram of an expressway road domain risk prediction device according to an embodiment of the invention.
Fig. 5 is a schematic diagram of an entity structure of an electronic device according to an embodiment of the present invention.
Detailed Description
For the purpose of making the objects, technical solutions and advantages of the embodiments of the present application more apparent, the embodiments of the present application will be described in further detail with reference to the accompanying drawings. The exemplary embodiments of the present application and their descriptions herein are for the purpose of explaining the present application, but are not to be construed as limiting the application. It should be noted that, without conflict, the embodiments of the present application and features of the embodiments may be arbitrarily combined with each other.
Fig. 1 is a flow chart of a highway road domain risk prediction method according to an embodiment of the present invention, and as shown in fig. 1, the highway road domain risk prediction method according to the embodiment of the present invention includes:
Step S1: and acquiring multi-source data related to road domain risk prediction of the expressway.
Step S2: performing road domain risk prediction on the multi-source data based on a preset expressway road domain risk prediction model to obtain an expressway road domain risk prediction result;
The expressway road domain risk prediction model is obtained by training a combined model according to expressway road domain risk prediction sample data, the combined model is obtained by taking a fully connected neural network as a meta classifier, taking a Bayesian network, a support vector machine and K nearest neighbors as base classifiers, and each base classifier is respectively connected with the meta classifier.
Step S3: and obtaining a road domain risk optimization prediction result of the expressway according to the output result of the statistical model and the expressway road domain risk prediction result which are obtained in advance.
In the above step S1, the apparatus acquires multi-source data related to road domain risk prediction of the expressway. The apparatus may be a computer device performing the method and may comprise, for example, a server. It should be noted that, the data acquisition and analysis according to the embodiments of the present invention are authorized by the user.
The multi-source data may include collecting real-time data with a highway toll gate industrial computer, ETC portal, camera, weather monitoring IoT devices, the real-time data including video monitoring data, toll gate lane flow data, ETC portal log data, weather monitoring device data, and the like;
The multi-source data can also comprise non-real-time data collected through road engineering design drawings, road condition detection vehicles, road blocking report forms and traffic accident report forms, wherein the non-real-time data comprise road geometric linear data, road surface anti-skid performance data, traffic accident data, road blocking data and the like.
In the step S2, the device performs road domain risk prediction on the multi-source data based on a preset expressway road domain risk prediction model to obtain an expressway road domain risk prediction result; as shown in fig. 2, the multi-source data may be input to a preset expressway road domain risk prediction model (real-time model), and an output result of the preset expressway road domain risk prediction model is used as an expressway road domain risk prediction result.
The expressway road domain risk prediction model is obtained by training a combined model according to expressway road domain risk prediction sample data, the combined model is obtained by taking a fully connected neural network as a meta classifier, taking a Bayesian network, a support vector machine and K nearest neighbors as base classifiers, and each base classifier is respectively connected with the meta classifier. The structure of the combined model is shown as "model build" in fig. 3.
In the step S3, the device obtains a road domain risk optimization prediction result of the expressway according to the output result of the statistical model and the road domain risk prediction result of the expressway, which are obtained in advance. The statistical model, i.e. the statistical model constructed based on historical traffic accident data, can be obtained by adopting the prior art means. The output result of the statistical model may represent a static highway road domain risk prediction result.
Corresponding output result weights can be respectively set for the output result of the statistical model and the expressway road domain risk prediction result, so that the road domain risk optimization prediction result of the expressway comprehensively considering static and dynamic is obtained.
The prediction using the trained model is described as follows:
developing a data fusion system, unifying real-time data of different sources into a format, and rapidly processing the data;
Processing and analyzing the real-time data stream using a stream processing technique (APACHE KAFKA);
Based on real-time traffic data, outputting a highway road domain risk prediction result through model calculation;
And carrying out decision-stage data fusion on a statistical road domain risk prediction result based on historical accident data and a dynamic road domain risk prediction result based on real-time traffic flow data to obtain the road domain risk situation of the whole road section.
In a subsequent step, an early warning system can also be integrated and an automated intervention strategy developed to prevent accidents.
The risk quantification can be further performed on the expressway road domain risk prediction result, and the method is described as follows:
Risk quantification:
And establishing a risk measure model based on a mutual information theory, and quantifying the expressway road domain risk.
Risk assessment model:
① And constructing a highway accident risk measure model based on a mutual information theory, wherein Pi represents the stepping probability of accidents under different characteristic combinations, and after the number of the accident samples reaches a certain number, the number of the accident samples is equal to the proportion of the number of the accidents under each risk factor combination to the total accidents in value.
② The model predicts the probability of an incident occurrence based on the input data features.
Risk quantification:
① And quantifying the risk into a specific numerical value according to the output of the risk assessment model, wherein the numerical value represents the risk level.
Map and visualization:
① And mapping the risk quantification result to a specific position of the expressway based on a Geographic Information System (GIS) by adopting a dynamic segmentation technology, and displaying the risk level of the whole road network so that expressway operation management staff can intuitively know the risk distribution.
Updating and monitoring in real time:
and collecting new data in real time, generating new data characteristics, iterating the risk assessment model, and monitoring the change of the risk level.
The step of predicting the road domain risk of the multi-source data based on a preset expressway road domain risk prediction model to obtain an expressway road domain risk prediction result comprises the following steps:
And carrying out data preprocessing and feature extraction on the multi-source data, and carrying out road domain risk prediction on the extracted features based on the preset expressway road domain risk prediction model to obtain an expressway road domain risk prediction result. The data preprocessing and the feature extraction are performed on the multi-source data, and the preprocessing and the related description of the feature extraction can be performed on the acquired data in the subsequent model training stage.
Training a combined model according to expressway road domain risk prediction sample data to obtain an expressway road domain risk prediction model, wherein the method comprises the following steps:
preprocessing the acquired data;
extracting features according to the preprocessed data;
performing feature optimization on the extracted features, and performing feature fusion on the features after feature optimization;
and training a combined model according to the characteristics after the characteristic fusion to obtain the expressway road domain risk prediction model.
The preprocessing of the acquired data comprises:
Collecting data to obtain a data set; the dataset includes a first type of data having an unstructured data type and a second type of data having a log data type;
Carrying out structural processing on the first type of data, and carrying out analysis and secondary calculation on the second type of data to obtain traffic flow data;
and cleaning and formatting the data in the data set to obtain standardized data, and dividing the standardized data to obtain a training set, a verification set and a test set.
Data preprocessing, described below:
carrying out structural processing on unstructured data such as road surface anti-skid performance data, road blocking data, traffic accident data, weather monitoring equipment data, video monitoring image data and the like;
Analyzing and secondarily calculating ETC portal log data to obtain traffic flow data including speed, flow and traffic density;
cleaning and formatting the data, removing abnormal values and missing values, and normalizing the data;
the data is divided into a training set, a verification set and a test set, so that the model can be generalized on unseen data.
The feature extraction according to the preprocessed data comprises the following steps:
oversampling is performed on the data in the training set, and no oversampling is performed on the data in the test set;
And selecting the features to be selected which are relevant to road domain risk prediction from the training set, the verification set and the test set.
The data set comprises video monitoring image data, traffic flow data and structured data; accordingly, the selecting the candidate features related to road domain risk prediction from the training set, the verification set and the test set includes:
Performing feature extraction on a first feature to be selected corresponding to the video monitoring image data by using a convolutional neural network;
Extracting features of a second to-be-selected feature corresponding to the traffic flow data by using a cyclic neural network or a long-short-term memory network;
and extracting the characteristics of the third candidate characteristics corresponding to the structured data by using a fully connected neural network or an embedded layer.
The feature extraction is described as follows:
① Data source analysis: analyzing the characteristics and the content of each data source, and determining the contribution of each data source to accident risk prediction;
② Sample design: designing a data sample by adopting a pairing case-control method, and solving the problem of unbalance of sample data by an SMOTE algorithm in order to improve the performance of a base classifier; it should be noted that only the training set data is oversampled, and the test set data is not oversampled to prevent overfitting, and the model is subsequently evaluated on the test set to ensure that the test set does not contain "synthetic" samples;
③ Feature selection: using a correlation analysis and information gain method to assist in selecting features (to-be-selected features) most relevant to accident risk prediction from the data samples;
④ Feature extraction:
For image data captured by a video surveillance camera, features are extracted using a Convolutional Neural Network (CNN). CNNs are able to automatically learn the characteristics of texture, shape, and color in an image.
For time series data of traffic flows, features are extracted using Recurrent Neural Networks (RNNs) or long short term memory networks (LSTM). These networks are capable of capturing time-dependent relationships and trends in data.
For other structured data such as traffic blocking, traffic accidents, weather, etc., the features are extracted using fully connected neural networks (FCNs) or embedded layers. These networks are able to learn patterns and associations in the data.
Convolutional Neural Networks (CNNs), cyclic neural networks (RNNs), and the like are trained to automatically learn features from raw data.
The model extracts more expressive and distinguishable features from the original data by multi-layer nonlinear transformation.
The extracted features are optimized, and the following description is given:
and optimizing the extracted features, including removing redundant features, merging similar features and the like.
And (3) optimizing a feature set by using a feature selection technology, such as a random forest feature selection method based on Principal Component Analysis (PCA), screening space-time features to obtain risk factors of expressway accident risk, and providing an evaluation index and a research basis for risk quantification.
Feature fusion is carried out on the features after feature optimization, and the following description is carried out:
The feature vectors from different data sources are directly fused by adopting a feature level fusion method in data fusion, and the method can keep the uniqueness of each feature and allow the model to learn the interaction among different features.
Feature transformation: the features to be fused are transformed using normalization, normalization and logarithmic transformation methods, making them more suitable for combination with other features.
Feature weighting: each feature is assigned a different feature weight based on its importance, and then the weighted values of these features are added.
Weight selection: and (3) evaluating the correlation among the features by adopting a correlation analysis method (the method adopts a Pierson correlation coefficient and a Szelman class correlation coefficient), selecting features highly correlated with the risk of the expressway accident, and giving higher weight.
Multimodal neural networks: using a multi-modal neural network, the Deep Belief Network (DBN), learns how to effectively fuse different features. The network can automatically learn the hierarchical structure and the combination mode of the features, so that the prediction performance of the model is improved.
The model was built into the training, described as follows:
① Generating: the purpose of this stage is to generate a set of base classifiers that contain the most suitable candidate classifiers for subsequent classifier selection and fusion. Based on the consideration of three key factors of the performance of the base classifier, the diversity among the classifiers and the fusion method (meta classifier) to be adopted, the invention adopts a Boosting algorithm to generate a classifier pool which comprises four classifiers of Bayesian Network (BN), support vector machine (SVN), K neighbor (KNN) and full-connected neural network (MLP).
② Selecting: the purpose of this stage is to determine a set of learners (fusion set) to improve the classification performance of the fusion model. The invention selects a full-connected neural network (MLP) algorithm with higher prediction performance as a meta-classifier, and other three classifiers as base classifiers. The most efficient classifier is selected mainly using SMOTE and cross-validation.
③ Fusion: the method is characterized in that the results of the base classifier are fused based on an optimal element classifier full-connected neural network (MLP), so that the overall error is minimum and the classification performance is highest, and the samples for training the base classifier are not applied to training the base classifier to prevent over fitting, the base classifier is mainly trained by adopting a training set, and the base classifier is trained based on a verification set.
④ Pre-training: pretraining is carried out on a large-scale data set, and fine tuning is carried out on a specific expressway road domain risk prediction task, so that the generalization capability of the model is improved.
⑤ Multiple output layers are added to the model using a multi-task learning architecture, with predictions for different combinations being predicted at the same time.
⑥ The model was trained using a back propagation algorithm and a gradient descent optimizer.
A multi-classifier system (MCS) is a machine learning technique that can provide higher performance results than using a single classifier. The invention designs a fusion framework to construct a traffic risk prediction model, and the model can fuse a single decision of a base classifier into a final decision.
Model verification and testing is described as follows:
① The performance of the model is evaluated over a validation set, using cross-validation techniques to avoid overfitting.
② The final performance of the model was evaluated on the test set using F1-score, G-mean and AUC, and the like.
The model deployment and monitoring is described as follows:
The trained model is deployed into an actual environment and used for predicting the risk of traffic accidents in real time, the performance of the model is continuously monitored, and the model can be automatically updated periodically according to new data by utilizing a self-adaptive learning mechanism.
The expressway road domain risk prediction method provided by the embodiment of the invention has the following advantages:
(1) The technical aspect of multi-source data fusion:
In the prior art, single or limited two three data sources are used for risk prediction, unstructured data and structured data can be processed at the same time, multiple types of data sources such as ETC (electronic toll collection) portal frame, toll station running water, monitoring video, meteorological information, road anti-skid performance coefficient, road blocking information, traffic accident information and the like are fused, and complex characteristic representation is automatically learned from original data by using an advanced deep learning algorithm, so that more comprehensive road domain risk prediction is provided.
(2) Feature extraction:
the invention reduces the dependence on artificial feature engineering, reduces the need of manual selection and construction of features and reduces the possibility of human error occurrence through automatic feature extraction.
Different deep learning models are adopted for extracting features for different types of data, and then the features are fused together to form a comprehensive feature set, so that the deep learning model can capture more potential relevance and modes, and the prediction accuracy is improved.
(3) End-to-end training:
The deep learning prediction model constructed by the invention and capable of processing multi-source data can directly learn from input data to output results without artificial characteristic engineering in the middle. The training mode can reduce errors possibly occurring in the process of feature selection and conversion, and improves training efficiency, generalization capability and prediction capability of the model.
(4) Risk prediction aspect:
Multitasking learning: according to the invention, a plurality of related tasks can be learned at the same time, on one hand, by sharing the characteristic representation of the bottom layer, the model can extract useful information among different tasks, so that the performance of each task is improved; on the other hand, road domain risks are predicted simultaneously through different tasks, and then prediction results of a plurality of related tasks are fused, so that the prediction accuracy of the model is improved.
Real-time dynamic prediction: the invention can respond to traffic condition change in real time, process new data in real time, dynamically adjust the model and predict the model, and adapt to the traffic condition continuously changing on the expressway.
(5) Adaptive learning mechanism:
in the prior art, a model may need to be updated manually, and the self-adaptive learning mechanism is introduced, so that the model can learn and optimize itself without manual intervention.
(6) Early warning and intervention system:
The prior art lacks an automatic early warning and intervention mechanism, the road domain risk and road domain risk factors possibly causing traffic accidents can be predicted, the size of the road domain risk can be estimated according to the prediction result through quantifying the road domain risk of the expressway, and then the road domain risk is early warned, and the main risk factors are intervened.
(7) Visual analysis tool:
the prior art may lack visual visualization tools, and the invention develops the easy-to-use expressway road domain risk visualization tools based on GIS, which can help traffic management personnel to better understand and utilize the prediction results.
The expressway road domain risk prediction method provided by the embodiment of the invention has the following beneficial technical effects:
(1) Accuracy rate is improved: the accuracy over the test set is improved over conventional models that use only a single or two three data sources.
(2) Real-time performance: the method can complete prediction in near real time after receiving new data, and basically meets the real-time requirement.
(3) In a specific expressway accident test case, the method successfully quantifies risks, and when the risk quantification result exceeds a threshold value, early warning is triggered, and the traditional model fails to respond.
The specific time range of the road domain risk which can be predicted by the invention depends on the quality and the computing resources of the multi-source heterogeneous data. Theoretically, the predicted time ranges from 5 minutes to 15 minutes, and even longer, and practical applications are limited by data quality and computational resources.
The key point of the method for enlarging the prediction time range is that:
1. A deep learning model such as a Recurrent Neural Network (RNN) or a long short term memory network (LSTM) capable of processing time series data is used, and time dependence and trend in the data are captured.
2. The model integration based on the fusion framework can effectively improve the accuracy and adaptability of prediction.
3. By utilizing the self-adaptive learning mechanism, the model can be automatically updated periodically according to new data, and the time range and accuracy of prediction are continuously improved.
4. And carrying out decision-making data fusion on the low-frequency statistical road domain risk prediction result based on the historical accident data and the high-frequency dynamic road domain risk prediction result based on the real-time traffic flow data.
Through the above implementation steps, advantages and positive effects, it can be determined that: the road domain risk prediction method based on the data fusion architecture has obvious advantages in the aspect of improving the accuracy of the accident risk prediction of the expressway, and overcomes the limitation of the prior art. The method brings new technical progress to the prediction of the accident risk of the expressway, and is helpful for improving traffic safety and reducing accident occurrence.
The expressway road domain risk prediction method provided by the embodiment of the invention acquires multi-source data related to the road domain risk prediction of the expressway; performing road domain risk prediction on the multi-source data based on a preset expressway road domain risk prediction model to obtain an expressway road domain risk prediction result; the expressway domain risk prediction model is obtained by training a combined model according to expressway domain risk prediction sample data, wherein the combined model takes a fully-connected neural network as a meta classifier, a Bayesian network, a support vector machine and K neighbors as base classifiers, and each base classifier is respectively connected with the meta classifier; according to the output result of the statistical model and the highway road domain risk prediction result which are obtained in advance, the road domain risk optimization prediction result of the highway is obtained, and the accuracy of road domain risk prediction of the highway can be improved.
Further, the predicting the road domain risk of the multi-source data based on the preset expressway road domain risk prediction model to obtain an expressway road domain risk prediction result, including:
And carrying out data preprocessing and feature extraction on the multi-source data, and carrying out road domain risk prediction on the extracted features based on the preset expressway road domain risk prediction model to obtain an expressway road domain risk prediction result. The description of the embodiments may be referred to above, and will not be repeated.
Further, training a combined model according to the expressway road domain risk prediction sample data to obtain an expressway road domain risk prediction model, including:
preprocessing the acquired data; the description of the embodiments may be referred to above, and will not be repeated.
Extracting features according to the preprocessed data; the description of the embodiments may be referred to above, and will not be repeated.
Performing feature optimization on the extracted features, and performing feature fusion on the features after feature optimization; the description of the embodiments may be referred to above, and will not be repeated.
And training a combined model according to the characteristics after the characteristic fusion to obtain the expressway road domain risk prediction model. The description of the embodiments may be referred to above, and will not be repeated.
Further, the preprocessing the collected data includes:
Collecting data to obtain a data set; the dataset includes a first type of data having an unstructured data type and a second type of data having a log data type; the description of the embodiments may be referred to above, and will not be repeated.
Carrying out structural processing on the first type of data, and carrying out analysis and secondary calculation on the second type of data to obtain traffic flow data; the description of the embodiments may be referred to above, and will not be repeated.
And cleaning and formatting the data in the data set to obtain standardized data, and dividing the standardized data to obtain a training set, a verification set and a test set. The description of the embodiments may be referred to above, and will not be repeated.
Further, the feature extraction according to the preprocessed data includes:
oversampling is performed on the data in the training set, and no oversampling is performed on the data in the test set; the description of the embodiments may be referred to above, and will not be repeated.
And selecting the features to be selected which are relevant to road domain risk prediction from the training set, the verification set and the test set. The description of the embodiments may be referred to above, and will not be repeated.
Further, the dataset includes video surveillance image data, the traffic flow data, and structured data; accordingly, the selecting the candidate features related to road domain risk prediction from the training set, the verification set and the test set includes:
Performing feature extraction on a first feature to be selected corresponding to the video monitoring image data by using a convolutional neural network; the description of the embodiments may be referred to above, and will not be repeated.
Extracting features of a second to-be-selected feature corresponding to the traffic flow data by using a cyclic neural network or a long-short-term memory network; the description of the embodiments may be referred to above, and will not be repeated.
And extracting the characteristics of the third candidate characteristics corresponding to the structured data by using a fully connected neural network or an embedded layer. The description of the embodiments may be referred to above, and will not be repeated.
Further, the training the combined model according to the features after feature fusion to obtain the expressway road domain risk prediction model includes:
training each base classifier with the training set, and training the meta classifier with the verification set; the description of the embodiments may be referred to above, and will not be repeated.
Adding a plurality of output layers into the combined model, and simultaneously predicting prediction results of different combinations; the description of the embodiments may be referred to above, and will not be repeated.
The combined model is trained using a back propagation algorithm and a gradient descent optimizer. The description of the embodiments may be referred to above, and will not be repeated.
Fig. 4 is a schematic structural diagram of an expressway road domain risk prediction device according to an embodiment of the present invention, and as shown in fig. 4, the expressway road domain risk prediction device according to an embodiment of the present invention includes an obtaining unit 401, a prediction unit 402, and an optimizing unit 403, where:
The acquiring unit 401 is configured to acquire multi-source data related to road domain risk prediction of an expressway; the prediction unit 402 is configured to perform a road domain risk prediction on the multi-source data based on a preset expressway road domain risk prediction model, so as to obtain an expressway road domain risk prediction result; the expressway domain risk prediction model is obtained by training a combined model according to expressway domain risk prediction sample data, wherein the combined model takes a fully-connected neural network as a meta classifier, a Bayesian network, a support vector machine and K neighbors as base classifiers, and each base classifier is respectively connected with the meta classifier; the optimizing unit 403 is configured to obtain a road domain risk optimization prediction result of the expressway according to the output result of the statistical model and the expressway road domain risk prediction result obtained in advance.
Specifically, the acquiring unit 401 in the device is configured to acquire multi-source data related to road domain risk prediction of the expressway; the prediction unit 402 is configured to perform a road domain risk prediction on the multi-source data based on a preset expressway road domain risk prediction model, so as to obtain an expressway road domain risk prediction result; the expressway domain risk prediction model is obtained by training a combined model according to expressway domain risk prediction sample data, wherein the combined model takes a fully-connected neural network as a meta classifier, a Bayesian network, a support vector machine and K neighbors as base classifiers, and each base classifier is respectively connected with the meta classifier; the optimizing unit 403 is configured to obtain a road domain risk optimization prediction result of the expressway according to the output result of the statistical model and the expressway road domain risk prediction result obtained in advance.
The expressway road domain risk prediction device provided by the embodiment of the invention acquires multi-source data related to the road domain risk prediction of an expressway; performing road domain risk prediction on the multi-source data based on a preset expressway road domain risk prediction model to obtain an expressway road domain risk prediction result; the expressway domain risk prediction model is obtained by training a combined model according to expressway domain risk prediction sample data, wherein the combined model takes a fully-connected neural network as a meta classifier, a Bayesian network, a support vector machine and K neighbors as base classifiers, and each base classifier is respectively connected with the meta classifier; according to the output result of the statistical model and the highway road domain risk prediction result which are obtained in advance, the road domain risk optimization prediction result of the highway is obtained, and the accuracy of road domain risk prediction of the highway can be improved.
The embodiment of the present invention provides an expressway road domain risk prediction device, which may be specifically used to execute the processing flow of each method embodiment, and the functions thereof are not described herein again, and may refer to the detailed description of the method embodiments.
Fig. 5 is a schematic diagram of an entity structure of an electronic device according to an embodiment of the present invention, as shown in fig. 5, where the electronic device includes: a processor (processor) 501, a memory (memory) 502, and a bus 503;
wherein, the processor 501 and the memory 502 complete communication with each other through a bus 503;
The processor 501 is configured to invoke the program instructions in the memory 502 to perform the methods provided in the above method embodiments, for example, including:
acquiring multi-source data related to road domain risk prediction of the expressway;
performing road domain risk prediction on the multi-source data based on a preset expressway road domain risk prediction model to obtain an expressway road domain risk prediction result;
The expressway domain risk prediction model is obtained by training a combined model according to expressway domain risk prediction sample data, wherein the combined model takes a fully-connected neural network as a meta classifier, a Bayesian network, a support vector machine and K neighbors as base classifiers, and each base classifier is respectively connected with the meta classifier;
And obtaining a road domain risk optimization prediction result of the expressway according to the output result of the statistical model and the expressway road domain risk prediction result which are obtained in advance.
The present embodiment discloses a computer program product comprising a computer program stored on a non-transitory computer readable storage medium, the computer program comprising program instructions which, when executed by a computer, are capable of performing the methods provided by the above-described method embodiments, for example comprising:
acquiring multi-source data related to road domain risk prediction of the expressway;
performing road domain risk prediction on the multi-source data based on a preset expressway road domain risk prediction model to obtain an expressway road domain risk prediction result;
The expressway domain risk prediction model is obtained by training a combined model according to expressway domain risk prediction sample data, wherein the combined model takes a fully-connected neural network as a meta classifier, a Bayesian network, a support vector machine and K neighbors as base classifiers, and each base classifier is respectively connected with the meta classifier;
And obtaining a road domain risk optimization prediction result of the expressway according to the output result of the statistical model and the expressway road domain risk prediction result which are obtained in advance.
The present embodiment provides a computer-readable storage medium storing a computer program that causes the computer to execute the methods provided by the above-described method embodiments, for example, including:
acquiring multi-source data related to road domain risk prediction of the expressway;
performing road domain risk prediction on the multi-source data based on a preset expressway road domain risk prediction model to obtain an expressway road domain risk prediction result;
The expressway domain risk prediction model is obtained by training a combined model according to expressway domain risk prediction sample data, wherein the combined model takes a fully-connected neural network as a meta classifier, a Bayesian network, a support vector machine and K neighbors as base classifiers, and each base classifier is respectively connected with the meta classifier;
And obtaining a road domain risk optimization prediction result of the expressway according to the output result of the statistical model and the expressway road domain risk prediction result which are obtained in advance.
It will be appreciated by those skilled in the art that embodiments of the present invention may be provided as a method, system, or computer program product. Accordingly, the present invention may take the form of an entirely hardware embodiment, an entirely software embodiment or an embodiment combining software and hardware aspects. Furthermore, the present invention may take the form of a computer program product embodied on one or more computer-usable storage media (including, but not limited to, disk storage, CD-ROM, optical storage, and the like) having computer-usable program code embodied therein.
The present invention is described with reference to flowchart illustrations and/or block diagrams of methods, apparatus (systems) and computer program products according to embodiments of the invention. It will be understood that each flow and/or block of the flowchart illustrations and/or block diagrams, and combinations of flows and/or blocks in the flowchart illustrations and/or block diagrams, can be implemented by computer program instructions. These computer program instructions may be provided to a processor of a general purpose computer, special purpose computer, embedded processor, or other programmable data processing apparatus to produce a machine, such that the instructions, which execute via the processor of the computer or other programmable data processing apparatus, create means for implementing the functions specified in the flowchart flow or flows and/or block diagram block or blocks.
These computer program instructions may also be stored in a computer-readable memory that can direct a computer or other programmable data processing apparatus to function in a particular manner, such that the instructions stored in the computer-readable memory produce an article of manufacture including instruction means which implement the function specified in the flowchart flow or flows and/or block diagram block or blocks.
These computer program instructions may also be loaded onto a computer or other programmable data processing apparatus to cause a series of operational steps to be performed on the computer or other programmable apparatus to produce a computer implemented process such that the instructions which execute on the computer or other programmable apparatus provide steps for implementing the functions specified in the flowchart flow or flows and/or block diagram block or blocks.
In the description of the present specification, reference to the terms "one embodiment," "one particular embodiment," "some embodiments," "for example," "an example," "a particular example," or "some examples," etc., means that a particular feature, structure, material, or characteristic described in connection with the embodiment or example is included in at least one embodiment or example of the invention. In this specification, schematic representations of the above terms do not necessarily refer to the same embodiments or examples. Furthermore, the particular features, structures, materials, or characteristics described may be combined in any suitable manner in any one or more embodiments or examples.
The foregoing description of the embodiments has been provided for the purpose of illustrating the general principles of the invention, and is not meant to limit the scope of the invention, but to limit the invention to the particular embodiments, and any modifications, equivalents, improvements, etc. that fall within the spirit and principles of the invention are intended to be included within the scope of the invention.
Claims (6)
1. The highway road domain risk prediction method is characterized by comprising the following steps of:
acquiring multi-source data related to road domain risk prediction of the expressway;
performing road domain risk prediction on the multi-source data based on a preset expressway road domain risk prediction model to obtain an expressway road domain risk prediction result;
The expressway domain risk prediction model is obtained by training a combined model according to expressway domain risk prediction sample data, wherein the combined model takes a fully-connected neural network as a meta classifier, a Bayesian network, a support vector machine and K neighbors as base classifiers, and each base classifier is respectively connected with the meta classifier;
Obtaining a road domain risk optimization prediction result of the expressway according to the output result of the statistical model and the road domain risk prediction result of the expressway, which are obtained in advance;
training a combined model according to expressway road domain risk prediction sample data to obtain an expressway road domain risk prediction model, wherein the method comprises the following steps:
preprocessing the acquired data;
extracting features according to the preprocessed data;
performing feature optimization on the extracted features, and performing feature fusion on the features after feature optimization;
Training a combined model according to the characteristics after the characteristic fusion to obtain the expressway road domain risk prediction model;
the preprocessing of the acquired data comprises:
Collecting data to obtain a data set; the dataset includes a first type of data having an unstructured data type and a second type of data having a log data type;
Carrying out structural processing on the first type of data, and carrying out analysis and secondary calculation on the second type of data to obtain traffic flow data;
Cleaning and formatting the data in the data set to obtain standardized data, and dividing the standardized data to obtain a training set, a verification set and a test set;
the feature extraction according to the preprocessed data comprises the following steps:
oversampling is performed on the data in the training set, and no oversampling is performed on the data in the test set;
selecting a feature to be selected related to road domain risk prediction from the training set, the validation set and the test set;
The data set comprises video monitoring image data, traffic flow data and structured data; accordingly, the selecting the candidate features related to road domain risk prediction from the training set, the verification set and the test set includes:
Performing feature extraction on a first feature to be selected corresponding to the video monitoring image data by using a convolutional neural network;
Extracting features of a second to-be-selected feature corresponding to the traffic flow data by using a cyclic neural network or a long-short-term memory network;
and extracting the characteristics of the third candidate characteristics corresponding to the structured data by using a fully connected neural network or an embedded layer.
2. The method for predicting the risk of an expressway according to claim 1, wherein the predicting the risk of the road domain for the multi-source data based on the preset expressway risk prediction model to obtain the prediction result of the risk of the expressway comprises:
And carrying out data preprocessing and feature extraction on the multi-source data, and carrying out road domain risk prediction on the extracted features based on the preset expressway road domain risk prediction model to obtain an expressway road domain risk prediction result.
3. The method for predicting the risk of an expressway road domain according to claim 1, wherein the training the combined model according to the features after feature fusion to obtain the expressway road domain risk prediction model comprises:
training each base classifier with the training set, and training the meta classifier with the verification set;
adding a plurality of output layers into the combined model, and simultaneously predicting prediction results of different combinations;
The combined model is trained using a back propagation algorithm and a gradient descent optimizer.
4. An expressway road area risk prediction apparatus, comprising:
the acquisition unit is used for acquiring multi-source data related to road domain risk prediction of the expressway;
the prediction unit is used for predicting the road domain risk of the multi-source data based on a preset expressway road domain risk prediction model to obtain an expressway road domain risk prediction result;
The expressway domain risk prediction model is obtained by training a combined model according to expressway domain risk prediction sample data, wherein the combined model takes a fully-connected neural network as a meta classifier, a Bayesian network, a support vector machine and K neighbors as base classifiers, and each base classifier is respectively connected with the meta classifier;
The optimization unit is used for obtaining a road domain risk optimization prediction result of the expressway according to the output result of the statistical model and the expressway road domain risk prediction result which are obtained in advance;
The expressway road domain risk prediction device is further used for:
training a combined model according to expressway road domain risk prediction sample data to obtain an expressway road domain risk prediction model, wherein the method comprises the following steps:
preprocessing the acquired data;
extracting features according to the preprocessed data;
performing feature optimization on the extracted features, and performing feature fusion on the features after feature optimization;
Training a combined model according to the characteristics after the characteristic fusion to obtain the expressway road domain risk prediction model;
the preprocessing of the acquired data comprises:
Collecting data to obtain a data set; the dataset includes a first type of data having an unstructured data type and a second type of data having a log data type;
Carrying out structural processing on the first type of data, and carrying out analysis and secondary calculation on the second type of data to obtain traffic flow data;
Cleaning and formatting the data in the data set to obtain standardized data, and dividing the standardized data to obtain a training set, a verification set and a test set;
the feature extraction according to the preprocessed data comprises the following steps:
oversampling is performed on the data in the training set, and no oversampling is performed on the data in the test set;
selecting a feature to be selected related to road domain risk prediction from the training set, the validation set and the test set;
The data set comprises video monitoring image data, traffic flow data and structured data; accordingly, the selecting the candidate features related to road domain risk prediction from the training set, the verification set and the test set includes:
Performing feature extraction on a first feature to be selected corresponding to the video monitoring image data by using a convolutional neural network;
Extracting features of a second to-be-selected feature corresponding to the traffic flow data by using a cyclic neural network or a long-short-term memory network;
and extracting the characteristics of the third candidate characteristics corresponding to the structured data by using a fully connected neural network or an embedded layer.
5. An electronic device comprising a memory, a processor and a computer program stored on the memory and executable on the processor, characterized in that the processor implements the steps of the method of any one of claims 1 to 3 when the computer program is executed by the processor.
6. A computer readable storage medium, on which a computer program is stored, characterized in that the computer program, when being executed by a processor, implements the steps of the method of any of claims 1 to 3.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN202410413853.1A CN118015839B (en) | 2024-04-08 | 2024-04-08 | Expressway road domain risk prediction method and device |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN202410413853.1A CN118015839B (en) | 2024-04-08 | 2024-04-08 | Expressway road domain risk prediction method and device |
Publications (2)
Publication Number | Publication Date |
---|---|
CN118015839A CN118015839A (en) | 2024-05-10 |
CN118015839B true CN118015839B (en) | 2024-07-09 |
Family
ID=90950974
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN202410413853.1A Active CN118015839B (en) | 2024-04-08 | 2024-04-08 | Expressway road domain risk prediction method and device |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN118015839B (en) |
Citations (2)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN109324604A (en) * | 2018-11-29 | 2019-02-12 | 中南大学 | A kind of intelligent train resultant fault analysis method based on source signal |
CN113256191A (en) * | 2021-07-15 | 2021-08-13 | 平安科技(深圳)有限公司 | Classification tree-based risk prediction method, device, equipment and medium |
Family Cites Families (6)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN111489553B (en) * | 2020-04-26 | 2022-02-25 | 百度在线网络技术(北京)有限公司 | Route planning method, device, equipment and computer storage medium |
CN115147155A (en) * | 2022-07-05 | 2022-10-04 | 西南交通大学 | Railway freight customer loss prediction method based on ensemble learning |
CN115565373B (en) * | 2022-09-22 | 2024-04-05 | 中南大学 | Expressway tunnel accident real-time risk prediction method, device, equipment and medium |
CN117133099A (en) * | 2023-08-28 | 2023-11-28 | 河南交院工程技术集团有限公司 | Automatic monitoring system for disaster of expressway high slope |
CN117238126A (en) * | 2023-08-29 | 2023-12-15 | 东南大学 | Traffic accident risk assessment method under continuous flow road scene |
CN117725537A (en) * | 2024-02-01 | 2024-03-19 | 深圳凯升联合科技有限公司 | Real-time metering data processing platform |
-
2024
- 2024-04-08 CN CN202410413853.1A patent/CN118015839B/en active Active
Patent Citations (2)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN109324604A (en) * | 2018-11-29 | 2019-02-12 | 中南大学 | A kind of intelligent train resultant fault analysis method based on source signal |
CN113256191A (en) * | 2021-07-15 | 2021-08-13 | 平安科技(深圳)有限公司 | Classification tree-based risk prediction method, device, equipment and medium |
Also Published As
Publication number | Publication date |
---|---|
CN118015839A (en) | 2024-05-10 |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN109508360B (en) | Geographical multivariate stream data space-time autocorrelation analysis method based on cellular automaton | |
CN110807562B (en) | Regional bridge risk prediction method and system | |
CN115578015B (en) | Sewage treatment whole process supervision method, system and storage medium based on Internet of things | |
CN109872003B (en) | Object state prediction method, object state prediction system, computer device, and storage medium | |
CN111861274A (en) | Water environment risk prediction and early warning method | |
Mokhtari et al. | Comparison of supervised classification techniques for vision-based pavement crack detection | |
CN112966714B (en) | Edge time sequence data anomaly detection and network programmable control method | |
CN114664091A (en) | Early warning method and system based on holiday traffic prediction algorithm | |
CN115148019A (en) | Early warning method and system based on holiday congestion prediction algorithm | |
KR102564191B1 (en) | Disaster response system that detects and responds to disaster situations in real time | |
CN115629160A (en) | Air pollutant concentration prediction method and system based on space-time diagram | |
CN116738192A (en) | Digital twinning-based security data evaluation method and system | |
Agarwal et al. | Camera-based smart traffic state detection in india using deep learning models | |
CN118015839B (en) | Expressway road domain risk prediction method and device | |
KR20220138250A (en) | A method and an electronic device for inferring occurrence of highly-concentrated fine dust | |
Yang et al. | Forecasting model for urban traffic flow with BP neural network based on genetic algorithm | |
CN115565388A (en) | Traffic light control method based on multi-channel vehicle detection and three-dimensional feature labeling | |
Li et al. | Information granularity with the self-emergence mechanism for event detection in WSN-based tunnel health monitoring | |
CN114078070A (en) | Multi-source data fusion text and travel safety monitoring and traceability analysis method and system | |
CN113936300A (en) | Construction site personnel identification method, readable storage medium and electronic device | |
Alshawabkeh et al. | Automated Pavement Crack Detection Using Deep Feature Selection and Whale Optimization Algorithm. | |
Hong et al. | Drainage network flow anomaly classification based on XGBoost | |
Venugopal | SHM for Intelligent Transportation Infrastructure using Machine Learning and AI-A Systematic | |
CN117975659B (en) | In-car child behavior monitoring system utilizing automobile microcontroller chip | |
CN117216722B (en) | Sensor time sequence data-based multi-source heterogeneous data fusion 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 | ||
GR01 | Patent grant | ||
GR01 | Patent grant |